Advances in Insect Physiology, Volume 29

Contributors S. Caveney D(7~m'lment ~/Zoo/ogy, University 0[ Western Ontario, Londml, Ontario N6A 5B7, Camula G. M. Coast D~Tmrtment ~?/"Biology, Birkbeek ( Universitv ~/ London ), Malet Street, Lottdon WC1E 7tt)C UK B. C. Donly Southern Crop Protection am/Food Research Centre, A~rieulture and A~ri-f~md Canadtt, Lomlon. Ontario, Canada N5V 4T3 M. J. Hall Deparmtent qf Biological Sciences, The Open University, Milton Keynes, MK7 6AA, UK M. L. Hudson D~7~armlent o/ Biologie~tl Structure and Function, 611, S W Cam~ms Drive, SD, Ores.on Health and Sciences University, Port[and, OR 97201. USA D. B. M o r t o n

Del~arlment qf Biological Struetm'e and k)mction, 611. S W Campus Drive, SD, Ore,~on Health and Sciences University, Portland, OR 97201. USA I. Orchard

Universilv ¢?/' Toro/lto, Department ~?/ Zoology, 25 Harbor Street, Toro/tm, Ontario M5S 3G5. Can~ttkt J. E. Phillips Universil|' o/ British CohmdEa, DC/)~IPllIIClZl ()/ Zoolo£,yl', ~/alleozlvel', British Colmnhia V6 T 1Z4, Cana~kt D. J. Robinson D~7~artment ~?/ Biolo~zical Science.,,'. The Opetl Universill', ,]/[illoll Keynes, MK7 6AA. ("K D. A. Schooley DeparmTent ~?/Biochemistry ( 330 ) , Universi O, ~?/Nevada, N V 89557-0014. USA

Cyclic GMP Regulation and Function in Insects David B. Morton and Martin L. Hudson Department of Biological Structure and Function, 611, SW Campus Drive, SD, Oregon Health and Sciences University, Portland, OR 97201, USA

1 Introduction 2 2 Cyclic GMP regulation 2 2.1 Guanylyl cyclases 2 2.2 Phosphodiesterases 22 3 Cyclic GMP function 26 3.1 Molecular targets 26 3.2 Physiological functions 32 4 Concluding remarks 43 Acknowledgements 44 References 44

Abstract Guanosine 3'5' cyclic monophosphate (cGMP) mediates a wide variety of physiological processes in many invertebrate and vertebrate species. Here we discuss our current understanding of c G M P regulation and function in insects, reviewing components of the c G M P signaling cascade and some of the major physiological roles played by c G M P in insects. The recent completion of the Drosophih~ genome project has enabled us to identify all the potential elements of the c G M P signaling cascade in a single insect. Most of these proteins have not been previously characterized, and by comparing their predicted sequences with identified proteins from other species (insects and mammals) we describe their expected properties. The list of potential proteins that regulate c G M P levels includes five receptor guanylyl cyclases (GCs), two receptor-like GCs, five soluble GC subunits, a possible GC-activating peptide, five possible GCAP-like proteins and five phosphodiesterases that are predicted to hydrolyze cGMP. Downstream elements of c G M P signaling include two phosphodiesterases that could be regulated by cGMP, three cGMP-dependent protein kinases and two ion channels that could be regulated by cGMP. ,M)VANC'ES IN INSECT PHYSIOLO(IY ]SBN I) 12 I)24229 X

VOL

29

('~)l~Ir~i~llt ~ 2002 E/~uw~,r S~iemc L i d 41/~i~,IiI~ ql r~Tn'nduction m aHl I m m ~'~'~'p vud

2

D.B. MORTON AND M. L. HUDSON

1 Introduction The intracellular messenger, guanosine 3'5' cyclic monophosphate (cGMP) was first discovered shortly after the seminal work of Sutherland and Rall (1957) had established the concept of second messengers (Ashman et al., 1963). Since that time many studies in a wide variety of tissues and organisms have demonstrated the crucial role that c G M P plays in many physiological processes. While the majority of these studies have been in vertebrate preparations. there have also been several major contributions to c G M P research that have utilized insects. These include one of the first demonstrations of a cGMP-dependent protein kinase (PKG) (Kuo and Greengard, 1970), the first identification of nitric oxide (NO)-insensitive soluble guanylyl cyclases (GCs) (Nighorn el al., 1999; Simpson et al., 1999), and the demonstration that a naturally occurring behavioral polymorphism is due to mutations in a P K G gene (Osborne et al., 1997). This review describes the molecular components of the c G M P signaling system that have been identified in insects, and surveys the newly available genomic sequence of Drosophila to demonstrate the variety of genes that might be involved in the regulation and function of cGMP in a single organism. In addition, we review the signaling cascades of several important physiological processes in insects that are regulated by cGMP.

2

Cyclic GMP regulation

The intracellular concentration of c G M P in a cell is regulated by a balance between its rates of synthesis and breakdown. The synthesis of c G M P from GTP is catalyzed by GCs, and c G M P is hydrolyzed by phosphodiesterases (PDEs) to form GMP. There are several classes of both enzymes, and the activity of both is regulated by a wide variety of factors. Molecular cloning has identified several members of GCs and PDEs in insects, and the recent completion of the Drosophila Genome Project has now allowed a more complete description of the variety of possible regulatory mechanisms for controlling c G M P levels in insects. 2. l

GUANYLYLCYCLASES

Early studies investigating GC enzyme activity revealed that most mammalian tissues contained both cytoplasmic and membrane-bound GC activity (Tremblay et al., 1988). Although GC activity has been measured in a number of insect tissues, there have been few studies that have separated cytoplasmic and membrane-bound activity. B o m b y x mori fat body contains exclusively particulate GC (Morishima, 1981) whereas silkmoth antennae and M a n d u c a sexta CNS (central nervous system) contain both soluble and particulate activities (Ziegelberger et al., 1990; Morton and Giunta, 1992). The initial

CYCLIC GMP REGULATION AND FUNCTION

3

purification and cDNA cloning of sea urchin and mammalian particulate GC and mammalian cytoplasmic GC revealed the fundamental structural difl'erences between these two types of GC (see Lucas et al., 2000). These are classified as receptor GCs and soluble GCs respectively. Until recently, all subsequently identified GCs could be assigned to one or the other of these classes, but studies reported in the last few years, particularly in invertebrates, have revealed that GCs are a far more diverse family of enzymes than previously thought. 2.1.1

Receptor GCs

2.1.1.1 Sequence analysis. Receptor GCs are integral membrane proteins that were first identified as the proteins associated with the membrane-bound GC activity described above. Seven different isoforms have been cloned in vertebrates. In rats they are named GC-A through GC-G (Garbers and Lowe, 1994; Ffille e! al., 1995; Lowe et al., 1995: Schulz et al.. 1998). Functional studies show that they can be further subdivided into two groups GC-A, GC-B and GC-C that are activated by extracellular ligands, and the retinal GCs, GC-E and GC-F that are activated by intracellular calcium-binding proteins (Lucas et ell., 2000). Sequence comparisons suggest that GC-G belongs to the former group and GC-D the latter (Lucas et al., 2000). BLAST analysis of the Drosophila genome reveals five genetic loci that are predicted to code for receptor GCs and reverse transcriptase-PCR (RT-PCR) with degenerate oligonucleotide primers has identified an additional member of this class in Mamluca, named MsGC-11 (A. Nighorn and D. B. Morton, unpublished data). A schematic diagram of the predicted protein structure of the insect receptor GCs is shown in Fig. 1. The predicted amino acid sequences share the same molecular features: a variable extracellular domain, a single transmembrane domain, an ATP binding/protein kinase-like domain and highly conserved dimerization and catalytic domains. No functional studies have been carried out on the insect receptor GCs and so whether they can be subdivided into the same functional subclasses as the vertebrate receptor GCs remains to be determined. GC-D, GC-E and GC-F are all expressed in sensory neurons, GC-D in olfactory receptor neurons and GC-E and GC-F in photoreceptors (Lucas et ell., 2000). Interestingly, a large family of receptor GCs has been identified in the nematode, Caenorhabdilis elegans, many of which are expressed in olfactory sensory neurons (Yu et al., 1997). The expression patterns of two of the Drosophila receptor GCs have been reported and are not specifically expressed in sensory neurons. Gyc32E is primarily expressed during oogenesis in the ovarioles (Gigliotti et al., 1993) and Gyc76C is expressed in a wide range of tissues, including optic lobe, central brain, thoracic ganglia, digestive tract, oocytes and muscles (Liu et al., 1995; McNeil el al., 1995). Sequence comparison of the insect receptor GCs with the mammalian receptor GCs does not immediately suggest that insect receptor GCs fall into

4

D. B. M O R T O N A N D M. L. H U D S O N

c-

E

A

,

Extracellular

Rat GC-A

)i

i i

Gyc76C

Ill II I

MsGC-II

)l J

o Kinase

~ dike

~_

C3

~11 I

I immmlm

)tim I

I

mJ

i

t

CG3216

])1 II

!

! A

I~lll

J)~ m

L

i711{

CG4224

CG 10738

~ m

A

Gyc32E

Catalytic

_

III

i!!ilmmmlmm [

J

i

~

H

B MsGC-I CG5719 CG9783

: ~ l

FIG. 1 Schematic representation of the insect receptor GCs and receptor-like GCs. A Receptor GCs. The insect receptor GCs are shown in comparison with a mammalian receptor GC, rat GC-A. MsGC-II is from Manduca sexta (A. Nighorn and D. B. Morton, unpublished data) and the other receptor GCs are from Drosophila, either previously published sequences (Gyc76C Liu et ell., 1995; McNeil et al., 1995; Gyc32E Gigliotti et al., 1993) or identified from the Drosophila genome sequence using a BLASTP analysis. Each of the five structural domains are shown and are aligned at the predicted transmembrane domain, the position of which was predicted using PSORT II (www.expasy.ch/tools/). The dimerization and catalytic domains were predicted by alignment with GC-A, using CLUSTALW. In the extracellular domains, the vertical lines represent the positions of cysteine residues, and the /x near the transmembrane domain represents the conserved proline-rich juxtamembrane hinge region (see text). Possible ATP binding sites in the kinase-like domain are shown as short, shaded boxes. B Receptor-like GCs. These GCs have catalytic and dimerization domains that are similar to receptor GCs but have no transmembrane or extracellular domains. The only example whose expression has been demonstrated is MsGC-I (Simpson et ell., 1999), but two sequences with a similar predicted structure are present in the Drosophila genome.

CYCLIC GMP REGULATION AND FUNCTION

5

these two functional groups. Figure 2 shows a phylogenetic dendrogram and reveals that the mammalian "sensory receptor GCs" form a separate cluster that does not include any of the insect receptor GCs. The natriuretic peptide receptors, G C - A and GC-B form a cluster that also contains the Drosophila GC, Gyc32E, and the guanylin/uroguanylin/heat stable enterotoxin receptor, GC-C, forms a separate group with CG4224 and CG3216. The mammalian orphan receptor GC, G C - G , and the other insect receptor GCs tk~rm separate branches from these clusters. Throughout the sequences there are several conserved features and domains, which have been reported to determine specific functional aspects of G C activity. The most extensive structure function analysis has been carried out on GC-A and is used here as the primary comparison. GC-A is a receptor G C that binds and is activated by the peptide hormone, atrial natriuretic peptide (ANP). There is relatively little primary sequence similarity in the extracellular domain between any of the vertebrate and insect receptor GCs, but there are a

Gyc76C

L

MsGC-II Rat GC-G

__~ Rat GC-F

[

Rat GC-E Rat GC-D

[_ Rat GC-A Rat GC-B Gyc32E

L CG4224 Rat GC-C CG3216 CG10738 FIG. 2 Phylogenetic dendrogram of the insect and rat receptor GCs. The dendrogram was compiled using the protein sequence parsimony method that infers an unrooted phylogeny (bioweb.pasteur.fl/seqanal/phylogeny/phylip-uk.html).

6

D. B. MORTON AND M. L. HUDSON

number of features that are conserved. The extracellular domains of the mammalian receptor GCs are all of a similar size, varying between 413 and 486 residues. The insect receptor GCs generally have a similar size, 358--476 residues, although CG4224 is predicted to have a much shorter extracellular domain with only 190 residues. The extracellular domain of GC-A contains six cysteine residues that form three intramolecular disulfide bonds (Miyagi and Misono, 2000). The most N-terminal pair forms a disulfide bond between cys-60 and cys-86 (numbering based on the rat GC-A sequence) and is present in all mammalian and insect receptor GCs, with the exception of CG4224, which has a substantially reduced extracellular domain (Fig. 1). The other highly conserved disulfide bond is close to the transmembrane domain (cys423 and cys-432 in GC-A) in a region known as the juxtamembrane hinge (Huo et al., 1999). This region is also notable for being rich in proline residues, in particular pro-419 that is essential for ANP-stimulated GC activity, but not ANP binding (Huo et al., 1999). Replacement of both the cysteine residues in the hinge region with serines also had no effect on ANP binding, but resulted in a constitutively active GC (Huo et al., 1999). These results suggest that this hinge region mediates transmembrane signal transduction for the peptideactivation of GC-A (Huo et al., 1999). Cysteine and proline residues in the equivalent positions are found in all vertebrate receptor GCs, with the exception of GC-C (Huo et al., 1999) and are found in all the insect receptor GCs, with the exception of CG3216 and MsGC-II (Fig. l). All mammalian receptor GCs, except GC-F, contain at least one consensus N-linked glycosylation site (Lucas et al., 2000). All the insect receptor GCs also contain these sites, which vary in number from two in CG3216 to nine in Gyc76C (not shown). Even the reduced extracellular domain of CG4224 contains 6 consensus N-linked glycosylation sites. A number of studies have shown that glycosylation of GC-A, GC-B and GC-C is necessary for ligand binding and activation (see Lucas et al., 2000). The other notable feature of the extracellular domain of GC-A is a binding site for chloride ions in the ANP binding domain, which was revealed when the crystal structure of the GC-A extracellular domain was solved (Van den Akker et al.. 2000). ANP binding is enhanced in the presence of chloride ions (Van den Akker et al., 2000), and it has been suggested that chloride-mediated ANP binding provides a feedback control mechanism for salt regulation (Misono, 2000). Residues that form this chloride-binding site are conserved in CG3216, suggesting that binding and hence activation by its ligand could also be regulated by chloride. Following the transmembrane domain is the kinase-like domain, so called because it contains a consensus ATP-binding site found in all protein kinases (Lucas et al., 2000). However, an essential aspartate residue required for kinase catalytic activity is absent in all receptor GCs (Lucas et al., 2000), including all the insect receptor GCs, and there is no evidence that this domain has kinase activity in any GC. The insect receptor GCs, however, do contain consensus ATP binding sites, although they are not all in the equivalent position to that

CYCLIC GMP REGULATION AND FUNCTION

7

found in GC-A (Fig. 1). The ATP binding site is an important regulatory domain in mammalian receptor GCs. ATP potentiates the effect of natriuretic peptides on GC-A and GC-B, and alterations in the ATP-binding site eliminate this effect (Duda et al., 1993). In addition, the ATP-binding site in GC-A contains six serine and threonine residues that are phosphorylated in the unstimulated receptor and are essential for peptide-activation of GC-A (Potter and Hunter, 1998). Activation of GC-A by ANP results in dephosphorylation and desensitization of the receptor (Potter and Hunter, 1998). All the insect receptor GCs have one or more serine or threonine residues in close proximity to their consensus ATP-binding site that might act in the same manner. The dimerization and catalytic domains show the highest sequence conservation between mammalian and insect receptor GCs. The dimerization domain, identified in GC-A as a sequence of 42 residues that forms an amphipathic helix (Wilson and Chinkers, 1995), shows 4 7 76% identity with the equivalent region in the insect receptor GCs. Analyzing this region for protein secondary structure reveals that in all the insect receptor GCs this region is also predicted to form an c~ helix (D. B. Morton, unpublished observations). Similarly, the catalytic domain is highly conserved, with the insect receptor GCs sharing 60 69% sequence identity with GC-A. Most importantly, all the residues that have been predicted to form the binding site for the Mg-GTP substrate (Liu et al., 1997) are 100% conserved in all insect receptor GCs. The C-terminal tail of receptor GCs, which immediately follows the catalytic domain, is extremely variable in its length and primary sequence. GC-A and GC-B have no extension beyond the catalytic domain, whereas the other mammalian receptor GCs have an additional 40 60 residues. The insect receptor GCs vary from having no extension in Gyc32E to 443 residues in Gyc76C. Little is known concerning potential functions for this domain, although there is speculation that for GC-C, -E and -F it may be involved in the association of the GC with the cytoskeleton (Lucas et al., 2000). In addition, GC-C undergoes ligand-mediated endocytosis that might be mediated by a YXXZ motif, where Z is L, l, V, M, C or A (Lucas et al., 2000). A similar motif is found in the C-terminal domain of CG4224. Sequence homology between some insect GCs and the natriuretic peptide clearance receptor (NPR-C) indicate a possible novel method of signaling cross talk in these GCs. The NPR-C has an extracellular domain with homology to GC-A, a transmembrane domain and a short cytoplasmic region but no catalytic domain (Lucas et al., 2000). Several studies have shown that this receptor inhibits adenylyl cyclase via activation of Gi (Murthy and Makhlouf, 1999; Pagano and Anand-Srivastava, 2001). The motif that actiw~tes Gi has been identified as containing two basic residues at the amino terminus and a motif that contains BBXXB, BBXB or BXB at the C-terminus, where B represents a basic residue and X is any other residue (Murthy and Makhlouf, 1999: Pagano and Anand-Srivastava, 2001). Interestingly, the Drosophila GCs, Gyc76C, CG10738 and CG9783 also contain these motifs.

8

D.B. MORTON AND M. L. HUDSON

It will be interesting to determine whether they are capable of activating Gi and hence possibly interacting with an additional signaling cascade. 2.1.1.2 Ligands and activators. There are several peptide hormones that are known to activate mammalian receptor GCs. The natriuretic peptides ANP, BNP and C N P activate GC-A and -B, while G C - C is activated by guanylin, uroguanylin and the bacterial heat stable enterotoxin (Lucas et al., 2000). A BLAST analysis of the Drosophila genome with each of these peptides showed no significant matches (D. B. Morton, unpublished observations). This is not too surprising as there is little similarity in the ligand-binding domains between the mammalian and insect receptor GCs. There are, however, examples of invertebrate peptides that activate membranebound GCs. The crustacean hyperglycemic hormone ( C H H ) activates G C activity in membrane preparations of several crustacean tissues (Goy, 1990; Scholz et al., 1996) and this peptide is part of a large family of related peptides, several of which are found in insects. Figure 3A shows an alignment of the known members of this family found in insects and also includes a sequence identified in Drosophila (CG13586) using a BLAST search with CHH-B. The best characterized of these peptides is the ion transport peptide of Schistocerca gregaria (SgITP) (Meredith et al., 1996) and the almost identical peptide in Locusta migratoria (LmITP). SglTP stimulates salt and water reabsorption in the ileum and acts through an increase in cAMP levels (Chamberlin and Phillips, 1988). A closely related peptide ITP-L is generated by alternative splicing of the m R N A and is not active in the locust ITP ileum bioassay (Meredith et al., 1996) and hence might act through other pathways such as c G M P in this or other tissues. The orthologous Drosophila peptide is part of a predicted protein that is much larger than the locust prohormone (430 compared with 130 residues). This could however be due to inaccurate predictions FIG. 3 Possible ligands and activators of insect receptor GCs. A CLUSTALW alignment of insect peptides that have sequence similarity with the lobster crustacean hypoglycemic hormone B (HaCHH-B). The peptides shown are Bombyx mori CHHlike peptide (BmCHH-L) (Endo et al., 2000), the ion transport and ion transport-like (ITP and ITP-L) of Schistocerca gregaria and Locusta migratoria (Meredith el a/., 1996) and part of a sequence from a predicted protein identified in the Drosophila genome (CG13586). Note that in the predicted Drosophila sequence there is an insertion of 39 residues that is absent in the other sequences. B CLUSTALW alignment of insect GCAP-Iike proteins compared with GCAP-1 from mouse. All of the related proteins are from Drosophila and include frequenin (Frq) (Pongs el al., 1993), neurocalcin (Nca) (Teng et al., 1994) and three genes identified from the Drosophila genome. The predicted N-terminal myristoylation site and the four EF hands arc also indicated. The first EF hand is predicted to be non-functional due to a conserved proline residue at position 4 (see text). The abbreviations for the residues that make each EF hand are as follows, o oxygen in carboxyl side chain (D, N, Q, E, S or T), * any residue, g glycine, j leucine, isoleucine or valine and e glutamate or aspartate.

CYCLIC GMP REGULATION AND FUNCTION

9

u l C~ ~

© 1 1 1 1 1

C~

~ A ~ ~ 0 ~

H

H

H

H

l l l l 1

l l l l 1

l l l l 1

l l l l 1

l l l l 1

H

t

r~

? ~:ii;i

? I I l l

~ m

l l

l l

E l ~ l

l l

N

. . . . .

~

m I I

I

1

~

rn

g~ggg

~ Z U U ~

10

D.B. MORTON AND M. L. HUDSON

of the intron/exon structure of the gene, as there are 39 residues in the middle of the predicted mature peptide that are absent in any of the other family members. Several features that are conserved in the peptide family are present in the Drosophila peptide, and include the N-terminal dibasic cleavage site, all six cysteine residues that form three disulfide bonds in CHH and a possible amidation site at the C-terminus of the peptide. Whether this peptide is indeed expressed in Drosophila or whether it activates one of the receptor GCs remains to be determined. The other known activators of receptor GCs are the intracellular, calciumregulated GC-activating proteins (GCAPs) that activate the retinal GCs, GC-E and GC-F (Palczewski et al.o 2000). The GCAPs bind to the retinal GCs and activate them in the presence of low (< I#M) calcium (Dizhoor and Hurley, 1999). BLAST analysis of the Drosophila genome with the GCAP-1 sequence reveals the presence of five closely related genes (Fig. 3B), two of which, frequenin and neurocalcin, have previously been described (Pongs et al., 1993; Teng et al., 1994). The GCAPs and four of the Drosophila GCAP-like proteins have a consensus N-myristoylation site and all contain four EF-hand domains. EF-hand domains are 12 residue calcium-binding loops that contain two helices (Palczewski et al., 2000). Residues at positions 1, 3, 5 and 9 contain an oxygen atom in their side chain (usually from D, N, Q or E but sometimes S or T); position 8 is a leucine, isoleucine or valine and position 12 is glutamate or occasionally aspartate. A glycine residue at position 6 is required for the correct turn between the two helices. The first EF-hand in the GCAPs is nonfunctional, primarily because the proline in position 4 disrupts the formation of the first helix. All five of the Drosophila sequences share these features, as they appear to have a non-functional EF-I and functional EF-2, -3 and -4. The GCAPs are part of a subfamily of myristoylated calcium-binding proteins that contain four EF-hands (Palczewski et al., 2000). This family includes the GCAPs that activate retinal GCs in low-calcium (Palczewski et al., 2000), neurocalcin, which activates GC-E in elevated calcium (Kumar et al., 1999) and recoverin, which inhibits rhodopsin kinase (Palczewski et al., 2000) and does not activate the retinal GCs (Gray-Keller et al., 1993), A more distantly related EF-hand calcium binding protein, SI00B also activates the retinal GCs in elevated calcium (Palczewski e: al., 2000). Hence it is not straightforward to predict which, if any, of the Drosophila proteins will regulate receptor GC activity, and, if they do regulate GC activity, whether it will be in response to elevated or reduced calcium levels. Frequenin was first isolated as a shaker-like mutant that was characterized as having altered synaptic efficacy (Pongs et al., 1993). h~ vitro assays showed that frequenin activated bovine retinal GCs at low calcium concentrations (Pongs et al., 1993). CG5744 is 96% identical to frequenin and would be expected to have similar properties. Drosophila neurocalcin, however, does not show activity in an in vitro assay for retinal GC activation, and has been suggested to function as an inhibitor of G protein-coupled receptor kinases (Faurobert et al., 1996). The observed lack of

CYCLIC GMP REGULATION AND FUNCTION

11

retinal G C activation could, however, be due to poor species cross reactivity as mammalian neurocalcin will only activate G C - E and has no effect on G C - F ( K u m a r et al., 1999). Hence, it appears that there are candidate G C A P s in Drosophila and presumably other insects, but their respective receptor GCs, and whether they are positively or negatively regulated by increasing calcium levels, remain to be determined. 2.1.2

Soluble GCs

The other major class of GCs are soluble proteins that exist as heterodimers with one c~ and one/4 subunit (Lucas et al., 2000). Activation of these enzymes by extracellular signals requires the generation of an intermediate intracellular messenger, of which the most widespread is believed to be the gas, nitric oxide (NO) (Lucas et al., 2000). Production of NO is achieved by the calciumdependent activation of NO synthase (NOS) (Bredt and Snyder, 1994). NOS and N O / c G M P signaling in insects has recently been reviewed (Davies, 2000; Bicker, 200l). Each soluble G C heterodimer contains a single heme group, which contains a central Fe 2+ or Fe ~+ ion that is coordinated in the plane of the heme by four nitrogen atoms (Lucas et al., 2000). The iron is also coordinated by a histidine residue provided by histidine 105 of the fi subunit (Zhao el al., 1998) that is necessary for NO activation of the GC. Although no spectroscopic studies have been reported on insect soluble GCs, their high degree of sequence conservation and their similar biochemical properties suggest that they bind heine in a similar manner (Liu et al., 1995: Nighorn el al., 1998). 2.1.2.1 Sequence analys& and biochemical properties. Both ~ and fi subunits for soluble GCs from Drosophila, (Liu et al., 1995: Shah and Hyde, 1995) and Mamhwa, (Nighorn et al., 1998) have been cloned and expressed in heteroIogous cells and a fi subunit has been cloned from the mosquito Anopheles ,~ambiae (Caccone et al., 1999). Each of these subunits is similar in size and sequence to their mammalian homologs. The soluble G C subunits can each be divided into two functional domains, a C-terminal catalytic domain that is very similar to the catalytic domain of receptor GCs, and an N-terminal regulatory domain that binds the heine group. Some reports also designate a dimerization domain located between the catalytic and regulatory domain, but this is primarily by analogy with the dimerization domain of the receptor GCs and because the o~and fl subunits do dimerize. There have been no functional studies to show that this region mediates dimerization in soluble GCs, and there is no primary sequence similarity with the dimerization domain of receptor GCs. However, a region of about 40 residues located at the N-terminal side of the catalytic domain shows a high probability of forming a two-stranded coiled coil structure in both the M a m h w a and Drosophila o~ subunits of the soluble GCs (D. B. Morton, unpublished obserw~tions). Coiled coils are common structural

12

D.B. MORTON AND M. L. HUDSON

motifs that often mediate protein protein interactions and can be predicted using the M U L T I C O I L program (Wolf et al., 1997). Interestingly, neither of the/3 subunits that form heterodimers with these oe subunits shows equivalent coiled coil motifs suggesting that if this region is the dimerization domain, only one subunit needs to form this structure. The crystal structure of the adenylyl cyclase catalytic domain, which has significant sequence similarity to that of GCs, has been solved, and it has been revealed that the active site forms a wreath-like dimeric structure (Zhang et al., 1997). Based on this information, homology modeling has been used to predict the catalytic doraain structure of both soluble GCs and receptor GCs, and has predicted which residues bind the M g - G T P substrate (Liu el al., 1997). The principal difference between the soluble and receptor GCs is that the soluble GCs are heterodimers, whereas receptor GCs are homodimers. One subunit of each dimer binds to the guanosine moiety (the fi subunit for soluble GCs) and the other to the Mg 2+ and triphosphate moieties (Liu et al., 1997). These predictions are shown in Fig. 4, and the conservation of the residues that associate with M g - G T P in the insect subunits is shown in Table 1. Because the soluble GCs are heterodimers, residues in the A strand do not need to be conserved in fi subunits, and conversely, residues in the B strand do not need to be conserved in oe subunits. This is not the case for receptor GCs, where all critical residues are present in both subunits. This arrangement leads to the formation of a single G T P binding site in soluble GCs and two in receptor GCs (Lucas el al., 2000). These active site residues of the insect soluble G C subunits are 100% conserved compared with their mammalian homologs. One deviation in the model compared with the actual sequences is in residue 12 that predicts a serine in the A strand to bind to the }, phosphate, but there is an asparagine at the homologous position of both insect and mammalian ot subunits. The presence of serine in mammalian and insect/4 subunits suggests that the B strand in soluble GCs provides this residue. Similarly, critical residues have been identified in the regulatory region and are also conserved in the five insect soluble G C subunits (Fig. 5). These include histidine 105 in the/4 subunit that forms the axial ligand to the iron of the heme group (Zhao et al., 1998) and two cysteines (at positions 78 and 214 in the rat ill) subunit that are necessary for NO activation (Friebe et al., 1997). One notable difference between the regulatory domains is that although the M a n & l e a regulatory domain contains the same number of residues as mammalian/4 subunits (Nighorn el al., 1998), the Drosophila/4 subunit contains an insertion of 118 amino acids near its N-terminus (Shah and Hyde, 1995). This results in a/4 subunit of 86 kDa, which is larger than the o~ subunit, whereas all other/4 subunits are smaller than the ot subunits. Interestingly, the Anopheles/4 subunit also has an insertion of 91 residues in the same position, although there is no sequence conservation of the inserts between the two flies (Caccone et al., 1999).

CYCLIC GMP REGULATION AND FUNCTION 1 R/

3-F

531BL. "1

2 c 535B// 4-G 5ssB ,.., "-.t'%./536B H

....% 1 % O"

13

l

+

N

-II ~ ""

~.

SH

"

1 o.....o\ "j[.['.....o _/ ' ,~N,/

17

15

F """"----

4~6B

~ 5- v

-..,.../;,\

o

"%,

,. 14-,

J""6"H .... •

k/"

"-,

544B

O

NH2

/""~ ~----'~

"-. . . . . . .

i ".. :.k...._--7-D

o

.... P

................

t~o"

,\NH2"................. ;....

/

O ......... P

/ o,

, f

..--

;

!"°:;i~,b:::;i[" ...... 3-, ,/" ',..,

"'o\

O

.......... H O / ' ' ' "

9- R 574A

12 S 604A

\ ~o-D 484A E 607A

FIG. 4 Predicted catalytic domain for the soluble GCs. The model for the Mg-GTP binding site is derived from the crystal structure of adenylyl cyclase and is redrawn from Liu el a/. (1997) with permission from The National Academy of Sciences, USA, copyright (1997). The residues that are predicted to be in close association with Mg-GTP are numbered, clockwise l¥om 1 to 17, starting from the arginine at position 53l. The position in either the A strand or B strand is indicated and the amino acid number is given for the Man&~ca soluble GC, either MsGC-oH for the A strand or MsGC-fll for the B strand.

14

TABLE 1

D.B. MORTON AND M. L. HUDSON

Residues in the G T P binding d o m a i n of the insect soluble G C s

A Strand position

6

7

9

10

11

12

14

Model prediction

T

D

R

D

E

S

I

-

N N -

DGC-c~ 1 MsGC-cel DGC-fl I MsGC-fil A g G C-fi 1 MsGC-fi3 CG4154 CG14885 C G 14886

. O A G

.

. -

. T T T

O O G

M M

Y Y

-

E E

V V V V V

-

B Strand position

1

2

3

4

5

8

13

15

16

17

Model prediction

R

C

F

G

V

N

R

F

M

E

D G C - / ~ I

-

MsGC-~I AgGC-~I DGC-ce I MsGC-oel MsGC-fi3 GC4154 GC14885 G C 14886

K

-

. -

.

. -

-

T T

K K -

C C

. -

The positions of each of the residues and their predicted interactions with GTP are shown in Fig. 4. A dash indicates that the residues in the insect GCs are identical to the mammalian receptor GC, GC-A, which was used in the original rnodeling study (Liu et al., 1997).

T h e i n s e c t s o l u b l e G C s u b u n i t s also s h o w s i m i l a r b i o c h e m i c a l p r o p e r t i e s t o t h e i r m a m m a l i a n h o m o l o g s w h e n e x p r e s s e d in h e t e r o l o g o u s cells. T h e s e e x p e r i m e n t s h a v e b e e n c a r r i e d o u t w i t h b o t h t h e Drosophila ( S h a h a n d H y d e , 1995) a n d M a n d u c a ( N i g h o r n et al., 1998) s u b u n i t s . B o t h , like t h e m a m m a l i a n e n z y m e s , a r e o b l i g a t e h e t e r o d i m e r s , i.e. n o a c t i v i t y is s e e n if e i t h e r s u b u n i t is e x p r e s s e d o n its o w n , b u t a c t i v i t y c a n b e m e a s u r e d if b o t h s u b u n i t s a r e c o t r a n s f e c t e d i n t o cells. T h i s a c t i v i t y w a s s t i m u l a t e d in t h e p r e s e n c e o f N O d o n o r s . I n a d d i t i o n , t h e M a n d u c a h e t e r o d i m e r h a d a h i g h e r level o f b a s a l a c t i v i t y a n d l o w e r N O - s t i m u l a t e d a c t i v i t y w h e n M n - G T P w a s u s e d as a s u b strate compared with Mg-GTP, and NO-activation was blocked by the soluble GC inhibitor ODQ (lH-[1,2,4]oxadiazolo[4,3-a]quinoxalin-l-one) (Nighorn et al., 1999). B o t h o f t h e s e c h a r a c t e r i s t i c s h a v e also b e e n d e s c r i b e d f o r m a m m a l i a n s o l u b l e G C s ( L u c a s et al., 2000).

CYCLIC GMP REGULATION AND FUNCTION Cys-78

RatGC ~i 76 MusGC ~i 76 HumGC-~I BovGC-~I

medGC MsGC-~I Dgc-~

AgGC ~i MsGC-~3 CG4154 CG14886 CG14885

76 76 76 76 76 75 75 75 75 57

15 Cys- 214

His-105 203 203 203 203 203 201 323 294 198 198 215 197

KHLPi~ASVLFEI~ KHLPI~AHVLFEIF~ KMPS~DLNVFLELF~P KMPT~KLDVFLDLFP

FIG. 5 Alignment of portions of the regulatory domain of the soluble GC [4 and filike subunits from vertebrates and insects. The residues highlighted in black are the highly conserved cysteines (78 and 214) and histidine 105. Gray highlighting shows residues that are conserved (identical and similar) between vertebrates and at least one insect GC. Sequences shown for vertebrate GCs are for the fll subunits of rat, mouse (mus), human (hum), cow (boy) and medaka fish (reed). Insect sequences are flom Manduca (MsGC-fil and MsGC-fl3), Anopheles (AgGC-fil) and Drosophila (Dgc-fl, CG45154, CG14886 and CG14885).

A BLAST analysis of the Drosophila genome also identified three additional GCs that had significant sequence similarity with /-,{ subunits of soluble GCs. These are discussed further, together with a unique Manduca fi subunit, named MsGC-/-]3, that appears to be a member of a separate class of GC.

2.1.3

Atypical GCs

As described above, GCs have classically been divided into two classes receptor and soluble GCs. Recent studies, however, have identified three additional types of GC, two of which have been identified in Man&tea. One, named MsGC-I, is related to receptor GCs, but is not an integral membrane protein and lacks the extracellular and transmembrane domains characteristic of receptor GCs (Simpson el al., 1999). The other, named MsGC-fl3, is related to soluble GCs but can function as a homodimer and is NO-insensitive (Nighorn et al., 1999). The third new class of GC is structurally very different. It is a bifunctional protein that has an N-terminal ATPase domain and a Cterminal domain that is structurally similar to adenylyl cyclases containing 12 transmembrane domains (Linder el al., 1999). The C-terminal adenylyl cyclaselike domain contains catalytic domains similar to GCs, and recombinant proteins have GC activity and no adenylyl cyclase activity (Carucci et al., 2000; Linder et al., 2000). GCs with this structure have been found in the protozoa Paramecium, Tetrahynwna and Plasmodium. Recently, a GC was also cloned from Dictoslelium that also had a predicted topology similar to adenylyl cyclases (Roelofs et ell., 2001). There is no evidence that similar GCs are

16

D.B. MORTON AND M. L. HUDSON

f o u n d in insects or other multicellular organisms a n d these will not be discussed further. A s u m m a r y of the different classes of G C a n d their phylogenetic d i s t r i b u t i o n is shown in Figure 6.

Ligandbinding

Heme-

Trans- ~ . membrane

~ase i -like

Regulatory?

||

Dimerization Catalytic

~

binding

II A

B

II

C ATPaseDomain

GC Domain

D

j u U v ouov E

Catalytic Domain

FIG. 6 Schematic diagram of the different classes of GCs identified in insects and other organisms. A Receptor GCs are homodimeric integral membrane proteins with ligand-binding, transmembrane, kinase-like, dimerization and catalytic domains. They have been identified in insects, mammals, fish, birds, sea urchins, and C. elegans. B NO-sensitive soluble GCs are heterodimeric cytoplasmic enzymes with heme-binding and catalytic domains. They have been identified in insects, mammals and fish and are predicted to be absent in C. elegans (Morton et al., 1999). C Receptor-like GCs are similar to receptor GCs, but lack the ligand-binding and transmembrane domains. The only definitive example is in Manduca, but genomic sequences exist that predict their presence in Drosophila, C. elegans and the acidian Ciona hltesthmlis. D NO-insensitive soluble GCs are active as homodimers and have a similar domain structure to NO-sensitive soluble GCs, but are not activated by NO. Manduca is the only organism where conclusive evidence for their presence has been demonstrated, but they are predicted to also be present in Drosophila and C. elegans. E Multiple transmembrane domain GCs have been found in the protozoa and Dictostelium. The protozoan GCs have an ATPase domain and a GC domain with 12 transmembrane domains, whereas the Dictostelium GC only has the GC domain.

CYCLIC GMP REGULATION AND FUNCTION

17

M s G C - I and related GCs. MsGC-I was isolated using RT-PCR and degenerate oligonucleotides designed to hybridize to regions of the catalytic domain that are conserved between soluble and receptor GCs (Nighorn et al., 1998). Sequencing the PCR product suggested that MsGC-I was a receptor GCs; however, isolation and sequencing full-length cDNAs showed that it did not contain a signal sequence, extracellular or transmembrane domains (see Fig. 6). Northern blots showed that the eDNA was similar in size to the 2.5 kb transcript, suggesting that an incomplete cDNA had not been cloned. In addition, antisera made to a fusion protein recognized a 55 kDa protein on Western blots of nervous tissue the size of protein predicted from the sequence (Simpson et ~ll., 1999). The catalytic domain is 77% identical to the catalytic domain of GC-B, a mammalian receptor GC, compared with 33% identity with soluble GCs, and contains all the residues necessary for Mg-GTP binding. In addition, MsGC-I contains a region that is 76°/,, identical to the dimerization domain of GC-B (Simpson et al., 1999). A schematic diagram of the domains of MsGC-I is shown in Fig. lB. MsGC-I contains a C-terminal extension that shows no homology to any GC or any other protein in the databases, and a short Nterminal extension that shows a low level of similarity to receptor GC kinaselike domains but does not contain a consensus ATP binding site. There is no similarity to the regulatory domain of soluble GCs. MsGC-1 belongs to a new class of GC that is related to receptor GCs, but is not an integral membrane protein and cannot be activated directly by an extracellular ligand. There is a report of a mammalian GC cloned from rat kidney, ksGC, that also has the same structure as MsGC-I (Kojima et al., 1995). No GC activity was measured t¥om recombinant ksGC, and there was no evidence that a protein of the predicted size was indeed expressed in vivo. A subsequent report has suggested that ksGC was a cloning artifact and was a partial clone of the receptor GC, GC-G (Schulz et al., 1998). When MsGC-I was transiently expressed in COS-7 cells it showed a high level of basal GC activity, was not stimulated by NO, and was located in the soluble traction of COS-7 cell homogenates (Simpson et al., 1999). Further experiments with recombinant MsGC-I showed that it formed homodimers when expressed in COS-7 cells (Simpson et al., 1999). Although it was located in the soluble fraction of COS-7 cells, it was present in the particulate fraction of Manchu'a nervous tissue homogenates, suggesting that it was bound to another protein in vivo (Simpson e t a / . , 1999). hi situ hybridization and immunocytochemistry showed that MsGC-I was expressed in a small population of neurons in the posterior of each segmental ganglion of pre-pupal animals (Simpson et al., 1999). Analysis of the distribution of MsGC-I has also been extended to the brain and antennae of adult M a m h u ' a . MsGC-I showed extensive expression throughout the brain, especially in sensory neuropil of the olfactory and visual systems and the higher order neuropil of the mushroom bodies and central complex (Nighorn et al., 2001 ). Expression of MsGC-I was also seen in the cell

2.1.3.1

18

D.B. MORTON AND M. L. HUDSON

bodies and dendrites of the olfactory sensory neurons in the antennae (Nighorn et al., 2001). Several studies have noted the role of N O / c G M P signaling in olfaction and olfactory processing in insects (see section 3.2.1). The distribution of MsGC-I in the olfactory system raises the possibility that there is a parallel pathway of NO-independent signaling, although co-localization of both NO-sensitive GCs and MsGC-I in the same cells has yet to be demonstrated. A major question to address is the mechanism of MsGC-I regulation. It is not activated by NO and it cannot be directly activated by extracellular ligands. When MsGC-I was expressed in COS-7 cells and the accumulation of c G M P in intact cells measured, the level of c G M P present was higher compared with cells that were transfected with the NO-sensitive soluble GC and stimulated with NO (Simpson et al., 1999). It seems unlikely that MsGC-I exhibits this high level of basal activity in ~,ivo, unless it is co-expressed with a constitutively active PDE. As MsGC-I has a different subcellular distribution in COS-7 cells (cytoplasmic) compared with its distribution itl ri~,o (particulate fraction), it is tempting to speculate that iu vivo MsGC-I interacts with a membrane associated protein (possibly a receptor) and its activity is inhibited in this state. Activation could then be achieved by dissociation from this protein, yielding an active GC. Another possible mechanism of activation has been revealed recently by examining the expression of MsGC-I in peripheral neurons. The epidermis of M a n d u c a is innervated by a population of dendritic arborization neurons that are thought to be mechanosensory (Grueber and Truman, 1999). Most of these neurons responded to NO donors with an increase in c G M P (Grueber and Truman, 1999), and hence probably contain MsGC-od and MsGC-/?I. One specific neuron, ddaB, however, responded very weakly to NO but showed a large cGMP increase ill the presence of EGTA, which was assumed to reduce the levels of cytoplasmic calcium in the cell (Grueber and Truman, 1999). This neuron specifically stained with the antisera to MsGC-I, suggesting that a reduction in calcium might be involved in its activation in ~'iro (Grueber et al., 2001). This might involve a similar mechanism of activation described for the receptor GCs mediated by the GCAPs in mammalian retina (Dizhoor and Hurley, 1999, see section 2.1.1.2). In support of this model we showed that antisera that recognize the Drosophila GCAP ortholog frequenin, also stained ddaB (Grueber et al., 2001). MsGC-I is the only reported example of a GC with its particular structural characteristics whose in vivo expression has been demonstrated. Genetic loci that are predicted to code ['or proteins with similar structures to MsGC-I are present in C. elegans, Drosol)tfila and the Acidian Ciona intes'tinalis (D. B. Morton, unpublished observations). Although these sequences suggest that MsGC-I is not unique, experimental evidence of their expression is needed. Schematic diagrams of the two predicted Drosophihl proteins, CG5719 and CG9783 are shown in Fig, lB. The catalytic domains, like that of MsGC-I,

CYCLIC GMP REGULATION AND FUNCTION

19

are more similar to those of receptor GCs than soluble GCs, and all the residues involved in M g - G T P binding are again 100% conserved. Both CG5719 and CG9783 have regions with similarity to the dimerization domains of receptor GCs and both are predicted to form ~ helices. The two Drosophih~ sequences have a region N-terminal to the dimerization domain that is larger than the one found in M s G C - I and is similar in size to the kinase-like domains of receptor GCs. A search for protein motifs in this region of CG9783 identifies it as being similar to protein kinases, and a BLAST analysis shows that it has about 45% identity to the kinase-like domain of several mammalian receptor GCs. It contains a consensus ATP binding site that has two serine/threonine residues nearby that could be phosphorylated in a similar fashion to GC-A (Potter and Hunter, 1998). This region also lacks the critical aspartate residue present in all protein kinases but lacking in the kinase-like domain of all receptor GCs. By contrast, the N-terminal portion of CG5719 shows no similarity to the kinase-like domains of receptor GCs and shows no homology to any other protein. Additional studies are clearly required on the members of this novel class of G C to determine whether their predicted structures are correct, how they are activated and what role they play in regulating c G M P levels in specific cells.

2.1.3.2 MsGC-fi3. During the course of the study that identified the Mandm'a GCs, MsGC-c~I, MsGC-fll and MsGC-I, we isolated an additional G C that was also closely related to mammalian fi subunits. Two mammalian G C fi subunits have been identified, fil and fi2. The new Manduca fi subunit had a number of specific differences that led us to believe that it was not a homolog of either the fil or fi2 subunits, and we named it MsGC-fi3 (Nighorn et al., 1999). The principal novel structural feature that MsGC-fi3 contains is a C-terminal extension of 315 residues that has no similarity to any other protein in the databases (see Fig. 6). There are no identifiable domains in this Cterminal sequence, but there is a predicted C-terminal isoprenylation site and several possible protein phosphorylation sites throughout its length (Nighorn el al., 1999). Although the regulatory domain of MsGC-fi3 is similar to other/31 subunits (both insect and vertebrate), there are a number of specific differences. The most notable are the lack of two cysteines (78 and 214) that are 100% conserved in all other fl subunits and are necessary for NO activation (Friebe et al., 1997) (Fig. 5). This suggested that MsGC-fi3 would form a G C that is insensitive to NO activation. It does, however, contain a histidine at the equivalent position of histidine 105, believed to be the axial ligand for the heme group in soluble GCs (Zhao el al., 1998). The catalytic domain of MsGC-fi3 also revealed some interesting features. All of the residues that are predicted to associate with the M g - G T P substrate are conserved in MsGC-fi3, here it is more similar to the homodimeric receptor

20

D.B. MORTON AND M. L. HUDSON

GCs than the heterodimeric soluble GCs (Table 1). The A strand residues 6-T, 9-R and 10-D that are replaced by non-conservative substitutions in all other fll subunits are conserved in MsGC-fi3 and the B strand 8-N and 16-M that are replaced in all other c~ subunits are also conserved in MsGC-fiY This is particularly noteworthy for 6-T and 8-N that are involved in the condensation reaction of the 3' hydroxyl of the ribose to the oe phosphate group and for 9R and 10-D that are involved in stabilization of the metal-triphosphate moiety (Fig, 4). Thus MsGC-fl3 provides all the residues needed for binding Mg-GTP without the need of a second subunit, suggesting that it might be able to form active homodimers. Transient expression of MsGC-/~3 in COS-7 cells showed that both of the predictions based on sequence analysis were valid. Although all the NOsensitive soluble GCs that have been examined are obligate heterodimers, MsGC-fi3 showed significant basal activity in the absence of additional subunits (Nighorn et al., 1999) and subsequent gel filtration data strongly supported the prediction that MsGC-fi3 formed homodimers (D, B. Morton, unpublished data). When MsGC-fl3 was co-expressed with either of the other M a n d u c a soluble G C s , MsGC-oel or MsGC-fll, the total level of enzyme activity was slightly reduced, suggesting that heterodimers were formed and that they had either reduced or negligible activity compared with homodimeric MsGC-fl3 (Nighorn et al., 1999). The activity of MsGC-fi3 was also insensitive to activation by NO (Nighorn et al., 1999). Like other GCs, MsGC-fl3 showed higher levels of activity when manganese was present, as compared with magnesium, but in neither case was the activity stimulated by NO. In addition, co-expression with either MsGC-otl or MsGC-fil did not yield a NO-sensitive GC (Nighorn et al., 1999). These properties have not been reported for any other native GC, but when the rat/41 subunit was mutated to remove cysteines 78 and 214, and subsequently co-expressed with wild type rat otl, a NO-insensitive GC was produced (Friebe et al., 1997). In this case, however, NO-sensitivity could be restored by incubating the mutant [31/oel G C with excess heine (Friebe et al., 1997). By contrast, when heine reconstitution experiments were carried out with either MsGC-/43 expressed alone or co-expressed with MsGC~1, the activity was still unaffected by NO donors (Nighorn e t a / . , 1999), It is not clear whether MsGC-fi3 is unable to bind heme, or whether it does so in a configuration that renders it insensitive to NO. The soluble GC inhibitor, 1H[1,2,4]oxadiazolo[4,3-a]quinoxalin-l-one (ODQ), acts on NO-sensitive soluble GCs by oxidizing the ferrous heme to ferric heme (Zhao et al., 2000) and we found that MsGC-fl3 was insensitive to ODQ at concentrations up to 100 # M (Nighorn et al., 1999). This suggests that MsGC-fl3 does not require ferrous heme for activity, but spectroscopic studies will be needed to unequivocally determine the presence and nature of the heme group. The unusual properties of this novel GC raise several important questions. These include understanding the physiological functions that it mediates, its mechanism of activation and whether it forms homodimers or heterodimers in

CYCLIC GMP REGULATION AND FUNCTION

21

viro. A significant contribution to these issues came from the finding that MsGC-fi3 is expressed in cells that respond to the neuropeptide, eclosion hormone (EH). EH is a central component of the endocrine cascade that triggers ecdysis behavior (shedding of the old cuticle) at the end of the molt (see Chapman, 1998). A series of studies (discussed in more detail in section 3.2.3) have shown that EH acts through an increase in c G M P and probably activates a novel NO-insensitive soluble GC (Morton and Truman, 1985: Morton and Giunta, 1992; Morton 1996; Kingan el al., 1997). Three specific EH target cell populations have been identified: a population of about 50 neurons in the ventral nerve cord (VNC) that contain crustacean cardioactive peptide (CCAP) (Ewer et al., 1994), the Inka cell in the peripheral epitracheal glands (Zitnan el al., 1996) and a population of non-neuronal cells in the transverse nerve of the abdominal ganglia (STNR cells) (Hesterlee and Morton, 2000). RT-PCR experiments showed that MsGC-fi3 was expressed in the transverse nerve and the epitracheal glands (Nighorn et al., 1999) and immunocytochemistry showed expression of MsGC-fi3 in the STNR and Inka cells (Morton, 2000). These studies provide circumstantial evidence for the activation of MsGC-fi3 by EH, but as the sequence of MsGC-fi3 predicts that it is cytoplasmic and as EH is an extracellular hormone, this activation must involve intermediate signals. Possible pathways for this activation are discussed in section 3.2.3. MsGC-fi3 is the only GC with the sequence characteristics described above that has been demonstrated to be a NO-insensitive GC. There are, however, several other GCs with these same features present in a variety of organisms. The nematode, Caenorhabditis elegasls has seven genes that code for soluble GCs and none of them have a cysteine in the equivalent positions as cys78 and cys214 (Morton el al., 1999). The lack of an identifiable NOS gene in the genome of C. eh,,gans (Bargmann, 1998) and the finding that GC activity in ('. el
22

D.B. MORTON AND M. L. HUDSON

In addition to predicting that these three Drosophila GCs are insensitive to NO, it is also possible to predict whether they are likely to form homodimers or heterodimers. Table 1 shows the residues in the catalytic domains of each G C that are predicted to interact with Mg-GTP. CG4154 is 90% identical to MsGC-fi3 over the whole catalytic domain, and all the predicted M g - G T P binding residues are 100% conserved suggesting that it will form active homodimers. By contrast, CG14885 and CG14886 only share 51% identity with MsGC-fl3 in the catalytic domain. The M g - G T P binding residues on the B strand are all conserved in both CG14885 and CG14886, whereas this is not the case for the A strand. The residues at positions l0 and 14 are both conservative substitutions and hence could fulfill the same roles. The threonine at position 6 is substituted for a tyrosine, which could provide the hydrogen bond to the 3' hydroxyl of the ribose moiety, although it is considerably bulkier. Most importantly, the aspartic acid at position 7 is replaced by a methionine. This aspartate is one of three residues required to form the catalytic center of the enzyme (the others are 8-N and 13-R from the B strand) (Liu et al., 1997). Because CG14885 and CG14886 lack this aspartate, they are unlikely to be able to tbrm an active homodimer and would have to combine with a subunit that provided this residue. The only Drosophila genes that contain all of the required residues for the A strand are CG4154 and the Drosophila ce subunit D G C ~ I . In this regard it is interesting to note that in the locust, Schks'tocerca gregaria, the soluble G C oe subunit is sometimes localized to neurons that do not show NO-stimulated c G M P immunoreactivity ( O t t e t al., 2000). If a similar situation exists in Drosophila, it is possible that the active G C present in these cells is a DGCoel/CG14885 or DGCoel/CG14886 heterodimer that would be predicted to be NO insensitive. Further support for this model is provided by analyzing the sequences using the M U L T I C O I L program for predicting coiled coil motifs ( W o l f e / a l . , 1997). Both MsGC-tq3 and CG4154 contain a region of about 40 residues that have a high probability of forming a two-stranded coiled coil in the same position as that in the Manduca and Drosophila oe subunits of the soluble GCs, whereas neither CG14885 nor CG14886 are predicted to lbrm coiled coils in this region. Thus, if this region does mediate dimer formation and there needs to be at least one subunit that forms a coiled coil, it is again unlikely that CG14885 and CG14886 form homodimers, but they could form heterodimers with the oe subunit. 2,2

PHOSPHODIESTERASES

Both synthesis and breakdown regulate the c G M P levels in a cell. A large family of enzymes, the cyclic nucleotide phosphodiesterases (PDEs), mediates the breakdown of cGMP. In addition to reducing cellular c G M P levels by converting it to G M P , some PDE family members are also downstream effectors and are either stimulated or inhibited by cGMP. Studies of mammalian PDEs have identified at least 19 different genes that are grouped into 11

CYCLIC GMP REGULATION AND FUNCTION

23

separate families, which differ in their substrate specificity and regulatory properties (Table 2) (Soderling and Beavo, 2000). At the present time, only a single insect PDE gene has been identified the Drosophila learning and memory gene dunce (dnc) (Davis el al., 1995), but analysis of the Drosophila genome identifies six additional genetic loci that are predicted to code for PDEs. BEAST analysis of each of these sequences places them in six of the 11 families (Table 2). Two of the Drosophila genes appear to be specific for cAMP. dnc is a member of the PDE4 family and has been shown to preferentially hydrolyze cAMP (Davis et al., 1995), and CG5411 is most closely related to the PDE8 family. Two additional sequences show most similarity to PDE families that exhibit little or no substrate specificity between cAMP and cGMP. These include CG14940 that is related to the PDEI family (calcium/calmodulin activated PDEs), and CG8729, which appears to be a P D E I I family member. Of most interest to the present discussion are three genes that are predicted to code for cGMP-specific PDEs CG10231 that is related to PDE5 and CG1627 and CG3765 that are most similar to PDE9 family members. There are several biochemical studies that support the sequence predictions that insects contain multiple PDEs with different substrate specificities and regulatory properties. Experiments examining PDE activity using fat body from the silkmoth Hvaloplu~ra cecropia (Filburn et al., 1977), accessory gland from crickets (Fallon and Wyatt, 1977) and Mamhu'a CNS (Albin et al., 1975) showed

TABLE 2 Classification of the predicted Drosophila PDEs PDE family

Substrate specificity

Regulatory domains

1 2 3 4 5 6 7 8 9

cAMP/cGMP cAMP/cGMP cAMP/cGMP cAMP cGMP cGMP cAMP cAMP cGMP

Ca:CAM GA F UCRI and 2 GAF GAF

I0 11

cAMP cGMP cAMP cGMP

GAF GAF

PAS

Regulation by cGMP

Drosophihl orthologs CG 14940

Stimulation Inhibition Stimulation

Dunce CG 10231 CG5411 CG 1627 CG3765 CG8729

The substratc specilicies for the manmlalian PDEs arc given and are classified into three classes: cAMP specific, cGMP specific and those ,aith no substrate specificity (Francis et ell., 2001). The regulatory domains include the calcium/cahnodulin-binding domain (Ca/CAM), the cGMP binding domain (GAF), the upstrealn conserved regions (UCRI and 2) and the PAS domain (named after the Per, ARNT and Sire proteins).

24

D. B. MORTON AND M. L. HUDSON

that there were distinct enzymatic activities for cAMP and cGMP-specific PDEs. In addition, Drosophila head extracts contained two distinct PDEs, one that was eliminated in &Tc mutants and the other that was activated by calcium and calmodulin (Sold et al., 1983). The crystal structure of the cAMP-specific PDE4 catalytic site has been solved, and the residues that bind the cAMP substrate have been identified (Xu el al., 2000). An alignment of these residues for mammalian and Drosophila PDEs is shown in Fig. 7A. The catalytic site contains three regions, two metal ion binding sites (ME1 and ME2) and a hydrophobic pocket that binds the cyclic nucleotide. All of the residues that contact cAMP in PDE4 are conserved in d~Tcand CG5411 the Drosophila sequences that are predicted to be cAMP specific. The majority of these residues are also conserved in the Drosophila PDEs, CG10231, CG14940 and CG8279, which are predicted to be either c G M P specific or to have dual specificity. The sequence alignment also reveals that two of the predicted Drosophila PDEs have incomplete catalytic domains. The sequence predicted for CG1627 begins after the two metal ion binding sites and only contains the predicted cNMP-binding site, whereas the predicted sequence for CG3765 only contains ME2. This could mean that these genes do not code for PDEs or that the sequence prediction software has incorrectly predicted the intron/exon structure of the gene, and additional coding sequences need to be identified. Specific residues that confer substrate specificity within the cNMP binding pocket have been identified using site-directed mutagenesis, and these are indicated with an asterisk (*) in Fig. 7A. When the residues W718, Q721, A725 and L727 in PDE5 were mutated to the corresponding residues in PDE4, thc resulting PDE showed selectivity towards cAMP (Turko et al., 1998a). Conversely, when D241 in PDE4 was mutated to an asparagine (present in PDE5), the selectivity was switched to c G M P (Herman et al., 2000). CG10231 shares all five of these residues with PDE5 and thus is likely to be cGMP specific. Interestingly, although PDEI 1 shows equal specificity for cAMP and c G M P (Fawcett et al., 2000) four of the five residues are conserved and the Drosophila PDEI 1 ortholog CG8729, shares all five with PDE5. It will be interesting to determine the specificity of CG8729, as it might provide additional information on cyclic nucleotide selectivity. These residues cannot, however, be the only determinants of selectivity, because PDE9, (which is also c G M P specific), does not share any of the five residues with PDES. One possible reason is that PDE9 has a very low K,,, for c G M P (40 170 times lower than PDE5 or PDE6) (Soderling el al., 1998) and has a very high Vma~ (Fisher et al., 1998). Thus it might be expected that the cGMP-binding site in PDE9 would differ significantly from the other cGMP-specific PDEs. Although the predicted sequence for CG1629 does not include the entire catalytic domain, the predicted cNMP-binding pocket is very similar to that of PDE9, with 28 of 35 residues conserved. If CG 1629 does code for a PDE it might also be a high affinity, cGMP-specific PDE.

CYCLIC GMP REGULATION AND FUNCTION

25

~ ~ ~:¢ t o

,0

~

-68 =

~'~

t°~.

=~'

m m o ~ m o ~ m m ~ m m o ~ m o

r--mp-m

i C~C~O~O~O~C~

C~ O~ C~ O~ O~

,.0

.=

"-

.<

~

.

,

.¢-.0 ~D

tO

N N ~ N e'-, "<21 " - -

~ > > > ~ ~ M M p~ ~

C)~ ~

n~ r~

-r~

~o~

m 0 ~ m o ~ m m m m m m m m

m ~ m ~ m

8 ~

o ~ m ~ m m ~ m

r q r q m ~ m

CI A

&

H

Cl ~

m

~

m

~

r--

m

o~

m

m M M M

cq tn H O~ ~ c-- ~ m Pc~ m c~ ~ ce

m

m ~ m m o ~ ~ c q m m m m m

m ~ m ~ m

>

eeee

H

~ m m H m m

'4 H m o ~1 c /

o m ~ m m

U

H

~

r

e

-,4

o

0

~

>

m ~o r -

H

m

~,~ SS

C~ ~Z

m

a

H

a

H

m ~

m ~0 co en ~n m

~ m Lrt ~0 r--

O4

H m

0 m

,~ ca Z

~ o ~ .~NNNN~

rn

,

¢aj

,~ . ~

,~

26

D. B. MORTON AND M. L. HUDSON

Several of the PDE families contain G A F regulatory domains (named after the cGMP-binding PDEs, the Anabaena adenylyl cyclase and the fhlA gene) that form allosteric binding sites for c G M P in addition to the catalytic site (Soderling and Beavo, 2000). These domains have been studied in most detail in PDE5, where they appear to be involved in a negative feedback loop for regulating cGMP levels. PDE5 contains two c G M P binding sites that are required for its phosphorylation by both cAMP-dependent protein kinases (PKA) and cGMP-dependent protein kinases (PKG) (Turko et al., 1998b), and this phosphorylation increases the catalytic activity of PDE5 (Corbin et al., 2000). As the levels of cGMP increase in a cell, c G M P first binds to the catalytic site, which increases the affinity of the G A F binding sites. This allosteric binding exposes the phosphorylation site of PDE5 to PKG, which is also activated by increased c G M P levels. PDE5 is then activated by PKG phosphorylation, bringing the c G M P back to resting levels (Corbin et al., 2000). PROSITE analysis shows that both CG10231 and CG8279 contain two predicted G A F domains. A number of the residues required for c G M P binding have been identified (Turko et al., 1996) and form a NKX,,D motif that is conserved in both G A F domains of CG10231 and CG8279 (Fig. 7B). In addition, a serine in a similar position to the set92 that is phosphorylated in PDE5 is also present in both the Drosophila sequences. This suggests that both CG10231 and CG8279 in addition to regulating the levels of c G M P may be downstream effectors that bind c G M P and might be stimulated by PKG phosphorylation.

3 3.1

Cyclic GMP function MOLECULARTARGETS

There are three primary protein families that act as c G M P receptors within cells. In addition to the cGMP regulated PDEs (see section 2.2), these include cGMP-dependent protein kinases (PKG) and cGMP-gated ion channels. All three families are characterized by containing cyclic nucleotide binding sites. A search of the Protein Family (Pfam) databases for cyclic nucleotide (cNMP) binding sites in insects revealed 17 separate protein sequences (D. B. Morton, unpublished data). All but one (a voltage and cyclic nucleotide-gated ion channel in Heliothis virescens) were from Drosophila. Five of the sequences were protein kinases (two for the regulatory subunits of PKA and three PKGs), nine were ion channels and three additional proteins were identified. Two of these additional proteins were similar to cAMPdependent guanine nucleotide exchange factors (Kawasaki et al., 1998) and the third was the protein product of the swiss cheese (sws) gene (Kretzschmar et al., 1997). Mutants of the sws gene were isolated in screens for structural brain defects and all five known alleles show age dependent

CYCLIC GMP REGULATION AND FUNCTION

27

neurodegeneration (Kretzschmar et al., 1997). The protein product of sws has three N-terminal cyclic nucleotide binding domains that are most similar to those domains in PKA regulatory subunits (Kretzschmar et al., 1997). However, a multiple sequence alignment of all cNMP-binding sites in insects showed that they were also closely related to the c N M P binding sites of PKGs (D. B. Morton, unpublished observations). The relative affinities of sws for cAMP and c G M P are not known. The importance of the cNMP binding site for normal sws function was demonstrated by the finding that one of the sws mutant alleles contained a point mutation that substituted a glycine that is present in all 17 of the insect cNMP-binding sites. Thus sws probably represents an additional molecular target for cNMPs and might be a target for cGMP. 3.1.1

Proteh7 kinases and substrates

Both mammals and insects have two distinct families of protein kinases that are activated by cyclic nucleotides. The cAMP-dependent protein kinases are tetramers consisting of two catalytic subunits and two cAMP-binding regulatory subunits, whereas the PKGs are homodimers, with each monomer containing both catalytic and cGMP-binding regulatory domains (Siegel el al., 1994). Protein kinase activities that are preferentially regulated by c G M P have been described in a variety of insect tissues, including ttyalophora cecropia body wall (Kuo et al., 1971), fat body from H. cecropia, Antheraeapolyphemus, Mamtuca and the cockroach Blaberus discoMalis (Kuo et al., 1971), B o m b v x pupae (lnoue et al., 1976) and eggs (Takahashi, 1985), Manduca CNS (Morton and Truman, 1986) and whole animal extracts from Drosophila (Kuo et al., 1971) and Ceratitis capitata (Haro et al., 1983). Two genes for P K G have been cloned from Drosophila (Kalderon and Rubin, 1989) and a third gene that appears to code for a PKG has also been identified fi'om the Drosophila genome (Morrison et al., 2000). A BLAST analysis shows that the two cloned PKG genes, Pkg21D (also known as DG1) a n d j b r a g i n g (fbr, also known as DG2), are most similar to mammalian PKG type I and the third, CG4839, is most similar to mammalian P K G type II. All three Drosophila PKGs have a similar domain structure to mammalian PKGs (Fig. 8). The N termini of the proteins are the most variable, but all three appear to have a dimerization domain that contains a leucine/isoleucine zipper motif (Foster et al., 1996). The central region of the predicted protein contains two cyclic nucleotide-binding domains and there is a C-terminal protein kinase catalytic domain. The./br P K G has four different splice variants (Kalderon and Rubin, 1989) that utilize three different initiation codons. The splice variants differ primarily in the N-terminus, although they all contain leucine/isoleucine zippers and hence are predicted to form dimers. Two of the splice variants have a shortened first cyclic nucleotide-binding domain, and hence might only contain one functional cGMP-binding site.

28

D. B. MORTON AND M. L. HUDSON cNMP LZ

1

2

PK

for-P1 for-P2

for-P3 for-p4

[] ........

~

~

Pkg21 D CG4839

[]

~

~

l

n

cGK-I cGK-II

~

l

n

FIG. 8 Schematic representation of the insect PKGs. The domains shown are the leucine/isoleucine zipper (LZ), the two cNMP binding sites and the kinase catalytic domain (PK). The insect PKGs shown are the four splice variants ofjor, Pkg21D and the predicted PKG CG4839. Note that in./br-P3 and/br-P4 exons within the first cNMP-binding domain are absent. Also shown are the two mammalian PKGs, cGKI and cGKII. The biochemical properties of Pkg21D have been investigated and were similar to the properties of mammalian PKGs (Foster et al., 1996). Recombinant Pkg21D formed homodimers, was preferentially activated by c G M P and was autophosphorylated in the presence of c G M P (Foster et al., 1996). Activation of Pkg21D with c G M P showed positive cooperativity, probably reflecting the presence of two cGMP-binding sites (Foster et al., 1996). Although no biochemical investigations have been carried out on t h e / o r gene product, there is experimental evidence, rather than only sequence homology, that suggests it does code for a PKG. The jbr gene was first identified as a naturally occurring polymorphism for food-search behavior (Sokolowski, 1980) and subsequently shown to code for a predicted PKG (Osborne et al., 1997) (see also section 3.2.3). Two naturally occurring alleles o f j o r showed different levels of expression o f J o r mRNA and protein product (Osborne et al., 1997). Correlated with the levels of tbr expression, different levels of cGMPdependent protein kinase activity were measured in animals expressing the different alleles (Osborne et al., 1997), providing circumstantial evidence that Jbr does code for a functional PKG. Relatively few studies have directly investigated specific substrates for PKGs in insects. This is partly because there is considerable overlap between substrate specificities for PKG and PKA. In most tissues, the levels of PKA are much

CYCLIC GMP REGULATION AND FUNCTION

29

higher than PKG and hence it is difficult to identify specific substrates For PKG. For example, in silkworm eggs, vitellin was shown to be phosphorylated by highly purified preparations of both PKA and PKG, and the K,, for both enzymes was similar (Takahashi, 1985). During early development, the levels of PKG was higher than PKA (Takahashi, 1985) and hence vitellin could have been the endogenous substrate, although no evidence was presented that vitellin was phosphorylated in vivo. In M a m t , ca ventral nerve cords (VNCs) two proteins have been described that are phosphorylated specifically in response to c G M P and appear to be phosphorylated by the action of eclosion hormone (EH) in vivo (Morton and Truman, 1986). These proteins, named the EGPs (EH and cGMP regulated Phosphoproteins) both have a mass of 54 kDa but differ in isoelectric point (about 5.6 and 6.6) (Morton and Truman, 1986). In VNC homogenates, cAMP was more effective than c G M P at stimulating their phosphorylation, but in intact nervous tissue the reverse was true, with c G M P being more effective at stimulating their phosphorylation than cAMP (Morton and Truman, 1988a). This difference is probably because both PKA and PKG can phosphorylate the EGPs, and as levels of PKA are about 10 fold higher than PKG in Manduca VNCs (Morton and Truman, 1986), PKA will preferentially phosphorylate the EGPs in homogenates. This is supported by the finding that partially purified preparations of the EGPs can be phosphorylated by mammalian PKA (D. B. Morton, unpublished data). Presumably, in the specific cells that express the EGPs there are either comparable levels of PKA and PKG, or there is more PKG than PKA, and hence in the intact tissue c G M P is more effective. The EGPs were also phosphorylated by exposure of intact VNCs to EH (Morton and Truman, 1988a). EH acts on the VNC to stimulate an increase in cGMP but has no effect on the levels of cAMP (Morton and Truman, 1985), providing further evidence that the EGPs are specific substrates for PKG. The identity. function and cellular location of the EGPs remain to be determined. Studies on mammalian cGMP-stimulated protein phosphorylation have identified several specific substrates for PKGs (see Smolenski er al., 1998). Several of these appear to be mammalian specific as BLAST analysis of the Drosophihl genome fails to identify orthologs (D. B. Morton, unpublished observations). These include the protein phosphatase inhibitors G-substrate and DARPP-32 and the inositol trisphosphate receptor (IP~R) associated PKG substrate (IRAG). Orthologs of other protcins, shown in mammalian systems to be substrates for PKG, are present in Drosol)hihl and hence are possible substrates for one or more of the Drosophila PKGs. These include the IP3R, L-type calcium channels, calcium sensitive potassium channel, phospholipase C and the cystic fibrosis transmembrane conductance regulator. Another family of proteins that includes substrates for PKG in mammals is of particular interest because of the relationship between c G M P and axonal outgrowth (see section 3.2.2). Many studies have shown that c G M P is a major regulator of vascular smooth muscle relaxation (see Lincoln el a/., 1994) and

30

D.B. MORTON AND M. L. HUDSON

the vasodilator-stimulated phosphoprotein (VASP) is a substrate for PKG in smooth muscle, vascular endothelium and platelets (Smolenski et al., 1998). VASP is a member of a family of related proteins that includes the product of the Drosophila enabh, d (enb) gene, the mammalian ortholog of enb, mena, and an additional related protein, Ena/VASP-like protein (EVL) (Gertler et al., 1996). The Drosophila enb gene was first identified as a suppressor of the tyrosine kinase, Abels, on (Gertler et al., 1995). Mutations in the enb gene include defects in neuronal dendritic and axonal outgrowth and branching (Gao et al., 1999; Wills el al., 1999) and interacts with robo, a Drosophila axon guidance receptor (Bashaw et al., 2000). BLAST analysis of the Drosophila genome has identified another member of this family, CG10155 (D. B. Morton, unpublished observations). There is accumulating evidence that members of the Ena/VASP family can act as adapter proteins linking extracellular signaling pathways to actin polymerization (Prehoda el al., 1999) and that phosphorylation of VASP regulates its interactions with actin (Harbeck et al., 2000). The functional conservation in the Ena/VASP family has been demonstrated by the rescue of lethal enb mutations with human VASP (Ahern-Djamali et al., 1998), although it is not known whether VASP will rescue dendritic and axon guidance defects. The cGMP-dependent phosphorylation sites of VASP are conserved in mena and EVL (Gertler et al., 1996). Although there is no PKG phosphorylation site (two basic residues 2 and -3 of a serine or threonine) at the equivalent position in the ent~ and CG10155 proteins, there is at least one potential PKG site elsewhere in both sequences. There is accumulating evidence that cGMP plays a critical regulatory role in neuronal outgrowth and pathfinding (section 3.2.2) and hence it is intriguing that the protein product of enb (a gene that affects dendritic and axonal outgrowth and branching) and CG 10155 could be substrates for PKG. 3.1.2 ()'clio nucleotMe-gated channels The Drosophila genome contains predicted genes for three classes of ion channels that have intracellular cNMP-binding domains (Littleton and Ganetski, 2000). These classes are the cNMP-gated ion channels that are orthologs of the mammalian retinal cGMP-gated ion channels (Zagotta and Siegelbaum, 1996), the Ih channels that are activated by hyperpolarization and contain a cNMP-binding site (Ludwig et al., 1998) and the EAG class of voltage activated potassium channels that also contain a cNMP-binding site (Briiggeman et al., 1993). There are four genes in the Drosophilci genome that appear to code for cyclic nucleotide-gated ion channels (CNGs) (Littleton and Ganetski, 2000). Two have been cloned and partially characterized, cyclic nuch, otide-gated ion channel protein (cng) (Baumann el al., 1994) and cng-like (cngl) (Miyazu et al., 2000), while two additional genes, CG3536 and CG17922, have been identified from sequencing the Drosophila genome. The mammalian CNG

CYCLIC GMP REGULATION AND FUNCTION

31

channels are composed of two homologous (oe and /3) subunits that form tetramers in vivo. although the stoichiometry is not known (Zagotta and Siegelbaum, 1996). Mammalian c~ subunits form functional homomeric channels in heterologous expression systems and although the /3 subunits do not form functional channels on their own, heteromeric channels formed from o~ and /3 subunits have distinct kinetic properties compared to homomeric oe channels (Finn el al., 1998). The cng channel has been expressed in heterologous cells and functions as a homomeric cGMP-gated non-specific cation channel that is 50-fold more sensitive to c G M P than to cAMP (Baumann el al., 1994). The cngl channel, by contrast, does not l~rm a functional homomeric channel in heterologous cells, suggesting that it is more similar to a /3 subunit (Miyazu et al., 2000). Varnum et al. (1995) have identified an aspartate residue near the C-terminal end of the cNMP-binding site that determines the selectivity towards c G M P in the bovine rod CNG. A C L U S T A L W sequence alignment of the predicted cNMP binding sites in the four Drosophila CNGs showed that the c~Tgchannel has an aspartate at the equivalent position, but cngl, CG3536 and CG17922 have valine, tyrosine and asparagine residues in place of the aspartate (D. B. Morton unpublished data). Varnum el al. (1995) found that replacing the negatively charged aspartate with an uncharged polar residue made the channel non-selective towards c G M P and cAMP, whereas replacement with a non-polar residue made the channel selective towards cAMP. Thus, CG3536 and CG17922 would be predicted to have little selectivity towards c G M P or cAMP, whereas cngl would be expected to form a channel specific to cAMP. Finn et al. (1998) report cloning a Drosophila CNG /3 subunit that does not form active homomeric channels. This subunit is likely to be the same as the CG3536 channel, as their sequences are 97.5% identical (D. Krautwurst. personal communication). Co-expression of the /3 subunit with the rat oe subunit of the olfactory CNG (rOCNGc0 yielded a channel that had a larger cGMP-induced current than cAMP-induced currenL whereas homomeric rOCNGoe channels yielded similar cGMP- and cAMPinduced currents (Finn eta/., 1998). This suggests that CG3536 might confer c G M P selectivity to the subunit with which it forms channels in fifo. The cellular expression patterns of cn~ and oHg/ suggest that these two subunits do not form heterodimers in s,ivo. The cHg channel is expressed in the antennae and eyes. although a snore detailed analysis of the expression pattern in specific cells is not known (Baulnann et al., 1994). cn~,~lis expressed in antennal lobes, mushroom bodies, neurons in the thoracic ganglia and in muscle fibers (Miyazu el a/., 2000). In addition to the C N G channels, there are two other families of ion channels that contain cNMP-binding sites. These are the hyperpolarizationactivated (lh) channels (Ludwig el al., 1998) and the ether-(t-,4o-~o (eag) family of voltage activated potassium channels. Two lh channels have been reported in insects, one in Helioflfis ~,irescens (Krieger et al., 1999) and the other in Drosophiht (Marx er al., 1999). The mammalian lh channels are activated b~

32

D. B. MORTON AND M. L. HUDSON

hyperpolarizing currents and the presence of either cAMP or c G M P shifts the voltage dependence, making the channel open at more positive voltages (Ludwig e t a ] . , 1998). The Helioth& lh channel has been expressed in Spodoplera cells and was activated by both cAMP and hyperpolarizing current pulses (Krieger el al., 1999). Both insect channels show prominent expression in sensory tissues such as the eye, antennae and auditory organs in Drosophila (Marx el al., 1999) and in antennal olfactory sensory neurons in Heliothis (Krieger el al., 1999). The Drosophila eag family of ion channels contains three members eag,, eag-like K" channel (elk) and sei:ure (sei) (Littleton and Ganetsky, 2000). When expressed in heterologous cells, depolarizing voltages activated the eag channel and cAMP shifted the threshold for activation to more negative voltages (Brfiggeman eta]., 1993). The application of c G M P appeared to have no direct effect, although it is possible that P K G might modulate eag (Brfiggeman el a]., 1993). There are no reported studies on the effects of cAMP or c G M P on the elk or sei channels.

3.2

PHYSIOLOGICAL FUNCTIONS

The two systems where c G M P regulation and function have been most extensively studied are probably mammalian photoreceptors (Stryer, 1986) and vascular smooth muscle (Lincoln, 1989). In vertebrate photoreceptors, c G M P is the primary transduction signal, c G M P binds to and opens cGMPgated channels maintaining the dark current. Subsequently, light-triggered c G M P hydrolysis causes channel closing and hyperpolarization (Stryer, 1986). Both NO-sensitive soluble GCs and receptor GCs regulate the levels of c G M P in vascular smooth muscle. NO is released from the neighboring endothelial cells that line the blood vessels and activates soluble GC (Bredt and Snyder, 1994), while the atrial natriuretic peptides are released from the heart and regulate receptor GCs (Drewett and Garbers, 1994). The c G M P thus formed activates cGMP-dependent protein kinases, resulting in the phosphorylation of several proteins and smooth muscle relaxation (Lincoln et al., 1994). Both systems act in concert to regulate smooth muscle tone and blood pressure. Neither of these systems has a direct parallel in insects. Insect phototransduction does not involve cGMP, but rather utilizes a light-stimulated phospholipase C and activation of TRP (transient receptor potential) and TRP-like channels (Montell, 1999). The open circulatory system of insects means that there is no parallel to vascular smooth muscle, and there are no reports that c G M P plays a role in modulating the tension of insect skeletal or visceral muscles, although glutamate appears to increase c G M P levels in insect skeletal muscle (Robinson et al., 1982). Nevertheless, there are several examples of physiological processes where there is substantial evidence that c G M P plays a major regulatory role.

CYCLIC GMP REGULATION AND FUNCTION

3.2.1

33

Sensory physiology

The genetic analysis of phototransduction in Drosophila has described the signal transduction pathways in great detail and shows that lnsP3 formation is the primary signal (Montell, 1999). There is, however, evidence showing cGMP involvement. There are several examples where photoreceptors express genes, which are known to participate in the formation or action of cGMP, that implies a role for cGMP. NO-sensitive soluble GC activity has been detected in locust eyes and immunoreactivity for the ce subunit has been localized to locust photoreceptors, where it has been suggested to play a role in dark adaptation (Jones and Elphick, 1999). Electroretinogram (ERG) recordings from locust eyes showed that cGMP and NO donors increased the response of the photoreceptors to light (Schmactenberg and Bicker, 1999). NO-sensitive GC activity has also been detected in Drosophila photoreceptors during adult development (Gibbs and Truman, 1998), where it has been suggested to play a role in synapse formation (Gibbs and Truman, 1998). Support for this idea has been provided by an analysis of DgceH mutations, which failed to show phototaxis (Gibbs el el/., 2001). ERGs showed that this was not due to a failure of the primary phototransduction cascade but to defects in the post-synaptic responses to light stimuli (Gibbs el al., 2001). The Drosophila cng and ]h channels are also expressed in photoreceptors (Baumann el ell., 1994; Marx el al., 1999). Application of exogenous cGMP induced membrane currents and also enhanced light-induced currents in Drosophila photoreceptors (Bacigalupo et ell., 1995). Similarly, although the primary transduction pathway in insect olfactory neurons does not seem to be mediated by cGMP, the presence of various elements of the cGMP pathway suggests that it is involved in some aspect of olfactory sensation. In most vertebrate olfactory neurons, the primary transduction signal is cAMP, which acts on a CNG channel (Brunet et ell., 1996). A large family of G protein-coupled receptors has been described that are believed to act as the olfactory receptors that couple to adenylyl cyclase (Buck and Axel, 1991). Drosophila also has a large family of G protein-coupled receptors that are expressed in olfactory and gustatory neurons (Clyne el al., 1999; Vosshall et el/., 1999). Pheromone stimulation of silkmoth antennae triggers a rapid and transient increase in lnsP3 (Breer el ell., 1990) and mutations in the Drosophila noq)A gene that encodes a PLC significantly reduced olfactory signaling (Riesgo-Escovar et a/., 1995). These studies suggest that the primary signal for most olfactory stimuli is InsP3. Both soluble and particulate GC enzyme activity have been measured in silkmoth antennae (Ziegelberger et al., 1990), and the Mamtuca receptor-like GC, MsGC-I has been detected in olfactory receptor neurons (Nighorn et al., 2001). In addition, the Drosophila cng, lh and Heliothis lh channels are expressed in antennae (Baumann et ell., 1994; Marx et ell., 1999; Krieger et al., 1999). Pheromone stimulation of silkmoth antennae also triggered an

34

D.B. MORTON AND M. L. HUDSON

increase in c G M P levels (Ziegelberger et al., 1990), but the time course of increase was much slower than the increase in InsP3, and was suggested to play a role in adaptation. Interestingly, the GC enzyme activity and the pheromone-stimulated c G M P increases measured in silkmoth antennae were detected in whole antennae but not in the olfactory dendrites, providing additional evidence against a primary signal transduction role for c G M P in olfaction (Ziegelberger et a/., 1990). Further evidence for a modulatory role of c G M P in olfactory processing comes from studies on C. ele~ans. A large family of receptor GCs has been described in C. e/egans, many of which are expressed in olfactory neurons (Yu et al., 1997). Two of these, ODR-I and DAF-I 1, are expressed in the chemosensory neuron, AWC, and mutations to either gene show that both are necessary for chemotaxis to all AWC-sensed odorants (Bargmann et al, 1993; Birnby et al., 2000). Deleting the extracellular portion of ODR-I showed that it was not required and hence did not function as the olfactory receptor itself (L'Etoile and Bargmann, 2000). Additional experiments showed that ODR-1 was necessary for odor adaptation and discrimination (L'Etoile and Bargmann, 2000), again suggesting a modulatory role of cGMP in olfactory signaling. In addition to these potential roles in visual and olfactory sensory neurons, c G M P signaling has also been implicated in sensory information processing in primary sensory interneurons and higher order centers (Mfiller, 1996; Nighorn et al., 1998, 2001). Most of the evidence for a role in these processes is the localization of GC and NOS expression to the antennal and optic lobes and central brain structures, although some functional studies have been carried out (see Mfiller, 1996: Bicker, 2001).

3.2.2

Neuronal development

There is accumulating evidence for a role of c G M P in neuronal development, particularly in axonal pathfinding and synapse formation. A wide variety of studies have identified many cell surface and secreted proteins that regulate axonal pathfinding and synapse formation in a wide variety of insects (TessierLavigne and Goodman, 1996). Although several studies have described downstream effectors for some of these proteins (Mueller, 1999L for the most part the signal transduction events that mediate axonal pathfinding are still relatively poorly understood. Nevertheless, evidence from a variety of different experimental systems have shown that c G M P can play an important role in these developmental events. Several studies have shown that NOS, NOsensitive soluble GCs and/or NO-stimulated c G M P immunoreactivity are localized to neurons at times in development when they are growing or nearing their targets. These studies include locust motor neurons (Truman et al., 1996; Ball and Truman, 1998) and antennal pioneer neurons (Seidel and Bicker, 2000), Drosophila photoreceptors (Gibbs and Truman, 1998) and Manduca

CYCLIC GMP REGULATION AND FUNCTION

35

enteric (Wright et al., 1998) and olfactory neurons (Gibson and Nighorn, 2000). There have also been studies where investigators have blocked NOS and/or NO-sensitive soluble GC and have shown that these treatments have disrupted outgrowth and patterning. During development of the enteric nervous system of M a n d u c a , neurons migrate along the midgut and then extend processes on to the gut musculature (Copenhaver and Taghert, 1989, 1990). Prior to migration none of the enteric neurons showed NO-stimulated GC activity, but soon after the onset of migration a subset (about 50%) of the neurons showed strong NOstimulated c G M P immunoreactivity (Wright e t a / . , 1998). These cells continued to show a NO-stimulated increase in c G M P as they ended migration, extended axons and elaborated processes on to the gut musculature (Wright et al., 1998). Blocking either NOS or NO-stimulated soluble GC activity with a variety of reagents had no effect on the extent of migration or axonal outgrowth (Wright et al., 1998). However, the extent of the terminal branches formed during the final stages of differentiation was significantly reduced when embryos were treated with NOS and soluble GC inhibitors. In addition, the level of immunoreactivity to the synaptic vesicle protein, synaptotagmin, was also significantly decreased by blocking NOS and soluble GC (Wright et al., 1998). These results suggest that the migration and axon formation of enteric neurons does not require NO-stimulated cGMP, but some aspect of terminal differentiation or synapse formation requires increased c G M P levels (Wright et aL, 1998). A similar situation is seen with the development of photoreceptors in Drosophila. During metamorphosis, the photoreceptor axons grow into the CNS to form a precisely ordered projection pattern (Wolff and Ready, 1993). When the tissue was incubated with NO donors during this time, the photoreceptors' axons showed an increase in cGMP, whereas at earlier and later times in development the sensitivity to NO-donors was not apparent. These results suggested that c G M P was involved in the formation of these patterns (Gibbs and Truman, 1998). These predictions were confirmed by maintaining the developing eyes and CNS in tissue culture. When the tissue was incubated with NOS and soluble GC inhibitors the axonal projection patterns were disrupted, but could be rescued with the addition of exogenous c G M P (Gibbs and Truman, 1998). The axons tended to overgrow their normal termination zones, suggesting that NO-stimulated c G M P was necessary to signal the axons to stop growing and to differentiate and form synapses (Gibbs and Truman, 1998). Surprisingly, when flies with a disrupted gene for the c~ subunit of NO-stimulated soluble GC were examined, no defects were seen in the gross organization of the axonal projection patterns (Gibbs et al., 2001). Closer examination of these flies, however, showed defects in their phototaxis behavior and electroretinograms revealed deficits in photoreceptor synaptic transmission (Gibbs et a/., 2001). Thus, in a similar manner to the development of M a n d u c a enteric neurons, NO-stimulated c G M P is not

36

D. B. MORTON AND M. L. HUDSON

necessary for axonal outgrowth, but is required for termination of growth and synapse development. There is also evidence that the activity of the Drosophila PKG./br mediates neuronal outgrowth at the neuromuscular junction (NMJ). As described previously (section 3.1.1) there are two naturally occurring alleles of the jbr gene that express different levels of P K G (Osborne et al., 1997). When the NMJ of flies bearing the different alleles were examined they showed significantly different branching patterns and projections, suggesting that PKG played a role in their formation (Renger et al., 1999). Here, as with the photoreceptor axons, the alleles with the lower levels of P K G activity showed additional branching of motor neuron axons (Renger et al., 1999). It is possible that again, the reduced levels of a c G M P signaling system prevent the normal terminal differentiation and formation of synapses. These effects might not be direct, however, as different levels of spontaneous electrical activity and evoked transmitter release were also observed and activity patterns are well known to influence neuronal branching and outgrowth (e.g. Budnik et al., 1990). A contrasting situation exists in the developing antennae of locusts. In this tissue, axonal outgrowth rather than the terminal differentiation of the developing olfactory receptor neurons is inhibited by the addition of NOS and soluble GC blockers (Seidel and Bicker, 2000). Thus, there appears to be more than one role for c G M P in developing neurons of insects. Future studies are likely to reveal additional roles and it is especially intriguing to note that the presence of c G M P in the nuclei of developing neurons has been described, suggesting a role for c G M P in the regulation ofgene expression (Truman el al., 1996; Seidel and Bicker, 2000). Although these studies strongly support a role for c G M P in neuronal development, there is no information on the nature of the signals that activate the N O / c G M P pathway. A possible candidate for this has, however, emerged from studies in vertebrate neurons. The secreted protein, semaphorin III, induces the directed growth of growth cones in cultured Xenopus spinal neurons. Simultaneous application of c G M P analogs with semaphorin 1II will reverse the direction of this growth (Song et al., 1998). Semaphorin III also acts as a chemoattractant for dendrites of pyramidal neurons of mice, yet acts as a repulsive agent for the axons of the same neurons. Interestingly, there is a differential distribution of soluble GC in these cells, with a higher concentration in dendrites compared to axons (Polleux el al., 2000). This suggested that a differential distribution of c G M P led to the different responses - a model supported by the finding that blocking soluble GC activity prevented dendritic outgrowth but had no effect on axonal outgrowth (Polleux et al., 2000). Semaphorin family members have been identified in grasshoppers and Drosophila, where they are also essential in axonal pathfinding and the formation of correct synaptic junctions between motor neurons and muscles

CYCLIC GMP REGULATION AND FUNCTION

37

(Kolodkin et al., 1997: Winberg el al., 1998). Thus the possibility exists that in insects the effects of semaphorin are also mediated by cGMP. 3.2.3

Ecd~,sis"

Probably one of the best studied roles for c G M P in insects is the action of eclosion hormone and the regulation of ecdysis behavior. At the end of each molt, insects need to escape from the cuticle of the previous instar. This is accomplished by a stereotyped behavioral sequence known as ecdysis, or eclosion for adult ecdysis (Reynolds, 1980). Initiation of ecdysis and the preparatory behavior, pre-ecdysis, involves a positive feedback loop between two peptides, eclosion hormone (EH) and ecdysis-triggering hormone (ETH) (Zitnan el al., 1996, 1999: Ewer el al., 1997: Gammie and Truman, 1997a, 1999: Zitnan and Adams, 2000). EH is a 62 amino acid peptide that is located in two pairs of neurons in the brain (Truman and Copenhaver~ 1989) and is released both centrally into the neuropil of the ventral nerve cord (VNC) and peripherally into the circulatory system (Hewes and Truman, 1991). Early studies on the role of EH in triggering ecdysis behavior showed that the action of EH was mediated by an increase in the levels o f c G M P in the VNC (Truman et al., 1979, Morton and Truman, 1985). Subsequent immunocytochelnical studies identified three specific cell populations that increase their c G M P levels in response to EH. Circulating EH acts on peripherally located epitracheal glands (EG). The epitracheal glands are located near each of the 18 spiracles and each contains four cells (Zitnan el al., 1996, 1999: Klein el al., 1999). The glands respond to the action of EH with an increase in c G M P (Ewer et al., 1997: Kingan et el/., 1997) and one of the four cells, the Inka cell, releases its content of ETH (Zitnan el ell., 1996: Kingan el al., 1997). ETH then forms a positive feedback loop by acting on the EH-containing cells in the brain to trigger further release of EH (Ewer et al., 1997). Centrally released EH acts on a population of about 50 neurons in the ventral ganglia that release another neuropeptide, crustacean cardioactive peptide (CCAP) (Ewer el al., 1994: Gammie and Truman, 1997a). The CCAP cells also show an increase in c G M P and the EH/cGMP-triggered release of CCAP is believed to directly activate the ecdysis motor program (Gammie and Truman, 1997a, 1999). A third EH target has been identified that does not appear to be directly involved in ecdysis behavior. Isolated VNCs from prepupal M a m l u c a responded to EH with a large increase in c G M P (Morton and Giunta, 1992) that was restricted to the transverse nerves of each abdominal ganglion (Morton, 1996). This EH-stimulated c G M P increase was localized to a population of intrinsic cells located in the posterior region of the transverse nerve, named the sub-transverse nerve region (STNR) (Hesterlee and Morton, 2000). The STNR cells are believed to develop into the ventral diaphragm muscles that ultimately lie dorsal to the VNC in the adult (Champlin el a/., 1999: Hesterlee and Morton, 2000).

38

D. B. MORTON AND M. L. HUDSON

A series of studies showed that the EH-stimulated c G M P increase occurs via a novel mechanism that does not utilize an NO-sensitive soluble GC, or a receptor GC, but instead appears to activate an NO-insensitive soluble GC (reviewed in Morton and Simpson, 2001). The primary evidence for this was that EH stimulated a c G M P increase in intact tissue, but failed to do so in tissue homogenates (Morton and Giunta, 1992). This property is a characteristic of pathways that utilize a soluble GC, as ligand-stimulated receptor GC activity can often be demonstrated in cell free preparations. The EH-stimulated c G M P increase in VNCs was not blocked by a variety of NOS inhibitors and NO donors did not mimic the action of EH in any of the EH target cells described above (Ewer et al., 1994; Morton, 1996; Kingan et al., 1997). This demonstrated that the GC present in the target cells was not an NO-sensitive soluble GC. These findings triggered the search for novel GCs that led to the identification of MsGC-I and MsGC-fl3, both potential candidates for the EHstimulated GC as they are NO-insensitive soluble GCs (see section 2.1.3). MsGC-I is not expressed in any of the target cells, ruling it out as a candidate (Simpson et al., 1999). Recent data have shown, however, that MsGC-/33 is expressed in both the STNR and the Inka cells providing circumstantial evidence that MsGC-fl3 is activated by EH in these cells (Morton, 2000; Morton and Simpson, 2001). A major question that remains to be resolved is the mechanism that is used to activate MsGC-fl3. Studies using M a n d u c a VNCs showed that EH stimulated an increase in lnsP3 levels (Morton and Simpson, 1995), and a variety of inhibitors of lipid metabolism blocked the EH-stimulated increase in c G M P (Morton and Giunta, 1992; Morton and Simpson, 1995). These studies suggested that production of a lipid intermediate was necessary for c G M P production, and the most likely pathway appears to involve activation of protein kinase C that then leads to the activation of MsGC-fl3 (see Morton and Simpson, 2001). As MsGC-fl3 is not expressed in the CCAP ceils, an alternative pathway is likely to be used in these ceils. Whether this involves another NO-insensitive soluble GC or whether they express an EH-stimulated receptor GC is not known. The physiological events downstream of the EH-stimulated increase in c G M P are cell specific. In the Inka cell of the epitracheal glands, the EHstimulated c G M P increase mediates the release of ETH (Kingan et a/., 1997). The release of ETH also involves increases in cytoplasmic calcium levels. Studies on isolated epitracheal glands showed that release of intracellular stores of calcium in the lnka cells using thapsigargin stimulated ETH release and EH triggered an increase in intracellular calcium levels (Kingan et al., 2001). Incubation of epitracheal glands with 5raM cobalt in the absence of extracellular calcium had no effect on the EH-stimulated release of ETH~ suggesting that release did not require extracellular calcium (Kingan et al., 1997). It has been suggested that EH triggers a parallel increase in both c G M P and calcium in the Inka cells, which act in concert to trigger ETH release (Kingan et al., 2001). Protein phosphorylation acts downstream of

CYCLIC GMP REGULATION AND FUNCTION

39

the c G M P accumulation to mediate EH-stimulated ETH release. Staurosporine, a broad spectrum protein kinase inhibitor, blocked ETH release but had no effect on EH-stimulated c G M P levels (Kingan et al., 2001). In addition, the protein phosphatase inhibitor calyculin A, potentiated both basal and EHstimulated ETH release (Kingan et al., 2001). These results suggest that activation of a cGMP-dependent protein kinase (PKG) is a likely mediator of ETH release (Kingan et al., 2001). EH also appears to act on the CCAP cells as a peptide-releasing hormone, stimulating the release of CCAP, which then triggers the ecdysis motor program (Gammie and Truman, 1999). Although there is considerable circumstantial evidence that EH acts directly on the CCAP cells, stimulating CCAP release via an increase in c G M P levels, there is, as yet, little direct proof for this pathway. Electrical stimulation of the VM cells (to stimulate EH release) and application of EH to desheathed abdominal ganglia cause c G M P increases in the CCAP cells and initiation of the ecdysis motor program (Gammie and Truman, 1999; Zitnan and Adams, 2000). Application of CCAP to desheathed ganglia also triggers the ecdysis motor program (Gammie and Truman, 1997a). Ecdysis triggered with the application of CCAP is initiated with a shorter latency than with EH and there is no corresponding increase in c G M P indicating that EH action is upstream of both the c G M P increase and CCAP release (Gammie and Truman, 1999). However, the ability of EH to induce the ecdysis motor program varies between individual preparations, and this variability has been interpreted to indicate that the effects of EH on the CCAP cells are not direct (Zitnan and Adams, 2000). Another explanation is that multiple inputs on to the CCAP cells are necessary for CCAP release, one of which is EH. Evidence to support this comes from Drosophila, where the absence of EH does not prevent ecdysis in all animals, although it does result in a less robust expression of the behavior (McNabb et al., 1997). This is consistent with the presence of multiple pathways activating the CCAP cells. As EH increases c G M P by direct action on the epitracheal glands and the STNR cells, it seems likely that it also acts directly on the CCAP cells to increase c G M P levels. The effect of increased c G M P levels in one of the CCAP cells, cell 27, has also been investigated (Gammie and Truman, 1997b). At times in development prior to the endogenous increase in cGMP associated with ecdysis, cell 27 showed a higher threshold for firing action potentials compared with times when c G M P was elevated. In addition, there were no spontaneous action potentials in the absence of an increase in cGMP, whereas, after c G M P levels had risen, spontaneous action potentials at frequencies of up to 2 Hz were recorded (Gammie and Truman, 1997b). Application of the cell-permeant analog, 8-bromo-cGMP mimicked the decrease in action potential threshold (Gammie and Truman, 1997b). Pharmacological and ion-substitution experiments suggested that this cGMP-induced reduction in threshold was mediated by enhancing an inward calcium current (Gammie and Truman, 1997b). These

40

D. B. MORTON AND M. L. HUDSON

electrophysiological changes could then lead to CCAP release, either directly or by enhancing the efficacy of additional synaptic inputs. The STNR cells of the abdominal transverse nerves also respond to EH with an increase in cGMP. The physiological significance of this increase is, however, unclear. The STNR cells are present throughout larval life and go through a period of enhanced proliferation during pupal development (Champlin et al., 1999: Hesterlee and Morton, 2000). The cells also become sensitive to EH during pupal development. They show no increase in c G M P in response to EH at larval ecdysis or during the final larval instar, but first show an EH-stimulated c G M P increase soon after the wandering stage at the onset of pupal development (Hesterlee, 1999). Although it has not been demonstrated directly, the STNR cells are likely targets for circulating EH at pupal ecdysis (Morton, 1996, 1997). Soon after ecdysis, the cells spread out from their tightly grouped position along the transverse nerve to become a thin sheet of cells covering each abdominal ganglion (Hesterlee and Morton, 2000) and subsequently develop into the ventral diaphragm muscles in the adult (Champlin el al., 1999; Hesterlee and Morton, 2000). Several possible roles for the EH-stimulated c G M P increase have been suggested, such as triggering the migration or initiating the differentiation of the cells, but as yet there is no evidence to support these possible actions (Hesterlee and Morton, 2000). Although the physiological significance of the EH-stimulated c G M P increase in the STNR cells is unknown, there is circumstantial evidence that the EGPs (described in section 3.1.1 ), proteins that are phosphorylated in response to EH and c G M P are located in the STNR cells. The most compelling evidence for this comes from experiments with isolated VNCs exposed to EH. Immunocytochemistry showed that in these preparations EH stimulated a c G M P increase only in the STNR cells (Morton, 1996). No increase was seen in the CCAP (crustacean cardioactive peptide) cells, as circulating or bath-applied EH cannot cross the blood brain barrier (Gammie and Truman, 1999). Nevertheless, incubation of isolated VNCs with EH stimulated the phosphorylation of the EGPs in a cGMP-dependent manner (Morton and Truman, 1988a), suggesting that the EGPs are localized to the STNR cells. Thus, the E H / c G M P pathway plays a central role in the physiology of ecdysis. In the Inka and CCAP cells it leads to the release of additional peptides and activation of the ecdysis motor program. It is likely that EH also functions as a master coordinating hormone regulating additional physiological events associated with ecdysis. The programmed degeneration of the intersegmental muscles of the silk moth Antheraea polyphemus lbllowing adult ecdysis is thought to be triggered by the direct action of EH via an increase in c G M P (Schwartz and Truman, 1982, 1984). The Verson's glands in Manduca release their contents at ecdysis a process that is prevented when peripherally released EH is blocked (Hewes and Truman, 1991) although it is not known whether this is the result of the direct action of EH or whether it is mediated by cGMP. Studies of eclosion in Drosophila revealed that an increase in c G M P

CYCLIC GMP REGULATION AND FUNCTION

41

was occasionally seen in tracheae that correlated with the release of EH (Baker et al., 1999). These c G M P increases were absent in animals that lacked EH, suggesting that tracheae are also EH targets (Baker et al., 1999). Many of the animals that lacked EH were unable to correctly fill their tracheae with air following ecdysis, suggesting a role for the E H / c G M P pathway in this process (Baker el al., 1999). The action of EH on the STNR cells is also likely to be part of this repertoire of physiological events that are associated with ecdysis. 3.2.4

k~)od-seareh behavior

A genetic analysis of food searching strategies in Drosophila has shown that c G M P plays a central role in this behavior (Sokolowski, 1998). Two naturally occurring alleles of the Jbraging (/or) gene exhibit different behaviors. In the presence of food, flies with the sitter allele stayed relatively stationary and remained in a single patch of food. By contrast, flies with the rover allele continued to forage for additional food and moved between patches of food (Sokolowski, 1980). In the absence of food, however, both alleles showed similar levels of locomotory activity, showing that sitter flies were not simply more sluggish and that the behavioral differences were foraging specific. Several lines of evidence showed that the jbr gene codes for a P K G (Osborne et al., 1997). Both the naturally occurring alleles and additional mutations in t h e j b r gene mapped close to a region that contains a previously identified gene for PKG (Osborne et al., 1997). The levels of total P K G enzyme activity were measured in these strains and showed that in flies with the naturally occurring sitter alleles, and in jbr mutations that exhibited a sitter phenotype, there was a slightly reduced, but significantly lower level of P K G present (Osborne et al., 1997). Flies containing a P element that was located in, and disrupted the open reading frame of, one of the jbr PKG transcripts, showed a sitter phenotype, and excision of the P element caused flies to revert back to the rover phenotype (Osborne et al.. 1997). Further proof that ./br encodes a P K G came fi'om overexpression of a cDNA for the PKG in sitter flies. This manipulation caused an increase in PKG levels and a rover phenotype (Osborne et al., 1997). Northern and western blots showed that both RNA and protein levels for t h e / b r P K G were lower in sitters than rovers (Osborne et al., 1997). A preliminary description of the expression patterns o f l o t showed widespread distribution in olfactory, gustatory, gut and brain tissues (Sokolowski, 1998). This leaves open several possible levels of organization that could be modulated by the c G M P / P K G pathway resulting in the behavioral differences. Expression o f . / o r in chemosensory tissues suggests that different levels of chemosensation associated with food might generate the different foraging strategies. The finding that the behavioral differences are only apparent in the absence of food supported this idea and, in addition, many different studies have provided evidence for a role of c G M P in modulating olfactory pathways

42

D.B. MORTON AND M. L. HUDSON

(see section 3.2.1). Behavioral differences associated with the presence or absence of food could also be activated by signals originating in the gut, and the presence of Jbr in gut tissue suggests that this is also a possibility. Expression of Jbr in the central regions of the brain, and the finding that differences of PKG activity could be detected in the heads of sitter and rover strains, suggest that differences in the central processing of information are also likely to represent a large component of the behavioral phenotype. A series of physiological studies are consistent with this suggestion. Cultured giant neurons from strains that showed the sitter phenotype (both natural alleles and induced mutations) showed hyperexcitability and reduced voltagedependent potassium currents compared with animals with rover phenotypes (Renger et al., 1999). Similarly, recordings from the neuromuscular junction showed that sitter strains exhibited higher levels of spontaneous activity and larger evoked post-synaptic potentials as compared with rovers (Renger et al., 1999). Thus it is likely that cGMP/PKG-mediated modulation of the nervous system is necessary at several levels of organization to generate the highly specific phenotypical differences in the two strains of flies. Clearly, many more studies are needed to gain a clearer understanding of this complex behavior and the role that cGMP signaling plays in its production. 3.2.5 Malpighian tubule regulation The regulation of fluid secretion from Malpighian tubules is another system for which there is good evidence for a central role of cGMP. Malpighian tubules are the principal excretory organ in insects, responsible for the production of primary urine (Chapman, 1998). In most insects, potassium is actively pumped into the tubules, and other solutes and water then follow passively. Resorption of ions, amino acids, water, sugars and other necessary components is then accomplished in the Malpighian tubules themselves or in the ileum and rectum (Chapman, 1998). The rates of both excretion and resorption are regulated by a variety of circulating hormones (Chapman, 1998). The majority of the studies that implicate cGMP as a regulator of Malpighian tubule fluid secretion have been carried out in Drosophila. RT PCR experiments have shown that Drosophila Malpighian tubules express the PKGs DG1 (Pkg21D) and DG2 (/br) (Dowet al., 1994), NOS (Davies et al., 1997) and the cGMP-gated ion channel cn~, (MacPherson et al., 2001). Experiments on isolated tubules show that both NO donors and exogenous cGMP stimulate fluid secretion and zaprinast, an inhibitor of cGMP-specific PDEs, and okadaic acid, a protein phosphatase inhibitor, potentiates the NOstimulated and cGMP-stimulated fluid secretion (Dowet al., 1994). At least two neuropeptides also stimulate fluid secretion: leucokinin (Terhzaz et al., 1999) and cardioacceleratory peptide 2b (CAP2b) (Davies et al., 1997). Both leucokinin and CAP2b stimulate an increase in intracellular calcium levels, although the effects of this increase differ with the two peptides. Leucokinin

CYCLIC GMP REGULATION AND FUNCTION

43

does not affect c G M P levels and stimulates an increase in chloride permeability, probably acting on the stellate cells of the tubule (O'Donnell el al., 1996; MacPherson, 2001). By contrast, CAP2b acts on the principal cells of the tubule to increase the activity of an apical vacuolar type H +-ATPase (VATPase) that stimulates cation transport across the tubule epithelium (O'Donnell el al., 1996). The increase in intracellular calcium triggered by CAP2b stimulates NOS, which in turn stimulates an increase in the levels of c G M P (Davies el al., 1997). The increase in c G M P probably has two effects. The major effect is to stimulate P K G activity, which activates the V-ATPase (O'Donnell et al., 1996) and there is evidence that cGMP also activates a cation-selective channel (possible the cng channel) to cause a longer lasting increase in intracellular calcium (MacPherson el a/., 2001). It is not clear how widespread this role of c G M P is in other insects. In Rhodnius prolixus Malpighian tubules, CAP2b also stimulates an increase in c G M P levels, but both CAP2b and c G M P inhibit rather than stimulate fluid secretion (Quinlan et al., 1997). Blood-feeding insects face special problems in urine production, as they need to rapidly excrete large quantities of water (Chapman, 1998). The anti-diuretic effects o f c G M P may therefore be a special case, but studies on the role of c G M P in Malpighian tubules of additional insects will be needed to determine whether stimulation or inhibition of fluid secretion is the more general action.

4

Concluding remarks

In this review we have described our current understanding of c G M P regulation and function in insects. In addition to reviewing published reports of the c G M P signaling cascade components in insects, we have summarized several of the major physiological systems that utilize cGMP. These studies highlight the diverse roles that c G M P plays in cellular physiology. There have been relatively few molecular studies that have characterized elements of the c G M P cascade in insects, but the availability of the Drosophila genome sequence now enables us to identify and predict the properties of all these components in a single organism. This analysis has revealed a wealth of genes, some of which have previously been described, but most have yet to be characterized. The list of genes that are predicted to code for components of the c G M P cascade include five receptor GCs, two receptor-like GCs, five soluble GC subunits, a possible GC-activating peptidc, five possible GCAPlike proteins, seven PDEs (of which five are predicted to hydrolyze c G M P and two might be regulated by cGMP), three PKGs, and a total of eight ion channels with cNMP-binding sites, of which at least two could be regulated by cGMP. Remarkably, there have been few genetic studies that describe the outcome of mutations in these genes. The systems that have been studied in this manner have shown profound effects: alterations in food search behavior and

44

D.B. MORTON AND M. L. HUDSON

defective synaptic connections, suggesting that this will be a fertile area for future investigations.

Acknowledgements We wish to thank Dr Alan Nighorn of the University of Arizona for providing unpublished sequence information, and Kristofor Langlais and Steve Matsumoto for comments on the manuscript. This work was supported by N I H grant NS29740.

Note added in proof A recent report (Tanoue et al., 2001) has described the cloning of a receptor G C from the silkmoth B o m b y x mori, The sequence of the extracellular domain of this GC, named BmGC-I, is almost 30'70 identical to that of mammalian atrial natriuretic peptide (ANP) receptors, including 100% conservation of the cysteine residues. This suggests that in B o m b y x , there may be ANP-like peptides that activate this GC. BmGC-I is widely expressed, including expression in olfactory sensory neurons in the antennae, glomeruli of the antennal lobes, thoracic ganglia, flight muscles, midgut and Malpighian tubules. Another recent report showed that the expression of BmGC-I in flight muscle is under circadian control (Tanoue and Nishioka, 2001). Tanoue, S., Sumida, S., Suetsugu, T., Endo, Y. and Nishioka, T. (2001). Identification of a receptor type guanylyl cyclase in the antennal lobe and antennal sensory neurons of the silkmoth, Bombvx mori. bisect. Biochem. Mol. Biol. 31, 971 979. Tanoue, S. and Nishioka, T. (2001). A receptor-type guanylyl cyclase expression is regulated under circadian clock in peripheral tissues of the silk moth. Light-induced shifting of the expression rhythm and correlation with eclosion. J. Biol. Chem. 276, 46 765-46 769.

References Ahern-Djamali, S. M., Comer, A. R., Bachmann, C., Kastenmeier, A. S., Reddy+ S. K., Beckerle, M. C., Walter, U. and Hoffmann, F. M. (1998). Mutations in Drosophih~ Enabled and rescue by human vasodilator-stimulated phosphoprotein (VASP) indicate important functional roles for the Ena/VASP homology domain I (EVH1) and EVH2 domains. Mol. Biol. Cell. 9, 2157 2171. Albin, E. E., Davison, S. J. and Newburgh, R. W. (1975). Properties of cyclic nucleotide phosphodiesterases in the central nervous system of Manduca se.vta. Biochim. Biophys. Acta 377, 364 380. Ashman, D. R., Lipton, R., Melicow, M. M. and Price, T. D. (1963). Isolation of adenosine 3'5' monophosphate and guanosine 3'5' monophosphate from rat urine. Biochem. Biophys. Res. Comm. 11,330 334.

CYCLIC GMP REGULATION AND FUNCTION

45

Bacigalupo, J., Bautista, D. M., Brink, D. L., Hetzer, J. F. and O'Day, P. M. (1995). Cyclic G M P enhances light-induced excitation and induces membrane currents in Drosophila retinal photoreceptors. J. Neurosei. 15, 7196-7200. Baker, J. D,, McNabb, S. L. and Truman, J. W. (1999). The hormonal coordination of behavior and physiology at adult ecdysis in Drosophila melanogaster. J. Exp. Biol. 202, 3037 3048. Ball, E, E. and Truman, J. W. (1998). Developing grasshopper neurons show variable levels of guanylyl cyclase activity on arrival at their targets. J. Comp. Neurol. 394, 113. Bargmann, C. 1. (1998). Neurobiology of the Caenorhahditis eleg~ans genome. Seieme 282, 2028 2033. Bargmann, C. 1., Hartweig, E. and Horvitz, H. R. (1993). Odorant-selective genes and neurons mediate olfaction in C. eh,~,,ans. ('ell 74, 515 527. Bashaw. G. J., Kidd, T., Murray, D., Pawsom T. and Goodman, C. S. (200(/). Repulsive axon guidance: Abelson and Enabled play opposing roles downstream of the roundabout receptor. ('ell 101,703-715. Baumann, A., Frings, S., Godde, M., Seifert, R. and Kaupp, U. B. (1994). Primary structure and functional expression of a Drosophih~ cyclic nucleotide-gated channel present in eyes and antennae. EMBO J. 13, 5040 5050. Bicker, G. (2001). Sources and targets of nitric oxide signalling in insect nervous systems. Cell 77ssue Res. 303, 137 146. Birnby, D. A., Link, E. M., Vowels, J. J., Tian, H., Colacurcio, P. L. and Thomas, J. H. (2000). A transmembrane guanylyl cyclase (DAF-I 1) and Hsp90 (DAF-21) regulate a common set of chemosensory behaviors in Caenorhahditis ele~ans. Genetics 155. 85 104. Bredt, D. S. and Snyder, S. H. (1994). Nitric oxide: a physiologic messenger molecule. Am~. Rev. Bioehem. 63. 175 195. Breer, H., Boekhoff, I. and Tarcilus, E. (1990). Rapid kinetics of second messenger formation in olfactory transduction. Nature 345, 65 68. Brfiggeman, A., Pardo~ L. A., Stfihmer, W. and Pongs, O. (1993). Elher-5-~o-go encodes a voltage-gated channel permeable to K ' and Ca 2- and modulated by cAMP. Natm'e 365, 445- 448. Brunet, L. J., Gold, G. H. and Ngai, J. (1996). General anosmia caused by a targeted disruption of the mouse olfactory cyclic nucleotide-gated channel. Neuron 17, 681 693. Buck, L. and Axel, R. (1991). A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell 65, 175 187. Budnik, V., Zhong, Y. and Wu, C. F. (1990). Morphological plasticity of motor axons in Drosophila mutants with altered excitability. J. Neurosci. I0, 3754 3768. Caccone, A., Garcia, B. A., Mathiopoulos, K. D., Min, G. S., Moriyama, E. N. and Powell, J. R. (1999). Characterization of the soluble guanylyl cyclase beta-subunit gene in the mosquito Anopheles ,gambiae. h~sect Mol. Biol. 8, 23 30, Carucci, D. J., Witney, A. A., Muhia, D. K., Warhurst, D. C., Schaap, P., Meima, M., Li, J. L., Taylor, M. C., Kelly, J. M. and Baker, D. A. (2000). Guanylyl cyclase activity associated with putative bifunctional integral membrane proteins in Plasmodium./~deiparum. J. Biol. Chem. 275, 22 147 22 156. Chamberlin, M. E. and Phillips, J. E. (1988). Effects of stimulants of electrogenic ion transport on cyclic AMP and cGMP levels in locust rectum. J. Exp. Zool. 245, 9 16. Champlin, D. T., Reiss, S. E. and Truman, J. W. (1999). Hormonal control of ventral diaphragm myogenesis during metamorphosis of the moth, Mamhtca sexta. Dev. GeHes Evol. 209, 265 274.

46

D. B. MORTON A N D M. L. H U D S O N

Chapman, R. F, (1998). "The Insects: Structure and Function'. Cambridge University Press, Cambridge, UK. Clyne, P. J., Wart, C. G., Freeman, M. R., Lessing, D., Kim, J. and Carlson, J. R. (1999). A novel family of divergent seven-transmembrane proteins, candidate odorant receptors in Drosophila. Neuron 22, 327 338. Copenhaver, P. F. and Taghert, P. H. (1989). Development of the enteric nervous system of the moth. II. Stereotyped cell migration precedes the differentiation of embryonic neurons. Dev. Biol. 131, 85 101. Copenhaver, P. F. and Taghert, P. H. (1990). Neurogenesis in the insect enteric nervous system: generation of pre-migratory neurons from an epithelial placode. Development 109, 17 28. Corbin, J. D., Turko, I. V., Beasley, A. and Francis, S. H. (2000). Phosphorylation of phosphodiesterase-5 by cyclic nucleotide-dependent protein kinase alters its catalytic and allosteric cGMP-binding activities. Fur. J. Biochem. 267, 2760 2767. Davies, S. A. (2000). Nitric oxide signalling in insects. Insect Biochem. Mol. Biol. 30, 1123-1138.

Davies, S. A., Stewart, E. J., Huesmann, G. R., Skaer, N. J. Maddrell, S. H. P., Tublitz, N. J. and Dow, J. A. T. (1997). Neuropeptide stimulation of the nitric oxide signaling pathway in Drosophila melanogaster Malpighian tubules. Am. J. Physiol. 273, R823 R827. Davis, R. L., Cherry, J., Dauwalder, B., Hart, P. L. and Skoulakis, g. (1995). The cyclic AMP and Drosophila learning. Mol. Cell. Biochem. 149, 271 278. Dizhoor, A. M. and Hurley, J. B. (1999). Regulation of photoreceptor membrane guanylyl cyclases by guanylyl cyclase activator proteins. Methods 19, 521 531. Dow. J. A. T., Maddrell, S, H. P., Davies, S. A., Skaer, N. J. V. and Kaiser, K. (1994). A novel role for the nitric oxide-cGMP signaling pathway: the control of epithelial function in Drosophila. Am. J. Physiol. 266, RI71~%RI719. Drewett. J. G. and Garbers, D. L. (1994). The family of guanylyl receptors and their ligands. Endocrhle Rev. 15, 135 162. Duda, T., Gorazniak, R. M. and Sharma, R. K. (1993). Core sequence of ATP regulatory module in receptor guanylate cyclases. FEBS Lett. 315, 143 148. Endo, H., Nagasawa, H. and Watanabe, T. (2000). Isolation of a cDNA encoding a CHH-family peptide from the silkworm Bombvx mori. Insect Biochem. Mol. Biol. 30, 355 361. Ewer, J., De Vente, J. and Truman, J. W. (1994). Neuropeptide induction of cyclic G M P increase in the insect CNS: resolution at the level of single identifiable neurons. J. Neurosci. 14, 7704-7712. Ewer, J., Gammie. S. C. and Truman, J. W. (1997). Control of insect ecdysis by a positive-feedback endocrine system: roles ofeclosion hormone and ecdysis triggering hormone. J. exp. Biol. 200, 869 881. Fallon, A. M. and Wyatt, G. R. (1977). Cyclic nucleotide phosphodiesterases in the cricket. Acheta domesticus. Biochhn. Biophys. Acta 480, 428M.41. Faurobert, E. Chen, C. K., Hurley, J. B. and Teng, D. H. (1996). Drosophila neurocalcin, a fatty acylated CaZ+-binding protein that associates with membranes and inhibits in vitro phosphorylation of bovine rhodopsin. J. Biol. Chem. 271, 1025610 262. Fawcett, L., Baxendale, R., Stacey, P., McGrouther, C., Harrow, I., Soderling, S., Herman, J., Beavo, J. A. and Phillips, S. C. (2000). Molecular cloning and characterization of a distinct human phosphodiesterase gene family: PDEI IA. Proc. Natl. Acad. Sci. USA 97, 3702-3707. Filburn, C. R., Karn, J. and Wyatt, G. R. (1977). Cyclic nucleotide phosphodiesterases of ttyalophora cecropia silkmoth fat body. Biochh~, Biophys. Acta 481, 152 163.

CYCLIC GMP REGULATION AND FUNCTION

47

Finn, J. T., Krautwurst, D., Schroeder, J. E., Chen, T. Y., Reed, R. R. and Yau, K. W. (1998) Functional co-assembly among subunits of cyclic nucleotide-activated nonselective cation channels, and across species fi'om nematode to human. Biophys..I. 74, 1333 1345. Fisher, D. A., Smith, J. F., Pillar, J. S., St. Denis, S. H. and Cheng, J. B. (1998). Isolation and characterization of PDE9A, a novel human cGMP-specific phosphodiesterase. J. Biol. (7hem. 273, 15559 15 564. Foster, J. L., Higgins, G. C. and Jackson, F. R. (1996). Biochemical properties and cellular localization of the Drosophila DG1 cGMP-dependent protein kinase. J. Biol. Chem. 271, 23 322- 23 328. Francis, S. H., Turko, 1. V. and Corbin, J. D. (2001). Cyclic nucleotide phosphodiesterases: relating structure to function. Proy,. Nucleic Acid Res. 65, 1 52. Friebe, A., Wedel, B.. Harteneck, C., Foerster, J., Schultz, G. and Koesling, D. (1997). Functions of conserved cysteines of soluble guanylyl cyclases. Biochemistry 36, 1194 1198.

Fiille, H.-J.. Vassar, R., Foster, D. C., Yang, R.-B., Axel, R. and Garbers, D. L, (1995). A receptor guanylyl cyclase expressed specifically in olfactory sensory neurons. Proc. N~ltl. Acad. Sci. 92, 3571 3575. Gammie, S. C. and Truman, J. W. (1997a). Neuropeptide hierarchies and the activation of sequential motor behaviors in the hawkmoth, Mamluca sex:a. J. Neurosci. 17, 4389 4397. Gammie, S. C. and Truman, J. W. (1997b). An endogenous elevation of cGMP increases the excitability of identified insect neurosecretory cells. J. Comp. Physiol. A 180, 329-338. Gammie. S. C. and Truman, J. W. (1999). Eclosion hormone provides a link between ecdysis-triggering hormone and crustacean cardioactive peptide in the neuroendocrine cascade that controls ecdysis behavior..I. Exp. Biol. 202. 343-352. Gao, F. B., Brenman, J. E., Jan, L. Y. and Jan Y. N. (1999). Genes regulating dendritic outgrowth, branching, and routing in Drosophila. Genes Dev. 13, 2549 2561. Garbers, D. L. and Lowe, D. G. (1994). Guanylyl cyclase receptors. J. Biol. Chem. 269, 31) 741-30 744. Gertler, F. B., Comer, A. R., Juang, J, L.. Ahem, S. M., Clark, M. J., Lieble, E. C. and Hoffman, F. M. (1995). crumbled, a dosage-sensitive suppressor of mutations I the Drosophila Abl tyrosine kinase, encodes an Abl substrate with SH3 domain-binding properties. Genes Dev. 9, 521 533. Gertler, F. B., Niebuhr, K., Reinhard, M., Wehland, J. and Soriano, P. (1996) Mena, a relative of VASP and Drosophila Enahled, is implicated in the control of microfilament dynamics. Cell 87. 227 239. Gibbs, S. M., Becker, A., Hardy, R. W. and Truman, J. W. (2001). Soluble guanylate cyclase is required during development for visual system function in Drosophila. J. Neurosci. 21, 7705 7714. Gibbs, S. M. and Truman, J. W. (1998). Nitric oxide and cyclic GMP regulate retinal patterning in the optic lobe of Drosophila. Neuron 20, 83 93. Gibson, N. J, and Nighorn, A. (2(/00). Expression of nitric oxide synthase and soluble guanylyl cyclase in the developing olfactory system of MaHduca se.\ta. J. Comp. Neurol. 422, 191 205. Gigliotti, S., Cavaliere, V., Manzi., A., Tino, A., Graziani, F. and Malva, C. (1993). A membrane guanylate cyclase Drosophila homologue gene exhibits maternal and zygotic expression. Dev. Biol. 159, 450M61. Goy, M. F. (1990). Activation of membrane guanylate cyclase by an invertebrate peptide hormone. J. Biol. ('hem. 265. 20 220 20 227.

48

D.B. MORTON AND M. L. HUDSON

Gray-Keller, M. P., Polans, A. S., Palczewski, K. and Detwiler, P. B. (1993). The effect of recoverin-like calcium-binding proteins on the photoresponse of retinal rods. Neuron 10, 523-531. Grueber, W. B. and Truman, J. W. (1999). Development and organization of a nitric oxide-sensitive peripheral neural plexus in larvae of the moth, Manduca sexta. J. Comp. Neurol. 404, 127--141. Grueber, W. B., Nighorn, A., Morton, D. B. and Truman, J. W. (2001). Co-localization of the guanylyl cyclase MsGC-I and frequenin suggests a mechanism for EGTAstimulated cGMP production in sensory neurons. Soc. Neurosci. Abs. 27. Harbeck, B., Huttelmaier, S., Schluter, K., Jockusch, B. M. and lllenberger, S. (2000). Phosphorylation of the vasodilator-stimulated phosphoprotein regulates its interaction with actin. J, Biol. CTwm. 275, 30817-30825. Haro, A., Garcia, J. L. and Municio, A. M. (1983). Purification and properties of a cGMP-dependent protein kinase from Ceratitis capitata pharate adults. ('omp. Biochem. Physiol. 74B, 417~424. Herman, S. B., Juilfs, D. M., Fauman, E. B., Juneau, P. and Menetski, J. P. (2000). Analysis of a mutation in a phosphodiesterase type 4 that alters both inhibitor activity and nucleotide selectivity. Mol. Pharmacol. 57, 991 999. Hesterlee, S. (1999). The identification and characterization of a novel target for eclosion hormone at the pupal ecdysis of Manduca sexta. Ph.D. Thesis, University of Arizona. Hesterlee, S. and Morton, D. B. (2000). Identification of the cellular target for eclosion hormone in the abdominal transverse nerves of the tobacco hornworm, Manduca sexta. J. Corllp. Neurol. 424, 339-355. Hewes~ R. S. and Truman, J. W. (1991). The roles of central and peripheral eclosion hormone release in the control of ecdysis behavior in Manduca sexta. J. Comp. Physiol. 168A, 697 707. Horodyski. F. M., Ewer, J., Riddiford, L. M. and Truman, J. W. (1993). Isolation, characterization and expression of the eclosion hormone gene of Drosophila melanogaster. Eur. J. Biochem. 215, 221-228. Huo, X., Abe, T. and Misono, K. S, (1999). Ligand binding-dependent limited proteolysis of the atrial natriuretic peptide receptor: juxtamembrane hinge structure essential for transmembrane signal transduction. Biochem. 38, 16 941-16 951. lnoue, M., Kishimoto, K., Takai, Y. and Nishizuka, Y. (1976). Guanosine 3', 5'-monophosphate-dependent protein kinase from silkworm, properties of a catalytic fragment obtained by limited proteolysis. J. Biol. Chem. 251, 4476-4478. Jones, 1. W. and Elphick, M. R. (1999). Dark-dependent soluble guanylyl cyclase activity in locust photoreceptor cells. Proc. R. Soc. Lond. B266, 413~419. Kalderon, D. and Rubin, G. M. (1989). cGMP-dependent protein kinase genes in Drosophila. J. Biol. Chem. 264, 10738-10748. Kawasaki, H., Springett, G. M., Mochizuki, N., Toki, S,, Nakaya, M., Matsuda, M., Houseman, D. E. and Graybiel, A. M. (1998). A family of cAMP-binding proteins that directly activate Rap 1. Science 282, 2275 2279. Kingan, T. G. and Adams, M. E. (2000). Ecdysteroids regulate secretory competence in Inka cells. J. Exp. Biol. 203, 3011-3018. Kingan, T. G., Gray, W., Zitnan, D. and Adams, M. E. (1997). Regulation of ecdysistriggering hormone release by eclosion hormone. J. Exp. Biol. 200, 3245 3256. Kingam T. G., Cardullo, R. A. and Adams, M. E. (2001). Signal transduction in eclosion hormone-induced secretion of ecdysis-triggering hormone. J. Biol. Chem. 276, 25 136 25 142. Klein, C., Kallenborn, H. G. and Radlicki, C. (1999). The 'Inka cell" and its associated cells: ultrastructure of the epitracheal glands in the gypsy moth, Lymantria di,TJar. J. hTsecl Physiol. 45, 65-73.

CYCLIC GMP REGULATION AND FUNCTION

49

Kojima, M., Hisaki, K., Matsuo, H. and Kangawa, K. (1995). A new type of soluble guanylyl cyclase, which contains a kinase-like domain its structure and expression. Biochem. Biophys. Res. Comm. 217, 993-1000. Kolodkin, A. L., Matthes, D. J., O'Connor, T. P., Patel, N. H., Admon, A., Bentley, D. and Goodman, C. S. (1997). Fasciclin IV: sequence, expression, and function during growth cone guidance in the grasshopper embryo. Neuron 9, 831 845. Kretzschmar, D., Hasan, G., Sharma, S., Heisenberg, M. and Benzer, S. (1997), The Swiss Cheese mutant causes glial hyperwrapping and brain degeneration in Drosophila. J. Neurosci. 17, 7425-7432. Krieger, ,I., Strobel, J., Vogl. A., Hanke, W. and Breer, H. (1999). Identification of a cyclic nucleotide- and voltage-activated ion channel fiom insect antennae, hzsect Biochem. Mol. Biol. 29, 255 267. Kumar, V. D., Vijay-Kumar, S., Krishnan, A., Duda, T. and Sharma, R. K. (1999). A second calcium regulator of rod outer segment membrane guanylate cyclase. ROSGCI: neurocalcin. Biochem. 38, 12614 12620. Kuo, J. F. and Greengar& P. (1970). Cyclic nucleotide-dependent protein kinases: IV. Isolation and partial purification of a protein kinase actiwtted by guanosine 3'5' rnonophosphate. J. Biol. Chem. 245, 2493 2498. K uo, J. F., Wyatt, G. R. and Greengard, P. (197l). Cyclic nucleotide-dependent protein kinases: IX. Partial purification and some properties of guanosine 3',5'-monophosphate-dependent and,adenosine 3',5'-monophosphate-dependent protein kinases from various tissues a n d species of arthropoda. J. Biol. Chem. 246, 7159 7167. L'Etoile, N. D. and Bargmann, C. I. (2000). Olfaction and odor discrimination are mediated by the C. elegans guanylyl cyclase ODR-I. Neuron 25, 575 586. Lincoln, T. M. (1989). Cyclic G M P and mechanisms of vasodilation. Pharmacol. Ther. 41. 479 502. Lincoln, T. M., Komalavilas. P. and Cornwell, T. L. (1994) Pleiotropic regulation of vascular smooth muscle tone by cyclic GMP-dependent protein kinase, t{ff~erte,sio, 23, 1141 1147. Linder. J. U., Engel, P., Reimer, A., Krtiger, T., Plattner, H., Schultz, A. and Schultz, J. E. (1999). Guanylyl cyclases with the topology of mammalian adenylyl cyclases and an N-terminal P-type ATPaseqike domain in Paramecium, Tetrahvmena and Plasmodium. EMBO J. 18, 4222 4232. Kinder, J. U., Hoffman, T., Kurz, U. and Schultz, J. E. (2000). A guanylyl cyclase from Paramecium with 22 transmembrane spans: expression of the catalytic domains and formation of chimeras with the catalytic domains of mammalian adenylyl cyclases. J. Biol. Chem. 275, 11 235 -11 240. Littleton, J. T. and Ganetzky, B. (2000). Ion channels and synaptic organization: analysis of the Drosophila genome. Neuron 26, 35~43. Liu, W., Moon. J., Burg, M., Chen, L. and Pak, W. L. (1995). Molecular characterization of two Drosophila guanylate cyclases expressed in the nervous system. J. Biol. Chem. 270, 12418 12427. Liu, Y., Ruoho, A. E., Rao, V. D. and Hurley, J. H. (1997). Catalytic mechanism of the adenylyl and guanylyl cyclases: modeling and mutational analysis. Pro(. Natl. Acad. Sci. USA 94, 13414-13419. Lowe. D. G.. Dizhoor, A. M., Liu, K., Gu, Q., Spencer, M., Laura, R., Lu, L. and Hurley, J. B. (1995). Cloning and expression of a second photoreceptor-specific membrane retina guanylyl cyclase (RetGC), RetGC-2. Proc. Natl. Acad. Sci. USA 92, 5535-5539. Lucas. K. A., Pitari, G. M., Kazerounian, S., Ruiz-Stewart. I., Park, J., Schulz, S., Chepenik, K. P. and Waldman, S. A. (2000). Guanylyl cydases and signaling by cyclic GMP, Pharmacol. Rev. 52. 375 413.

50

D. B. MORTON AND M. L. HUDSON

Ludwig, A., Zong, X., Jeglitsch, M., Hot'mann, F. and BieL M. (1998). A family of hyperpolarization-activated mammalian cation channels. Nctture 393, 587 591. McNabb, S. L., Baker, J. D., Agapite, J., Steller, H., Riddiford, L. M. and Truman, J. W. (1997). Disruption of a behavioral sequence by targeted death of peptidergic neurons in Drosophila, Neuron 19, 813 823. McNeil, L., Chinkers, M. and Forte, M. (1995). Identification, characterization and developmental regulation of a receptor guanylyl cyclase expressed during early stages of Drosophih~ development. J. Biol. Chem. 270, 7189-7196. MacPherson, M. R., Pollock, V. P., Broderick~ K. E., Kean, L., O'Connell, F. C., Dow, J. A. T. and Davies, S. A. (2001). Model organisms: new insights into ion channel and transporter function. L-type calcium channels regulate epithelial fluid transport in Drosophiht melanogaster. Am. J. Physiol. 280, C394-C407. Marx, T., Gisselmann, G., St6rtkuhL K. F., Hovemann, B. T. and Hart, H. (1999). Molecular cloning of a putative voltage- and cyclic nucleotide-gated ion channel present in the antennae and eyes of Drosophila melanogaster. Invert. Neurosci. 4, 55-63. Meredith, J., Ring, M., Macins, A., Marschall, J., Cheng, N. N., Theilmann, D., Brock, H. W. and Phillips, J. E. (1996). Locust ion transport peptide (ITP): primary structure, eDNA and expression in a baculovirus system. J. E.vp. Biol, 199, 1053 1061. Misono, K. S. (2000). Atrial natriuretic factor binding to its receptor is dependent on chloride concentration: a possible feedback-control mechanism in renal salt regulation. Chic. Res. 86, 1135 1139. Miyagi, M. and Misono, K. S, (2000). Disulfide bond structure of the atrial natriuretic peptide receptor extracellular domain: conserved disulfide bonds between guanylate cyclase-coupled receptors. Biochh~. Biophvs. Acta 1478, 30-38. Miyazu, M., Tanimura, T. and Sokabe, M. (2000). Molecular cloning and characterization of a putative cyclic nucleotide-gated channel from Drosophila melano~aster. bisect Mol. Biol. 9, 283 292. Montell, C. (1999). Visual transduction in Drosophiht. Am~. Rev. Cell. Dev. Biol. 15, 231 268. Morishima, I. (1981). Properties and distribution of guanylate cyclase in silkworm fat body. Comp. Biochem. Physiol. 68B, 567 563. Morrison, D. K., Murakami, M. S. and Cleghon V. (2000). Protein kinases and phosphatases in the Drosophila genome. J. Cell Biol. 150, F57 F62. Morton, D. B. (1996). Neuropeptide-stimulated cGMP immunoreactivity in the neurosecretory terminals of a neurohemal organ. J. Neurobiol. 29, 341 353. Morton, D. B. (1997). Eclosion hormone action on the nervous system: lntracellular messengers and sites of action. AmT, N Y Acad. Sci. 814, 40-52. Morton, D. B. (2000). Localization of the NO-insensitive soluble guanylyl cyclase, MsGC-/43, to the abdominal transverse nerves of Manduea suggests a role in eclosion hormone action. Soe. Neurosci. AhstracLs 26. Morton, D. B. and Giunta, M. A. (1992) Eclosion hormone stimulates cGMP levels in Mamh~ca sextet nervous tissue via arachidonic acid metabolism with little or no contribution from the production of nitric oxide. J. Neuroehem. 59, 1522 1530. Morton, D. B. and Simpson, P. J. (1995). Eclosion hormone-stimulated cGMP levels in Mamluca sexm CNS: Inhibition by lipid metabolism blockers, increase in inositol (1,4,5) trisphosphate and further evidence against the involvement of nitric oxide. J. Conq~. Physiol. B165, 417 427. Morton, D. B. and Simpson, P. J. (2001). Cellular signaling in eclosion hormone action. J. hlsecl. Physiol. 48, 1-13.

CYCLIC GMP REGULATION AND FUNCTION

51

Morton, D. B. and Truman, J. W. (1985). Steroid regulation of the peptide-mediated increase in cyclic G M P in the nervous system of the hawkmoth, Manduca sexta..l. Comp. Physiol. A157, 423-432. Morton, D. B. and Truman, J. W. (1986). Substrate phosphoprotein availability regulates eclosion hormone sensitivity in an insect CNS. Nature 323, 264 266. Morton, D. B. and Truman, J. W. (1988a). The EGPs the eclosion hormone and cyclic GMP regulated phosphoproteins. I. Appearance and partial characterization in the CNS of Manduca sexta. J. Neurosci. 8, 1326 1337. Morton, D. B. and Truman, J. W. (1988b). The EGPs the eclosion hormone and cyclic G M P regulated phosphoproteins, il. Regulation of appearance by the steroid 20hydroxyecdysone in Mandu~a sexta. J. Neurosci. 8, 1338 1345. Morton, D. B., Hudson, M. L., Waters, E, and O'Shea, M. (1999), Soluble guanylyl cyclases in C. elegans - NO is not the answer. Current Bioloy:y 9, R546 547. M ueller, B. K. (1999). Growth cone guidance: first steps towards a deeper understanding. Aml. Rev. Neurosci. 22, 351 388. Mfiller, U. (1996). Inhibition of nitric oxide synthase impairs a distinct form of longterm memory in the honeybee, Apis mell(/era. Neuron 16, 541 549. Murthy, K. S. and Makhlouf, G. M. (1999). Identification of the G protein-activating domain of the natriuretic peptide clearance receptor (NPR-C). J. Biol. Chem. 274, 17587 17592. Nighorn, A., Gibson, N. J., Rivers, D. M., Hildebrand, J. G. and Morton, D. B. (1998). The NO/cGMP pathway may mediate communication between sensory afferents and projection neurons in the antennal lobe of Matuh~ca se.vta. J. Neurosci. 18, 7244 7255. Nighorm A., Byrnes, K. A and Morton, D. B. (1999). Identification and characterization of a novel beta subunit of soluble guanylyl cyclase that is active in the absence of additional subunits and relatively insensitive to nitric oxide. J. Biol. Chem. 274, 2525 2531. Nighorn, A., Simpson, P. J. and Morton, D. B. (2001). The novel guanylyl cyclase MsGC-I is strongly expressed in higher order neuropils in the brain of Manduca sexm. J. F~\-p. Biol, 204, 305 314. O'Donnell, M. J., Dow, J. A. T., Huesmann, G. R., Tublitz, N. J. and Maddrell, S. H. P. (1996). Separate control of anion and cation transport in malpighian tubules of Drosophila melanogaster. J. Exp. Biol. 199, 1163 1175. Osborne, K. A., Robichon, A., Burgess, E.. Butland, S., Shaw, R. A., Coulthard, A., Pereira, H. S., Greenspan, R. J. and Sokolowski, M. B. (1997). Natural behavior polymorphism due to a cGMP-dependent protein kinase of Drosophila. Science 277. 834 -836. Ott, S. R., Jones, 1. W., Burrows, M. and Elphick, M. R. (2000). Sensory afferents and motor neurons as targets for nitric oxide in the locust. J. Comp. Neurol. 422, 521 532. Pagano, M. and Anand-Srivastava, M. B. (2001). Cytoplasmic domain of natriuretic peptide clearance receptor C constitutes Gi activator sequences that inhibit adenylyl cyclase activity. J. Biol. Chem. 276, 22 064-22 070. Palczewski, K., Polans, A. S., Baehr, W. and Ames, J. B. (2000). Ca~ -binding proteins in the retina: structure function and the etiology of human visual diseases. BioEssavx 22, 337 350. Polleux, F., Morrow, T. and Ghosh, A. (2000). Semaphorin 3A is a chemoattractant for cortical apical dendrites. Nature 404, 567 573. PoI~gs. O., Lindemeier, J., Zhu, X. R., Theil, T., Engelkamp, D., Krah-Jentgens. I., Lambrecht, H. G., Kock, K. W., Schwemer, J., Rivosecchi, R., Mallart, A., Galceran, J., Canal, 1., Barbas, J. A. and Ferrus. A. (1993). Frequcnin a novel

52

D.B. M O R T O N A N D M. L. H U D S O N

calcium-binding protein that modulates synaptic efficacy in the Drosophi& nervous system. Neuron ! 1, 15 28. Potter, L. R. and Hunter, T. (1998). Phosphorylation of the kinase homology domain is essential for activation of the A-type natriuretic peptide receptor. Mol. (k, II Biol. 18, 2164-2172. Prehoda, K. E., Lee, D. J. and Lira, W. (1999). Structure of the Enabled/VASP homology 1 domain peptide complex: a key component in the spatial control of actin assembly. Cell 97, 471-480. Quinlan, M. C., Tublitz, N. J. and O'Donnell, M. J. (1997). Anti-diuresis in the bloodfeeding insect Rhodnius prolixus Still: the peptide CAP2b and cyclic GMP inhibit malpighian tubule fluid secretion. J. Exp. Biol. 200, 2363 2367. Renger, J. J,, Yao, W. D., Sokolowski, M. B. and Wu, C. F. (1999). Neuronal polymorphism among natural alleles of a cGMP-dependent kinase gene, jbrag#,g, in Drosophila. J. Neurosei. 19, RC28 (1 8). Reynolds, S. E. (1980). Integration of behavior and physiology in ecdysis. Adv. Insect Physiol. 15, 475-595. Riesgo-Escovar, J., Piekos, W. and Carlson, J. R. (1995). Requirement for a phospholipase C in odor response: overlap between olfaction and vision in Drosophila. Proe. Acad. Natl. Sei. USA 92, 2864~2868. Robinson. N. L., Cox, P. M. and Greengard, P. (1982). Glutamate regulates adenylate cyclase and guanylate cyclase activities in an isolated membrane preparation from insect muscle. Nature 296, 354 356. Roelofs, J., Snippe, H., K[eineidam, R. G. and Van Haastert, P. J. M. (2001). Guanylate cyclase in Dietostelhm7 discoidium with the topology of mammalian adenylate cyclase. Biochem. J. 354, 697-706. Schmachtenberg, O. and Bicker, G. (1999). Nitric oxide and cyclic GMP modulate photoreceptor cell responses in the visual system of the locust. J. Exp. Biol. 202, 13-20. Scholz, N. L., Goy, M. F., Truman, J. W. and Graubard, K. (1996). Nitric oxide and peptide neurohormones activate cGMP synthesis in the crab stomatogastric nervous system. J. Neurosciem'e 16, 1614-1622. Schulz, S., Wedel, B. J., Matthews, A. and Garbers, D. L. (1998). The cloning and expression of a new guanylyl cyclase orphan receptor. J. Biol. Chem. 273, 1032 1037. Schwartz. L. M. and Truman, J. W. (1982). Peptide and steroid regulation of muscle degeneration in an insect. Science 215, 1420 1421. Schwartz, L. M. and Truman, J, W. (1984). Cyclic GMP may serve as a second messenger in peptide-induced muscle degeneration in an insect. Proc. Natl. Acad. Sci. USA 81, 6718 6722. Seidel, C. and Bicker, G. (2000). Nitric oxide and cGMP influence axonogenesis of antennal pioneer neurons. Development 127, 4541M549. Shah, S. and Hyde, D. R. (1995). Two Droxophik¢ genes that encode the oe and /4 subunits of the brain soluble guanylyl cyclase. J. Biol. (Twin. 270, 15 368 15 376. Siegel, G. J., Agranofl; B. W., Albers, R. W. and Molinoff, P. B. (1994). 'Basic Neurochemistry'. 5th edition, Raven Press, NY. Simpson, P. J., Nighorn, A. and Morton, D. B. (1999). Identification and characterization of a novel guanylyl cyclase that is related to receptor guanylyl cyclases, but lacks extracellular and transmembrane domains. J. Biol Chem. 274, 4440-4446. Smolenski, A., Burkhardt, A. M., Eiganthaler, M., Butt, E., Gambaryan, S., Lohmann. S. M. and Walter, U. (1998). Functional analysis of cGMP-dependent protein kinases I and II as mediators of NO/cGMP effects. Nau,yn-Schmiedeherg's Arch. Pharmacol. 358, 13+ 139.

CYCLIC GMP REGULATION AND FUNCTION

53

Soderling, S. S. and Beavo, J. A. (2000). Regulation of cAMP and cGMP signaling, new phosphodiesterases and new functions. Curt. Opinion Cell Biol. 12, 174 179. Soderling, S. S., Bayuga, S. J. and Beavo, J. A. (1998). Identification and characterization of a novel family of cyclic nucleotide phosphodiesterases..I. Biol. Chem. 273, 15553 15558. Sokolowski, M. B. (1980). Foraging strategies of Drosophila melam~,~aster: a chromosomal analysis. Behavior Genetics 10, 291 302. Sokolowski, M. B. (19981. Genes for normal behavioral variation: recent clues from worms and flies. Neuron 21,463-466. Solti, M., Devay, P., Kiss, 1., Londesborough, J. and Friedrich. P. (19831. Cyclic nucleotide phosphodiesterases in larval brain of wild type and dunce mutant strains of Dro.sophila melanogas,ter: isozyme pattern and activation by Ca2~,,cahnodulin. Biochem. Biophys, Res. Comm. 111, 652 658. Song, H.-J., Ming, G.. He, Z., Lehmann, M., McKerracher, L., Tessier-Lavigne, M. and Poo, M.-M. (1998). Conversion of neuronal growth cone responses from repulsion to attraction by cyclic nucleotides. Science 281, 1515 1518. Stryer, L. (19861. Cyclic GMP cascade of vision. Am1. Rev. Neurosei. 9, 87 119. Sutherland, E. W. and Rail, T, W. (1957). The properties of an adenine ribonucleotide produced with cellular particles. ATP, Mg e-~ and epinephrine or glucagon..I. Am. Chenz. Soc. 79, 3608-361 I. Takahashi, S. Y. (19851. Characterization of the guanosinc 3',5'-monophosphatedependent protein kinase from silkworm egg and analysis of the endogenous protein substrates. J. Comp. Physiol. B 155, 693-701. Teng, D. H., Chen, C. K. and Hurley, J. B. (19941. A highly conserved homologue of bovine neurocalcin in Drosol~hiht melam~gaster is a Ca2 -binding protein expressed in neuronal tissues. J. Biol. Chem. 269, 31 900 31 907. Terhzaz, S., O'Connell, F. C., Pollock, V. P., Kean, L., Davies, S. A., Veenstra. J. A. and Dow, J. A. T. (1999). Isolation and characterization of a leucokinin-like peptide of Drosophila melanoga,ster. .I. Exp. Biol. 202, 3667 3676. Tessier-Lavigne, M. and Goodman, C. S. (19961. The molecular biology of axon guidance. Scietwe 274, 1123 1133. Tremblay, J., Gerzer, R. and Hamet. P. (1988). Cyclic GMP in cell function. Adv. Secomt Mess. Phosphol~rotein Res. 22, 319 383. Truman. J. W. and Copenhaver, P. F. (1989). The larval eclosion hormone neurones in Manduca sexta: identification of the brain proctodeal neurosecretory system. J. Exp. Biol. 147, 457 470. Truman, J. W., Mumby, S. M. and Welch, S. K. (19791. Involvement of cyclic G M P in tl~e release of stereotyped behavior patterns in moths by a peptide hormone. J. Exp. Biol. 84, 201-212. Truman, J. W., De Vente, J. and Ball, E. E. (1996). Nitric oxide-sensitive guanylate cyclase activity is associated with the maturational phase of neuronal development in insects. Development 122, 3949 3958. Turko, I. V., Haik, T. L., McAIlister-Lucas, L. M., Burns, F., Francis, S. H. and Corbin, J. D. (19961. Identilication of key amino acids in a conserved cGMP-binding site of cGMP-binding phosphodiesterases, A putative NKXnD motif for cGMP binding. J. Biol. Chem. 271. 22 240 22 244. Turko, 1. V., Francis, S. H. and Corbin, J, D. (1998a). Hydropathy analysis and mutagenesis of the catalytic domain of the cGMP-binding cGMP-specific phosphodiesterase (PDE5). cGMP versus cAMP specificity. Bioche,t. 37, 4200 42(/5. Turko, I. V., Francis, S. H. and Corbin, J. D. (1998b). Binding of cGMP to both allosteric sites of cGMP-binding cGMP-specific phosphodiesterase (PDE5) is required I\w its phosphorylation. Biochem. J. 329. 5(15 510.

54

D.B. MORTON AND M. L. HUDSON

Van den Akker, F., Zhang, X., Miyagi, M., Huo, X., Misono, K. S. and Lee, V. C. (2000). Structure of the dimerized hormone-binding domain of a guanylyl-cyclasecoupled receptor. Nature 406, 101-104. Varnum, M. D., Black, K. D. and Zagotta, W. N. (1995). Molecular mechanism for ligand discrimination of cyclic nucleotide-gated channels. Neuron 15, 619 625. Vosshall, g. B., Amrein, H., Morosov, P. S., Rzhetsky, A. and Axel, R. (1999). A spatial map of olfactory receptor expression in the Drosophila antenna. Cell 96, 725 736. Wills, Z., Bateman, J., Korey, C. A., Comer, A. and Van Vactor, D. (1999). The tyrosine kinase Abl and its substrate e,abled collaborate with the receptor phosphatase Dlar to control motor axon guidance. Neuron 22, 301-312. Wilson, E. M. and Chirlkers, M. (1995). Identification of sequences mediating guanylyl cyclase dimerization. Biochem. 34, 4696~4701. Winberg, M. L., Mitchell, K. J. and Goodman, C. S. (1998). Genetic analysis of the mechanisms controlling target selection: complementary and combinatorial functions of netrins, semaphorins, and IgCAMs. Cell 93, 581 591. Wolf, E., Kim, P. S. and Berger, B. (1997). MultiCoil: a program for predicting twoand three-stranded coiled coils. Protein S(ience 6, 1179-1189. Wolff\ T. and Ready, D. (1993). Pattern formation in the Drosophila retina. In 'The Development of Drosophila mehmogaster" (M. Bate and A. M. Arias, eds), pp. 12771325. Cold Spring Harbor Press, Plainview, NY. Wright, J. W., Schwinof, K. M., Snyder, M. A. and Copenhaver, P. F. (1998). A delayed role for nitric oxide-sensitive guanylate cyclases in a migratory population of embryonic neurons. Dev. Biol. 204, 15-33. Xu, R. X., Hassell, A. M., Vanderwall, D., Lambert, M. H., Holmes, W. D., Luther, M. A., Rocque, W. J., Milburn, M. V., Zhao, Y., Ke, H. and Nolte, R. T. (2000). Atomic structure of PDE4: insights into phosphodiesterase mechanism and specificity. Science 288, 1822-1825. Yu, S., Avery, L., Baude, E. and Garbers, D. k. (1997). Guanylyl cyclase expression in specific sensory neurons: a new family of chemosensory receptors. Proc. Natl. Acad. Sci. USA 94, 3384-3387. Zagotta, W. N. and Siegelbaum, S. A. (1996). Structure and function of cyclic nucleotide-gated channels. Ann. Rev. Neurosci. 19, 235 236. Zhang, G., Eiu, Y., Ruoho, A. E. and Hurley, J. H, (1997). Structure of the adenylyl cyclase catalytic core. Nature 386, 247 253. Zhao, Y., Brandish, P. E., DiValentin, M., Schelvis, J. P. M., Babcocj, G. T. and Marietta, M. A. (2000). Inhibition of soluble guanylate cyclase by ODQ. Biochem. 39, 10848 10854. Zhao, Y., Schelvis, J. P. M., Babcocj, G. T. and Marietta, M. A. (1998). Identification of the histidine 105 in the /~1 subunit of soluble guanylate cyclase as the heine proximal ligand. Biochem. 37, 4502M-509. Ziegelberger, G., van den Berg, M. J., Kaissling, K. E., Klumpp, S. and Schultz, J. E. (1990). Cyclic G M P levels and gnanylate cyclase activity in pheromone-sensitive antennae of the silkmoths Antheraea polyphemus and Bomb)'.v mori. J. Neurosci. 10, 1217-1225. Zitnan, D. and Adams, M. E. (2000). Excitatory and inhibitory roles of central ganglia in initiation of the insect ecdysis behavioural sequence. J. Exp. Biol. 203, 1329-1340. Zitnan, D., Kingan, T. G., Hermesan, J. L. and Adams, M. E. (1996). Identification of ecdysis-triggering hormone from an epitracheal endocrine system. Science 271, 88 91. Zitnan, D., Ross, L. S., Zitnanova, 1., Hermesman, J. L., Gill, S. S. and Adams, M. E. (1999). Steroid induction of a peptide hormone gene leads to orchestration of a defined behavioral sequence. Neuron 23, 523-535.

Neurotransmitter Transporters in the Insect Nervous System Stanley Caveney a and B. Cameron Donly b ~Department of Biology, University of Western Ontario, London, Ontario N6A 5B7, Canada ~Southern Crop Protection and Food Research Centre, Agriculture and Agri-Food Canada, London, Ontario, Canada N5V 4T3

I

Introduction 56 1.1 Background 56 1.2 Molecular chemistry of insect neurons 56 1.3 Neurotransmitter uptake and vesicular storage 59 1.4 Scope of the review 61 2 Excitatory amino acid transporters 61 2.1 Net+:K+-dependent glutamate transporters 61 ") "~ Net ,K -dependent aspartate transporter 77 3 Na+/CI -dependent GABA and monoamine transporters 1 78 3.1 GABA transporters 79 3.2 Serotonin transporters 90 3.3 Dopamine transporters 99 3.4 Octopamine transporters 106 3.5 Orphan transporters 111 4 Na~:(71 -dependent transporters II 114 4.1 Choline transporters 114 5 Other Na ~-dependent transporters 121 5.1 Histamine transporter 121 6 Putative neurotransmitter transporters 123 6.1 Glycine transporters 124 6.2 Taurine transporters 124 7 Applications to insect control 125 7.1 Relevance of insect neurochemistry to pest control 125 7.2 Neurotransmitter transporters as new' targets for insect control 7.3 Future directions 127 7.4 Post-genomic prospects t\~r insect physiology research 128 Acknowledgements 129 References 129

125

Note: For the sake of consistency' this review uses a three-letter designation for neurotransmitter transporters described fiom invertebrates and a one-letter designation for those described from vertebrates. ADV,%N('t

S IN INSF(T

1%13N i i- I l 02422~) - X

I%tYSIOI O(iY

VOI.

2~)

( op ~i~ht ,

2(1()2 I:/,~e~lcJ ~,¢i¢lt~c l.td

I / / 117]lt~ ~4 ~c/iJodm tlotl lit ~m i h~lm ~'s~,/vl,I

56

1 I.l

S. CAVENEY AND B. C. DONLY

Introduction BACKGROUND

Neurotransmitters are chemical signals released from neurons at specialized synapses by exocytosis (Pennetta et al., 1999; Prokop, 1999). These neurotransmitter molecules diffuse within the synaptic space and bind to specific neurotransmitter receptors on the membranes of post-synaptic neurons, muscle cells, peripheral effector cells, glial cells surrounding the synapse, or even on the surface of the neuron that released the neurotransmitter (autoreceptors). The neurotransmitter receptor interactions involved normally change the behaviour of a target cell by altering its membrane potential or second messenger signalling pathways. This signalling event is rapidly terminated because the neurotransmitter molecules disappear quickly from the synaptic space, either by diffusion or by local uptake (as intact or partially degraded molecules) by neuroglial cells. A second class of cell surface proteins at the synapse, neurotransmitter transporters (NTTs), are responsible for the removal of neurotransmitter from the synaptic space (Fig. 1). These NTTs may be located on the surface of the pre- or post-synaptic neurons, or on the surface of glial ceils that surround the synapse. The pattern of rapid transmitter "reuptake' following transient transmitter release is an essential feature of chemical neurotransmission. Transmitter molecules taken up by axon terminals are concentrated and stored in synaptic vesicles for future use. The number of neuroactive chemicals taken up by neurons and stored in synaptic vesicles in insects is known to be rather limited. Fewer than ten amines and amino acids appear to fit the stringent criteria that define authentic neurotransmitters (Callec, 1985; Burrows, 1996), implying that only a small number of membrane transport systems may be involved in neurotransmitter recycling at the neural synapse. 1.2

MOLECULAR CHEMISTRY OF INSECT NEURONS

Formerly neurons were classified on the basis of their functional anatomy (sensory neurons, interneurons, motor neurons, region of brain, etc.) and mode of action (excitatory neurons, inhibitory neurons). These epithets reveal nothing about neuronal molecular chemistry, however. In modern times, a neuron is defined by the specific neurotransmitter molecule it stores, releases into the synaptic space and subsequently removes from it (for reviews, see Evans, 1980; Restivo and White, 1990; Buchner, 1991, Burrows, 1996: Osborne, 1996). The rate-limiting enzymes involved in the synthesis and degradation of particular neurotransmitters constitute a set of neuronal 'hallmark' proteins (Buchner, 199l; Lundell and Hirsch, 1994b). Implicit in this terminology is the notion that axon terminals contain plasma membrane transporters that return (or supply) specific neurotransmitters to the axoplasm for re-use.

NEUROTRANSMITTER TRANSPORTERS

57

Presynapse I-

GABA

~:~ Monoamines

H+

onoamines

Na+/CINa ÷

O

GI

Na+ Glu Na+/Cl -

:3K+

O

Glial cell FIG. 1 Neurotransmitter transport at the insect synapse. The diagram is a composite showing known families of Na+-dependent transport proteins in the plasma membrane of the presynaptic axon (left) and glial cells (right), and in the membrane of synaptic vesicles in the axon terminal (top left). Neurotransmitters are actively transported across the plasma and vesicular membrane against a concentration gradient. High-affinity Na+/K+-dependent excitatory amino acid transporters (EAATs, shown at O) transport the neuroactive amino acids glutamate and aspartate across the plasma membranes in glutamatergic neurons or glial cells (as shown). The inhibitory neurotransmitter GABA and several monoamine neurotransmitters (octopamine (OA), dopamine (DA) and serotonin (5hydroxytryptamine, or 5-HT)) are taken up selectively by cells in the CNS through the activity of transporters belonging to one family of N a ' / C 1 - d e p e n d e n t transporters. GABA transporters (GATs O) are found in the plasma membrane of GABA-ergic neurons and glial cells. Monoamine transporters (OATs, DATs, SERTs) are excusively neuronal in distribution (O). Each neuron selectively expresses the monoamine transporter taking up the specific monoamine neurotransmitter that the neuron releases into the synaptic space. Choline, the precursor of acetylcholine, is taken up by chotinergic neurons that express a transporter belonging to an unrelated family of Na +/C1 -dependent transporters (0). Synaptic vesicles further concentrate and sequester neurotransmitters within the axon terminal. In aminergic neurons, this involves the activity of vesicular monoamine transporters (vMATs O), vesicular acetylcholine transporters (vAchTs), vesicular excitatory amino acid transporters (vEAATs) and vesicular inhibitory amino acid transporters (vlAAT). Vesicular neurotransmitter transport is driven by transmembrane H ~ gradients (vMATs) and/ or H--dependent electrochemical gradients (vEAATs). The plasma and vesicular membranes of individual neurons normally contain transporters |\)r a single type:' class of neurotransmitter.

58

S. CAVENEY AND B. C. DONLY

Consequently, the high-affinity and Na+-dependent transporters involved are named after the neurotransmitter substrates they selectively transport (Table 1). On the other hand, neurotransmitter receptors, although named for the neurochemical they bind (or after an analogue of the natural transmitter, or an inhibitor of the receptor), are proteins primarily associated with the postsynaptic cells (although presynaptic autoreceptors modulate neurotransmitter release by some insect neurons; see Howell and Evans (1998). The specific proteins involved in neurotransmitter synthesis and transport at the nerve terminal that assign each neuron in the arthropod central nervous system its unique molecular identity are as follows. For cholinergic neurons. the hallmark proteins are the enzyme choline acetyltransferase (CHAT; Buchner et al., 1986; Kitamoto el al., 1998) and the choline transporter (CHT; Wang et al., 2001). Histaminergic neurons in the arthropod retina are identified by the presence of histidine decarboxylase (HDC: Maxwell el al., 1978; Elias and Evans, 1983b; Burg et al., 1993) and of the histamine transporter (HAT; Morgan ez al., 1999). Tyramine fi-hydroxylase (TflH; Monastirioti et al., 1996; Lehman et al.. 2000b), and the octopamine transporter (OAT; Malutan el al., 2002) together act as hallmark proteins for insect octopaminergic neurons. (Trill is the homologue of dopamine fl-hydroxylase that catalyses nor-epinephrine synthesis in mammals). For dopaminergic neurons in insects, as well as other animals, tyrosine hydroxylase (TH; Neckameyer and Quinn, 1989; Lundell and Hirsch, 1994b; Granholm el ell., 1995) and the dopamine transporter (DAT; P6rzgen el ell., 2001) are marker proteins. (DOPA decarboxylase (DDC) is not an unambiguous marker for

TABLE 1 Neurotransmitter transporters as key marker proteins for the different types of neurons in insects Neuron type Cholinergic Serotonergic

Dopaminergic Octopaminergic Histaminergic GABAergic Glutamatergic

Enzymes involved in neurotransmitter synthesis

Na+-dependent high-affinity neurotransmitter transporter

Choline acetyltransferase (CAT) Tryptophan hydroxylase (TPH), Aromatic amino acid decarboxylase (AADC or DDC) Tyrosine hydroxylase (TH), Dopa decarboxylase (DDC) Tyramine hydroxylase (TflH) Histidine decarboxylase (HDC) Glutamic acid decarboxylase (GAD) Glutaminase~' Amino acid transferases ~'

CHT SERT

~ConslittHiveenzymesin rnosl insect cells

DAT OAT HAT GAT EAAT

NEUROTRANSMITTER TRANSPORTERS

59

dopaminergic neurons since, as an aromatic amino acid decarboxylase (AADC), it catalyses the final decarboxylation step in both dopamine synthesis (Budnik and White, 1988; Kostal et al., 1998) and serotonin synthesis (Buchner, 1991; Lundell and Hirsch, 1994b) in Drosophila). In serotonergic neurons, tryptophan undergoes a two-step conversion to 5-hydroxytryptamine (5-HT) or serotonin, the first step being catalysed by tryptophan hydroxylase (TPH) (Wright, 1987; Budnik and White, 1988; Lundell and Hirsch, 1994b) and the second by DDC. In addition to TPH, the serotonin transporter (SERT: Corey et al., 1994b; Demchyshyn et al., 1994) is a hallmark protein of serotonergic neurons. The situation is less straightforward for amino acid transporters since many have non-neuronal distributions as well. GABA-ergic neurons in insects are defined by the presence of glutamate decarboxylase (GAD: Baxter and Torralba, 1975; Breer et al., 1989; Jackson et al., 1990) involved in converting glutamate to GABA, but not necessarily by the presence of high-affinity GABA transporters, which occur in glial cells. Because L-glutamate and L-aspartate are amino acids needed in intermediary metabolism and protein synthesis in many types of insect cells, it is not possible to identify glutamatergic neurons solely on the quality of specific enzymes involved in the synthesis or degradation of these amino acids, such as glutaminase and glutamine synthetase, or on the presence of high-affinity glutamate transporters (EAATs). (Glutarnine synthetase and tyrosine decarboxylase activity is found in many tissues in insects, including glial cells in the central nervous system (CNS).) 1.3

N H J R O T R A N S M I T T E R U P T A K E A N D VESICULAR STORAGE

To establish that a candidate molecule is a traditional neurotransmitter, two of the several criteria that must be satisfied are: (1) that a mechanism must exist to terminate the receptor-mediated actions of the putative transmitter following its presumed release into the synapse and (2) that the putative neurotransmitter must be concentrated within synaptic vesicles in the presynaptic terminal (Callec, 1985). With the notable exception of the neurotransmitter acetylcholine, which is degraded by cholinesterase in the synaptic space, neurotransmitters are cleared directly from the space by selective uptake systems (often called re-uptake systems) in the plasma membranes of neurons and/or glial cells, and then sequestered in synaptic vesicles by vesicular transporter systems. Neurotransmitter uptake into the nerve terminal and its synaptic vesicles is required for the nervous system to function normally (Attwell and Mobbs, 1994). Demonstration of a high-affinity membrane transport system and its corresponding lower-affinity vesicular transport system in a nerve terminal (or at least in a nerve preparation) is strong evidence for a candidate neurotransmitter acting as a local neurotransmitter. The molecular biology of the uptake systems that transfer neurotransmitters from the synaptic space into neurons and glial cells has been characterized

60

S. CAVENEY AND B. C. DONLY

during the last decade (for recent general reviews, see Reith, 1997; Amara, 1998; Beckman and Quick, 1998; Krantz et al., 1999; Masson et al., 1999). The proteins involved constitute several families of high-affinity Na +dependent neurotransmitter transporters (NTTs) that help to terminate the postsynaptic action of neurotransmitters released from neurons, and replenish their neurotransmitter content. The plasma membrane NTTs for GABA (GATs I 3), serotonin (SERT), dopamine (DAT), nor-epinephrine (NET), glutamate (EAATs 1 5), glycine (GLYT1 and 2), the NT precursor choline (CHT), as well as proline (PROT) and taurine (TAUT) (both putative NTTs) have been cloned from the vertebrate CNS (Masson et al., 1999). Notably absent from this list is a molecular description of the neuronal transporter for histamine. These NTTs belong to three structurally distinct families of co-transporter proteins, all of which are dependent on transmembrane gradients in K + and/or CI , in addition to Na +, for their normal activity. Following the selective uptake of neurotransmitter (or, in the case of acetylcholine, its precursor choline) into a nerve terminal, a low-affinity transport system located in the vesicle membrane concentrates the neurotransmitter in synaptic vesicles (Fig. 1). These vesicle-membrane carriers are powered by proton or voltage gradients across the vesicular membrane and have protein structures distinct from those of plasma membrane NTTs (Attwell and Mobbs, 1994; Worrall and Williams, 1994; Schuldiner, 1997). Vesicular transporters are generally less selective in their substrate preferences compared to plasma membrane transporters, and make up four functional groups, vesicular excitatory amino acid transporters (vEAATs, for glutamate and possibly aspartate), vesicular inhibitory amino acid transporters (vIAATs, for GABA and glycinc), vesicular monoamine transporters (vMATs, for serotonin, catecholamines, histamine, and presumably in insects, octopamine) and the vesicular acetylcholine transporter (vAChT). Molecular cloning studies indicate that there are two mammalian vMAT subgroups, vMATI and vMAT2, that belong to the same protein family as yAChT, and are distinct from the vlAAT (and likely the vEAAT) protein family (reviewed most recently in Schuldiner, 1997; Masson et al., 1999). A few hundred of the several thousand genes expressed in the adult brain of Drosophila code for the enzymes and proteins involved in the presynaptic synthesis, storage, triggered release, re-uptake and/or regeneration of neurotransmitter chemicals, and for the many pre- and post-synaptic ionotropic and metabotropic receptors to which neurotransmitters bind on release into the synaptic space (Buchner, 1991). Almost all brain-related genes in insects have corresponding homologues in mammals (Buchner, 1991; Pennetta el al., 1999). This genetic conservation underscores the value of Drosophila as a general model in neuroscience. Conversely, advances in mammalian molecular neurobiology during the last decade have also facilitated the identification of many new genes and their products in the insect CNS, Considerable progress has been made during the last 10 years in our understanding of the molecular

NEUROTRANSMITTER TRANSPORTERS

61

biology of neurotransmitter transport in the insect nervous system. Many of the cDNAs that encode selective high-affinity neurotransmitter transporters in insects have been cloned and characterized. I .4

SCOPE OF THE REVIEW

This review summarizes the current state of understanding of neurotransmitter transport across the plasma membrane in insect nervous tissue. Although research on neurotransmitter transport in the mammalian CNS has intensified over the last decade, a comprehensive review of neurotransmitter transport in the insect CNS has not been available, although Osborne (1996) presented a brief account. The sections in this review describe the molecular physiology of the various N T T systems in insects. The approach used is necessarily hierarchical and comparative. Much of what we know about the molecular biology of neurotransmitter uptake in the CNS of Dro,vophila and other insects is derived from discoveries in the mammalian CNS. The review concludes with an assessment of the potential of neurotransmitter transporters as selective neural targets for future insect control strategies.

2 2.1

Excitatory amino acid transporters Na ~ K ~ - D E P E N I ) E N T

GLUTAMATETRANSPORTERS

L-glutamate is the principal excitatory neurotransmitter in the vertebrate nervous system (Orrego and Villanueva, 1993; Meldrum, 2000). In insects, L-glutamate acts both as an excitatory and inhibitory neurotransmitter (Usherwood, 1994) and is the most abundant amino acid in the CNS (Pitman, 1985). In orthopteroid insects (locust, cockroach), glutamate activates several types of cation-selective channels (fast-acting quisqualate- and kainatesensitive ionotropic receptors), anion-selective channels (ibotenate-sensitive ionotropic receptors) and G-protein coupled receptors (slower-acting metabotropic receptors) (Usherwood, 1994). Several glutamate receptor subtypes have been cloned from Dro,s'ol~hila, and analysis of the FlyBase revealed the presence of other members of the ionotropic glutamate receptor superfamily (Littleton and Ganetzky, 2000). In the honeybee brain, glutamate has been detected immunocytochemically in descending interneurons and in visual interneurons (Bicker et al., 1988). Glutamatergic neurons may be involved in the processing of long-term olfactory memory in the mushroom bodies in the honeybee brain (Maleszka el al., 2000). L-glutamate acts as an inhibitory transmitter at extrajunctional receptors in the locust CNS (Giles and Usherwood, 1985: Wafford and Sattelle, 1989). The ibotenate-selective L-glutamatereceptor/Cl- channels present in insect neurons (Raymond et ell., 2000) are related in molecular structure to GABA-gated C1 channels and not to

62

S. CAVENEY AND B. C. DONLY

excitatory quisqualate-sensitive glutamate-gated Na + channels (Lummis et al., 1990; Usherwood, 1994; Osborne, 1996). e-glutamate is also the principal excitatory neurotransmitter at the arthropod neuromuscular junction (Usherwood, 1994; Burrows, 1996). Excitatory glutamatergic motor neurons have been shown, for instance, to innervate skeletal muscle in lobster (Takeuchi and Takeuchi, 1964), Drosophila (Jan and Jan, 1976; Johansen et al., 1989) and locust (Usherwood, 1994; Burrows, 1996), as well as hindgut visceral muscle in Leucophaea (Cook and Holman, 1979). Several hundred neurons, presumably mostly motor neurons, display glutamate-like immunoreactivity in locust thoracic and abdominal ganglia (Watson and Seymour-Laurent, 1993). The excitatory neurotransmitter status of other amino di-acids, such as k-aspartate, L-cysteate and L-cysteine sulphinate, in the insect nervous system is unresolved. These compounds could serve as excitatory neurotransmitters in the mammalian CNS (Griffiths, 1990; Orrego and Villanueva, 1993; Gundersen et al., 1998), and compete with glutamate for Na+-dependent uptake in insect tissues (Caveney et al., 1996). 2.1.1

Background

It has been known for a long time that specific Na+-dependent high-affinity transport plays a key role in inactivating glutamate released at insect neuromuscular junctions (Faeder and Salpeter, 1970; Faeder et al., 1974; Van Marie et al., 1983, 1985). Radiolabelled glutamate uptake was first demonstrated in nerve muscle preparations isolated from Gromphadorhina portentosa (Faeder and Salpeter, 1970), and then in intact abdominal nerve cords from Periplaneta americana (Evans, 1975) and thoracic and abdominal ganglia isolated from the moth Manduca sexta (Kingan and Hishinuma, 1987). The saturable component of glutamate uptake in the CNS was shown to be dependent on external Na +. The sheath (glial) cells are primarily responsible for glutamate uptake by the nerve muscle preparations (Salpeter and Faeder, 1971). Sodium-iondependent and high-affinity uptake of glutamate has also been reported in neurons derived from embryonic cerebral ganglia, but not in embryonic muscle cells, grown in vitro (Bermudez et al., 1988). Synaptosomes prepared from Drosophila also display high-affinity Na+/K+-dependent uptake of L-glutamate and k-aspartate (Ramarao et al., 1987). Employing current mammalian terminology, high-affinity Na+-dependent plasma membrane transporters for L-glutamate and k-aspartate (a putative insect neurotransmitter) are called excitatoo, amino acid transporters, or EAATs, notwithstanding the nonjunctional role of glutamate as an inhibitory neurotransmitter in insects (Usherwood, 1994; Raymond et al., 2000). The less-restrictive acronym GLUT was earlier assigned to cloned facilitated glucose transporters (Thorens, 1996). EAATs are also found in non-neural tissues in mammals (Hediger and Welbourne, 1999; Palacin et al., 1998) where they fulfill a variety of non-neurotransmitter related functions.

NEUROTRANSMITTER TRANSPORTERS

63

The molecular biology and pharmacology of EAATs cloned and characterized from the insect nervous system will now be described and compared in light of the wealth of vertebrate information currently available. However, it should be mentioned that several non-excitable insect tissues display Na +dependent uptake of glutamate. These include the epidermis (McLean and Caveney, 1993; Caveney el al., 1996), fat body and hindgut (H. McLean, unpublished) of the beetle T e n e b r i o m o l i t o r and the epidermis of several orthopteroid insects (S. Caveney, unpublished). A role proposed for this non-neuronal uptake is to keep glutamate levels in the haemoplasm below the activation threshold of the neuromuscular synapses (Irving el al,, 1979; McLean and Caveney, 1993; Tomlin el al., 1993). 2.1.2

Slructure

Glutamate transporters constitute a distinct family of Na+-dependent transporters, with no significant homology to any other known protein family. Members are known from all three superkingdoms, Eukaryota, Bacteria and Archaea, with those found among the eukaryotes being specific for either glutamate/aspartate or neutral amino acid substrates. The EAATs (those specific for glutamate/aspartate) are present in multiple forms in many organisms, with humans, for example, containing five different paralogues fl'om this protein family. The five human EAATs share 40-60% amino acid sequence identity among themselves, whereas comparisons of the same type of transporter among different species of mammals yield much higher identities o{" approximately 90%. The molecular biology and pharmacology of EAATs in the mammalian CNS have been reviewed extensively (Takahashi el ell., 1997; Palacin el al., 1998: Vandenberg, 1998; Masson el ell., 1999; Seal and Amara, 1999: Slotboom et al., 1999; Danbolt, 2001; Gadea and gopez-Colome, 2001a). The global structure of the mammalian EAATs has also been found by analysis of peptide hydrophobicity to be highly conserved. The profiles suggest a common tertiary structure comprising eight transmembrane c~-helical domains (TMDs) with several fi-pleated pore-loop structures in the vicinity of c<-helical TMDs 7 and 8 (Grunewald and Kanner. 2000; Seal et ell., 2000) (Fig. 2). This pore loop structure resembles part of the fi-barrel structure formed by the subunits of several multimeric pore-forming ion channels (such as the nicotinic acetyleholine receptor), implying that mammalian EAATs form homomultimeric complexes in the plasma membranes of neurons and glial cells (Danbolt et al., 1990: Haugeto et al., 1996: Danbolt, 2001). Both N- and C-termini of EAAT proteins lie within the cytoplasmic space. The study of EAATs in invertebrates has been facilitated by the wealth of structural information initially obtained from cloned vertebrate molecules. Amino acid sequence comparisons between the known mammalian transporters were used to identify, regions of conserved sequence homology characteristic of this family of transporters. Degenerate PCR (polymerase

64

S. CAVENEY AND B. C. DONLY

~, -AAXFIAQ

NH~"

-DWXLDR-

PKC /

-TA/~;SSS/A-

PKC COOH

FIG. 2 Membrane topology of Na '-dependent glutamate transporters. Hydrophobicity analysis suggests that excitatory amino acid transporters contain 6-8 or-helical transmembrane domains (shown here as TMDs I to 7, 10) and 2-4 fi-pleated domains (shown as TMDs 8 and 9). Most motifs specifically involved in glutamate transport and ion movements are found on the highly-conserved C-terminal half of EAATs. Amino acid substitution in this region, particularly around TMDs 8 and 9, alters the selectivity of substrate binding and transport kinetics. After Slotboom et al. (1999), Grunewald et al. (1998) and Seal et al. (2000). See text for details.

chain reaction) primers encoding the most conserved sequences were then used to amplify fragments of homologous c D N A s in invertebrates such as nematodes (Radice and Lustigman, 1996) and insects (Donly et al., 1997; Seal et al., 1998). Insect EAATs have now been cloned from the CNS of the cabbage looper Trichoplusia ni (trnEAAT1), the cockroach Diploptera punctata (dipEAAT1), the fruitfly Drosophila melanogaster (drmEAATs 1 and 2), and the honeybee Apis mellifera (apmEAAT) (Table 2). In a phylogenetic analysis of E A A T proteins, the amino acid sequences of dipEAAT1, trnEAAT1 and d r m E A A T I suggest that they are not similar to any particular subtype of mammalian glutamate transporter (Fig. 3). Based on amino acid sequence similarities, these three insect transporters show the greatest relatedness to hEAAT1 and hEAAT3, with dipEAAT1 showing 46% and 47%, trnEAATI showing 40% and 42%, and d r m E A A T I showing 46% and 47% sequence identity to these two human proteins, respectively. The same is true from a functional perspective, as mammalian EAAT1 and EAAT3 are also difficult to distinguish pharmacologically, either from each other or from the lepidopteran trnEAAT1 (Donly et al., 1997). There is however, a distinct dichotomy among the invertebrate EAATs in the dendrogram, with those above clearly separated from a group including the second fruitfly E A A T

NEUROTRANSMITTER TRANSPORTERS

65

>

>

c~

Z

z z

~ o ~ ° ~

r-

%

c~

L~ ,_m

©

Fi ©

o,,

%

C'4 0

S -,s

"4 c./)

S

..b ~

f-,l

..a .<

o

cq

~--

[.--

[.--

'<

.<

<:

c L

66

S. CAVENEY AND B. C. DONLY

apmEAAT

drmEAAT2

caeF_.MT

hEAAT2

hEAAT4~

981 -hEAAT3 1000

hEAAT1

dipEAAT1

hEAAT5 tmEAATI

FIG. 3 Dendrogram/bootstrap analysis of insect excitatory amino acid transporters (EAATs) compared with those in other animals. Sequence similarity was measured by aligning amino acid sequences using ClustalX 1.81 (Thompson el al., 1997) and an unrooted tree calculated using the neighbour joining method employed by the program. Confidence values for the derived tree were determined by bootstrapping the dataset using 1000 replicates and a generator seed value of 333 (ClustalX 1.81). Alignments were output in Phylip format to Treeview (1.6.5) for display. The accession numbers of the aligned amino acid sequences are: nematode (caeEAAT; QI0901), cockroach (dipEAAT1; AAF71701), caterpillar (trnEAAT1; AAB84380), honeybee (apmEAAT; AAD34586), fly (drmEAATI: AAD09142 and drmEAAT2: AAD47830), human (hEAATI; AAA50428, hEAAT2; AAA50429, hEAAT3; AAAS0430, hEAAT4; AAA75314 and hEAAT5; AAB53971).

NEUROTRANSMITTER TRANSPORTERS

67

(drmEAAT2), and those of the honeybee (apmEAAT) and nematode (caeEAAT), which cluster more closely with the mammalian EAAT2 type transporter (Fig. 3), The honeybee transporter apmEAAT shows greatest (54%) sequence identity to vertebrate EAAT2 (Kucharski et al., 2000) and nematode EAATs. A pharmacological comparison of hEAAT2 with caeEAAT and apmEAAT remains to be made. This suggests the likelihood that other invertebrates would, like the fly, have at least two EAAT type transporters. However, except for the fly, no more than one EAAT has as yet been reported in any invertebrate. Since the C. eh,gans genome is complete, it would seem that this nematode is equipped with only a single type (although it does potentially have at least two alternatively spliced forms).

2.1.2.1 Functional domains q f E A A T s . In addition to the significant advances made in the molecular and pharmacological characterization and localization of neurotransmitter transporters, research on mammalian highaffinity transporters has identified many structural components needed for them to function normally (Palacin et al., 1998: Slotboom et ell., 1999: Danbolt, 2001). Of particular interest are the amino acid residues involved in substrate specificity and affinity, substrate and ion translocation, and transport inhibition. Other residues have been found to be important in protein targeting, and conformation and assembly in the plasma membrane. The effects on protein function that result from changing specific amino acids in the protein have been analysed by site-directed mutagenesis. Substrate selectiviO' domains. Glutamate binding to mammalian metabotropic glutamate receptors (mGluRs) is coordinated by about twelve polar (tyr, ser) and acidic charged amino acid residues (asp, glu, lys, arg) spread along the extracellular ligand binding domains of the dimeric receptor molecule (Kunishima et al., 2000). Glutamate transporters, too, have many polar and charged residues in their extracellular and cytoplasmic domains. A serine-rich region, T(A/G)SS(S/A) found in most EAATs, shown on the cytoplasmic loop between TMDs 6 and 7 in Fig. 2, is thought to be involved in substrate binding. This serine-rich region is present in most insect glutamate transporters (TASSS in trnEAAT; TGSSS in drmEAATl; TSSNA in dipEAAT; TASSA in apmEAAT). A second motif, AAXFIAQ, is thought to bind a carboxylate group of the substrate diacid (Slotboom et al., 1999). This motif is located in putative TMD 7 in the model shown in Fig. 2. Ion permeation site, A highly conserved sequence, NMDGTLYEA, has been found in all insect and mammalian EAATs cloned to date. It constitutes part of the cation permeation channel (Kavanaugh et al., 1997). The position of this motif in different structural models for EAATs is controversial. In rat GLT-I (EAAT2) it is modelled as part of TMD 7 (Fig. 2) (Grunewald and Kanner,

68

S. CAVENEY AND B. C. DONLY

2000), while in GLAST (EAATI) it is either in a short cytoplasmic loop connecting fl-pleated regions lying between TMDs 8 and 9 (Wahle and Stoffel, 1996) or part of an extracellular loop associated with this Ê-structure (Seal et al., 2000). Mutation of the glutamate residue in the NMDGTLYEA sequence abolishes the normal interaction between an EAAT and cytoplasmic K + ions, converting it into a non-electrogenic glutamate/aspartate exchanger (Kavanaugh et al., 1997). The aspartate and glutamate residues in this sequence (D398 and E404 in rGLT-I) also appear to be required for glutamate transporters to discriminate between their primary transport substrate and Dand L-aspartate (Pines et al., 1995). The glutamate residue in the LYEA sequence, which may confer coupled co-transport of glutamate and sodium and counter-transport of potassium (Palacin et al., 1998), is not conserved in other Na+/K+-dependent amino acid transporters such as the Drosophila aspartate transporter, drmEAAT2 (Besson et al., 2000) and the mammalian neutral amino acid transporter, ACST (Slotboom et al., 1999). The Drosophila aspartate transporter drmEAAT2 has an isoleucine residue (I392) in place of this glutamate residue, suggesting that drmEAAT2 may act more in glutamate/ aspartate exchange than in net amino acid uptake (Besson et al., 2000). N a + binding sites. These sites presumably lie near or within the extracellular domains contiguous to the two or three putative/~-pleated structures identified by Wahle and Stoffel (1996) in TMDs 7, 8 and 9.

Zn ;+ is a non-competitive partial inhibitor of glutamate uptake in the vertebrate brain (Gabrielsson et al., 1986; Vandenberg et al., 1998; Spirodon et al., 1998). In the related two human glutamate transporters, hEAAT1 (GLASTI) and hEAAT4, this Zn 2+ sensitivity is thought to be conferred on the transporters by a Zn2+-binding site formed by two histidine residues (His146 and His156 in hEAAT1) in conjuction with a cysteine residue (C186) in the second extracellular loop (Vandenberg et al., 1998). Although a comparable Zn2+-binding domain is located at the leading edge of the second extracellular loop in dipEAAT1, containing a pair of histidines (His117 and His127) as well as the cysteine residue, its transport activity is unaffected by Zn 2+ (Donly et al., 2000). Nevertheless, the presence of this motif aligns dipEAAT1 more closely with hEAATI (which also contains one) than with hEAAT3. The Zn2+-binding motif has a limited distribution in glutamate transporters. These histidine residues are absent in trnEAATI and drmEAATI, and apmEAAT has only the first one. The Zn2+-binding site is also missing from hEAAT2 and hEAAT3. Z&c-binding site.

P K A and P K C phosphoo'&tion sites. Several potential sites exist on both N and C terminals in both insect and mammalian EAATs. trnEAAT1, for example, has a K-SR consensus sequence for PKC phosphorylation on its C-terminal cytoplasmic tail (Fig. 2).

NEUROTRANSMI-FFER TRANSPORTERS

69

N-linked glycosylation sites. All insect glutamate transporters have one to several potential glycosylation sites on asparagine residues in the large extracellular loop between TMDs 3 and 4 (Fig. 2). Dihy~h'okai, ate ( D H K ) binding site. A kainate-binding site has been identified as involving two serine residues, Ser441 and Ser444 in mammalian EAAT2 (Zhang and Kanner, 1999). The transport inhibitor, dihydrokainate, blocks glutamate uptake by mammalian EAAT2 isoforms that possess this motif. This pair of serines is present in the kainate-sensitive Drosophila aspartate transporter (Besson et al., 2000), but absent from other cloned insect EAATs, which appear to be insensitive to kainate (Donly et al., 1997; Seal el al., 1998). tfistidine +326". A histidine residue located in TMD 6 of all cloned glutamate transporters is thought to be involved in the translocation of protons accompanying Na+/K+-dependent glutamate uptake (Zhang et al., 1994). Although this histidine residue (His326 in rat GLT-1) is required for glutamate uptake (Zhang et al., 1994), it apparently is not needed for aspartate exchange by drmEAAT2, where it is replaced by tyrosine (Besson et al.. 1999). Chloride c h a m w l domain. A carboxy-terminal motif ES/xxv has been identified with the presence of a glutamate-gated, but transport-independent, chloride conductance in hEAAT5 (Arriza et al., 1997) (see section 2.1.4). This motif is present in several vertebrate ligand-gated ion channels and potassium channels, where it promotes subcellular targeting and clustering (Arriza el al., 1997). While this sequence is absent from trnEAAT1, dipEAAT1 and both drmEAATs, the ET part is preserved in apmEAAT. All five mammalian EAATs have an integral chloride channel (Seal and Amara, 1999) and drmEAATI has also been shown to have a glutamateinduced chloride conductance (Seal et al., 1998). 2.1.3

Distribution

The distribution of glutamate transporters EAATI (GLAST), EAAT2 (GLT-1), EAAT3 (EAAC1), EAAT4 and EAAT5 in the vertebrate brain has been studied immunocytochemicalIy using antibodies raised to their unique Cand N-terminal peptide domains. Mammalian EAATs have distinctive regional and cellular patterns of distribution. EAATs 1 4 are expressed in the cerebrum and cerebellum of brain (reviewed in Danbolt, 2001; Gadea and LopezColome, 2001a) and EAAT 5 in the retina (Arriza et al., 1997; Gegelashvili and Schousboe, 1998). EAATs 1 and 2 (GLT-1 is short for 'glial transporter protein 1') are exclusively glial in location, found both in astroglial processes and glial somata (Danbolt el al., 1992; Rothstein el al., 1995; Gegelashvili et a/,, 2000). Most GLT-1 protein is located near nerve terminals, axons and

70

S. CAVENEY AND B. C. DONLY

dendritic spines (Chaudhry et al., 1995). EAATs 3, 4 and 5 are expressed by neurons and are generally located some distance from the synaptic regions (Coco et al., 1997; Gegelashvili et al., 2000). EAAT3 was originally described as presynaptic in glutamatergic neurons although it is also found in presynaptic terminals and dendritic trees of GABA-ergic Purkinje neurons (Rothstein et al., 1995; Furuta et al., 1997). These five mammalian glutamate transporters are also expressed selectively in a variety of non-neuronal tissues (Palacin et al., 1998: Hediger, 1999). The cellular distribution of EAATs in the insect nervous system is less clearcut. Transcripts of caterpillar trnEAAT1 (Donly et al., 1997) and cockroach dipEAAT1 (Donly et al., 2000) are expressed in the CNS and at glutamatergic neuromuscular junctions (NMJ). In situ hybridization showed that trnEAATI m R N A is expressed in ganglia by a small population of interracial glial cells (Fig. 8) but not by neurons (Fig. 4). Peptide-specific antibodies raised to the Nand C-terminal regions of trnEAAT1 stain glial membranes throughout the ganglionic neuropile, implying that these interracial glial cells branch extensively within the neuropile (Fig. 4c,d). Glial EAAT expression throughout the neuropile presumably allows glutamate levels in the extra-axonal space to be kept low and favours net glutamate diffusion away from its synaptic release sites. These antibodies also stain the membranes of glial cells that surround motor neurons where they contact body wall muscle, but not elsewhere along the nerve fibres (Fig. 4a,b). The trnEAATl-immunoreactive motor endings are somewhat similar in appearance to glutamate-immunoreactive endplates in Drosophila skeletal muscle (Johansen et a/., 1989). The trnEAATl-positive glial membranes form a linear series of spots (of approximately 0.5#m in diameter) on both sides of the axon of the motor neuron where it runs over the muscle surface (Fig. 4b). These sites presumably demarcate the individual glutamatergic synapses. A similar pattern has been described at the NMJ in the tobacco hornworm M a n d u c a sexta (Rheuben and Reese, 1978). Sodium-iondependent glutamate uptake in insects with high blood Na + levels is unlikely to be restricted to the NMJ and CNS. In the crab Carcinus maenas, for instance, [3H]glutamate is taken up by non-synaptic glial cells surrounding peripheral nerves excised from the walking legs (Evans, 1974). Glial-specific expression of trnEAAT1 RNA in the CNS resembles that of mammalian EAAT1 rather than that of EAAT3, which in the CNS is expressed exclusively in neurons. Although the cell type(s) expressing dipEAATI in the cockroach CNS is not known, several other insect EAATs are thought to be neuronal. Transcripts coding for two other Na+-dependent glutamate transporters have been shown by in situ hybridization to be present in the brain of the honeybee (Apis mell(/era EAAT, Kucharski et al., 2000), and retina and brain interneurons in the fly (Drosophila melanogaster EAATs 1 and 2, Besson et al., 1999). The highest level of expression of apmEAAT is in the median region of the calycal cups of the mushroom bodies in the brain, an area packed with the cell bodies of the innermost compact Kenyon cells (Kucharski el al..

NEUROTRANSMITTER TRANSPORTERS

71

FIG. 4 Distribution of a high-affinity glutamate transporter m the nervous system of the caterpillar Trichoplusia hi. (a) Motor neuron running over the surface of at segmental muscle fibre probed in situ with trnEAAT1 antisense RNA. The punctate areas likely correspond with the perisynaptic areas of glial cell(s) that surround the axon of the motor neuron, as shown unstained in (b). The nucleus of a glial cell (n) is present at the nerve branch. (Micrograph courtesy of Tabita Malutan.) (b) Neuromuscular junction probed with a polyvalent antibody raised to a C-terminal peptide of trnEAAT1. The proximal (adsynaptic) surface of the glial cell (GC) layer surrounding the unstained axon terminal (ax) is strongly EAAT-immunoreactive~ particularly at spots (row of arrows) representing the synaptic sites. (c) Segmental ganglion stained with antibody raised to a C-terminal peptide of trnEAATI. The neuropile (np) is strongly EAAT-immunoreactive, particularly its outer margins (see also Fig. 6). The cortex and connectives are unstained. (d) lmmunogold localization of trnEAATI in the neuropile of a ganglion. EAAT1 staining is spread diffusely over the glial membranes (glm) surrounding the axons (ax). (Micrographs B, C and D courtesy of Richard Gardiner.)

72

S. CAVENEY AND B. C. DONLY

2000). The authors argue that the expression is neuronal, since glial cells are sparse in the calyx region of the mushroom bodies, apmEAAT is also expressed strongly in the optic lobes, and to a lesser extent in the lateral protocerebral lobe, central body and antennal lobe (Kucharski et al., 2000), suggesting that glutamatergic neuronal pathways play a significant role in the processing of sensory information in the insect brain. The human transporter closest in peptide sequence to apmEAAT is hEAAT2 (Fig. 3), which is exclusively glial in distribution. This raises the possibility that the honeybee may have separate glial EAAT and/or motor-neuron-associated glutamate transporters (as suggested by Maleszka et al., 2000). As discussed above, this again raises the general question of whether the nervous system of individual insect species possesses more than one type of EAAT, as seen in vertebrates. 2.1.4

Kinetics and pharmacology

The data available on the kinetic properties of cloned insect EAATs are summarized in Table 2. Most insect EAATs apparently have an equal or greater affinity for L-aspartate (a putative neurotransmitter) over L-glutamate (Seal et al., 1998: Besson et aL, 2000; Donly et al., 2000). dipEAAT1 and drmEAATI are more selective in their affinity for L-aspartate over D-aspartate. D-aspartate is normally a high-affinity competitive inhibitor of glutamate transport by most EAATs, but is a weak inhibitor of dipEAAT1 activity. The affinity of dipEAAT1 for [3H] D-aspartate is an order of magnitude lower than that of trnEAAT1 (Table 2), revealing an unusual stereoselectivity for aspartate in the cockroach transporter (Donly et al., 2000). A similar trend is seen for drmEAAT1 (Seal ez al., 1998). drmEAAT1 expressed in COS cells had a lower affinity for L-glutamate (Km = 72#M) than when expressed in Xenopus oocytes (Kin = 26 #M) (Seal et al., 1998). Glutamate transport by EAATs is electrogenic (i.e. causes a change in transmembrane potential) since it is coupled to an asymmetrical transmembrane flux in several inorganic ions (Fig.l). Typically, uptake of one glutamate anion is coupled to the co-transport of three Na + ions, one H +, and the countertransport of one K + ion, resulting in the net translocation of two positive charges per transport cycle (Zerangue and Kavanaugh, 1996; Levy et al., 1998). A transmembrane Na + gradient is absolutely essential for the transporter to pump glutamate against its concentration gradient into the cell. The sequence in which these ligands associate with EAAT during the translocation cycle is ordered and thought to involve: (1) Na + binding; (2) glutamate binding; (3) proton binding, on the outside; (4) their translocation into the cell; (5) K + binding on the inside: and (6) its translocation out of the cell (Kanner and Bendahan, 1982). Recent findings (Auger and Attwell, 2000) amend this sequence. The transmitter removal step may be extremely rapid compared with the overall transporter cycle time, with proton movement following glutamate translocation. Consequently, two models have been

NEUROTRANSMITTER TRANSPORTERS

73

proposed to describe the role of glutamate transporters at glutamatergic synapses in vertebrates. The first model proposes that EAATs cycle slowly, at about one-tenth the cycle time of a glutamate receptor (Wadiche et al., 1995), and promote diffusion-driven clearance of glutamate from the synaptic cleft by reversibly buffering the local glutamate concentration, In this model the transporter turnover time is much slower than the half-life of the postsynaptic ion channels gated by glutamate (less than 10ms). A second model, based on recent observations using Purkinje cells, suggests instead that glutamate uptake by transporters is extremely rapid (with time constants in the order of 0.8ms) and that most of the synaptically released glutamate is taken up directly by post-synaptic (largely glial) transporters (Auger and Attwell, 2000). There is, however, no a priori reason why both mechanisms might not exist in different regions of the nervous system. Glial glutamate transporters are clustered at the insect neuromuscular synapse in the larva of T. , i (model 2), but also spread diffusely throughout the ganglionic neuropile (model 1) in this insect, which may have few glutamatergic synapses. Here the role of EAAT may be to buffer the extracellular levels of glutamate throughout the neuropile. The ion Zn 2+ is a non-competitive partial inhibitor of glutamate uptake by the glial transporter EAAT1 in the vertebrate brain (Gabrielsson et al., 1986) and retina (Spirodon et al., 1998). Zn 2+ also increases the anion (chloride) conductance associated with transport activity in hEAATI. As mentioned above, the presence of a putative Zn 2+ binding domain in dipEAAT1 does not appear to confer Zn2+-sensitivity to glutamate transport in this cockroach transporter (Donly et al., 2000). Among the remaining insect EAATs, only apmEAAT contains even a partial motif. The reason for this difference is unclear in the absence of any demonstrable effect from its presence in the cockroach transporter. The uptake of [3HI glutamate by insect EAATs (including drmEAAT2, an aspartate transporter) is inhibited strongly by linear analogues of both aspartate and glutamate, such as L- and D- t h r e o - 3 - h y d r o x y a s p a r t a t e , L- and ~-cysteate, and by cyclic analogues of glutamate that are constrained in an open configuration, such as (2S,3S,4R)-cis-(carboxycyclopropyl)glycine (CCG 11I) and L - t r a n s - p y r r o l i d i n e - 2 , 4 - d i c a r b o x y l a t e (LPDC) (Donly et al., 1997, Seal et al., 1998: Besson et al., 2000: Donly et al., 2000). These cyclic compounds compete with L-glutamate for high-affinity Na+-dependent uptake into insect cells (Caveney et al., 1996: Donly et al., 1997). Based on relative affinities for these various analogues, trnEAATl is more similar in its pharmacology to hEAATI than the other four cloned human glutamate transporters (Donly et aL, 1997). LPDC is a particularly high-affinity blocker of mammalian EAAT2, and has been proposed to block L-glutamate uptake by apmEAAT in the bee brain (Maleszka et al., 2000). Other less potent competitive blockers of glutamate transport in insects include L-serine-O-sulphate, l~-aspartate-fihydroxamate and I - a m i n o c y c l o b u t a n e - t r a n s - l , 3 - d i c a r b o x y l a t e (Caveney et

74

S. CAVENEY AND B. C. DONLY

al., 1996; Donly et al., 1997). The algal toxin kainate (a mammalian EAAT2-

specific blocker) and cycad toxin 3-N-oxalyl-L-2,3-diaminopropionate are reported to block glutamate uptake by synaptosomes prepared from Drosophila brain (Ramarao et al., 1987). Kainates are potent blockers of glutamate uptake by drmEAAT2 (Besson et al., 2000), but are ineffective on drmEAATI (Seal et al., 1998) and trnEAAT1 (Donly et al., 1997). All of the compounds listed above are, to some extent at least, stereoconformers of L-glutamate and presumably interact with the EAAT glutamate-binding site. No non-competitive and selective blockers of high-affinity glutamate transporters are currently available to allow study of the pharmacological properties of EAATs, unlike the situation for other neurotransmitter transporters such as GATs and the various monoamine transporters. The properties of some EAATs have begun to break down the distinction between ion-coupled transporters and ion channels. One such property, first revealed in human cerebral EAAT4 (Fairman et al., 1995) and retinal EAAT5 (Arriza et al., 1997), is the presence of a glutamate-gated anion channel integral to the protein structure of the transporter. When the fly transporter drmEAAT1 was expressed in frog oocytes, the glutamate-induced current in voltage-clamped oocytes reversed at -2mV, consistent with an inwardmovement of chloride along a channel formed during glutamate transport (Seal et al., 1998). This current was abolished when gluconate replaced either external or internal C1 (Seal et al., 1998). In humans, glutamate is implicated in the hippocampal neuronal circuitry associated with short-term memory (Meldrum, 2000). Glutamate may have a comparable role in olfactory memory in the honeybee. Pharmacological disruption of glutamate uptake in the honeybee brain following injection of the selective glutamate transport blocker LPDC was shown to impair the olfactory associative learning and long-term memory in this insect (Maleszka et al., 2000). The site of action in the bee's brain appears to be the compact Kenyon neurons in the mushroom bodies, which express an EAAT (Kucharski et al., 2000) and are known to be involved in olfactory memory (Maleszka el al., 2000). Application of LPDC had no effect on acquisition and short-term memory. This may be the first observation of transporter involvement in memory and learning (albeit indirect through perturbing glutamatergic signalling in the brain).

2.1.5 Regulation A variety of signalling mechanisms are thought to differentially modulate the synthesis, surface expression and membrane trafficking of mammalian EAATs (reviewed in Beckman and Quick, 1998; Gadea and Lopez-Colome, 2001a). Prolonged exposure of astrocytes to glutamate induces the synthesis of EAATI mRNA (Gegelashvili et al., 1996). Shorter pre-incubation with glutamate

NEUROTRANSMITTER TRANSPORTERS

75

causes an activity-dependent upregulation of glutamate transport in glial cells expressing EAAT1 and 2 (Duan el al., 1999; Munir et al., 2000). This glutamate-induced activity, which is not seen in EAAT3, is also triggered by other high-affinity transport substrates such as D-aspartate and LPDC (Munir et al., 2000). This increase in transport activity results from a net trafficking of transporter protein to the plasma membrane (Davis et al., 1998; Gegelashvili el ell., 2000; Munir el al., 2000). In glioma cells, translocation of EAAT3 to the cell surface appears to be under protein kinase C control (Davis el al., 1998). The subcellular redistribution and consequent modulation of EAAT activity at the cell surface may involve an interaction with membrane proteins of the SNARE system (Gadea and Lopez-Colome, 2001a). In mammalian EAATs 1,4 and 5 this may involve a PDZ-like binding domain in the C-terminal region (Arriza el al., 1997; Marie and Atwell, 1999; Gadea and Lopez-Colome, 2001a). PDZ domains are 80-90 residue predominantly fi-structures that allow proteins to bind to the C-terminal 4 5 residues of target proteins, which include many transmembrane receptors and neurotransmitter-gated ion channels, The consensus sequence (typically XEs/TXv/FCOOH) includes valine, leucine, or sometimes methionine, as the hydrophobic C-terminal residue. The residues at -2 (polar) and -3 (acidic) positions are important in determining specificity. Mammalian EAAT1 (GLAST), EAAT4 and EAAT5 have the C-terminal sequence SETKM-COOH, NESAM COOH and LETNV COOH, respectively. None of the insect EAATs cloned to date, however, appear to have this combination at their C-termini, suggesting that EAAT protein interactions in insect cells may be based on other consensus sequences. These sequences are drmEAATI, HEMKE COOH; drmEAAT2, CNRRV COOH; trnEAATI, EKGDH COOH; dipEAATI, EEDGL COOH; apmEAAT,-FLSET COOH. The activity of mammalian EAATs is differentially modulated by protein kinase A- or protein kinase C- induced phosphorylation (Casado et ell., 1993; Davis et al., 1998; Schlag et al., 1998; Daniels and Vickroy, 1999). Phorbol ester-induced direct phosphorylation of EAATI by protein kinase C suppresses glutamate uptake in cells expressing EAATI (Conradt and Stoffel, 1997), whereas it stimulates glutamate uptake in cells expressing EAAT2 (Casado et ell., 1993). Arachidonic acid released into the extracellular space during glutamatergic signalling also modulates glutamate uptake (Nicholls and Atwell, 1990; Zerangue et al., 1995). Soluble factors acting via a receptor tyrosine kinase pathway stimulate EAAT2 expression, but not EAATI, in glial cells (Gegelashvili et al., 1997, 2000). The extent to which the modulation in transporter activity is due to a shift in net membrane trafficking or to in situ phosphorylation/dephosphorylation in the plasma membrane is unresolved. Preliminary data suggest that the activity of trnEAAT1 is down-regulated by arachidonic acid, but not up-regulated when cells expressing this EAAT are pre-incubated in saline containing glutamate (J. Burman, unpublished data). This promises to be a rewarding direction for future work.

76

2.1.6

S. CAVENEY AND B. C. DONLY

The glutamine etch"

The glutamine cycle is a concept based on observations of the vertebrate CNS. It states that glutamate and GABA released from presynaptic neurons are inactivated through their uptake by perisynaptic glial elements (reviewed in Robinson et al., 1998). These neurotransmitter molecules are rapidly converted to glutamine in the glial cells and then returned to the extracellular space, from where glutamine is recovered by the nerve terminals and converted back to glutamate or GABA, depending on the characteristics of the recipient neuron. This metabolic trafficking of amino acid neurotransmitters between neurons and glial cells (both astrocytes and oligodendrocytes) in the vertebrate CNS is a consequence of the restricted distribution of the enzymes involved in glutamine cycling. Glutaminase, which converts glutamine to glutamate and NH +, is enriched in nerve terminals, whereas glutamine synthetase (GS), which converts glutamate to glutamine by the energy-dependent addition of NH~-, is present only in glial cells (Robinson et al., 1998). Both glutamate transporters (GLT-I, and GLASTl) and GABA transporters (GATs 1 and 3) are expressed in vertebrate glial cells (reviewed in Danbolt, 2001; Gadea and Lopez-Colome, 200lb). The formation of glutamine from GABA in glia follows a more complex route, involving transamination to succinate, conversion to oe-ketoglutarate, transamination to glutamate and finally amidation to glutamine. The final step, the decarboxylation of glutamate to GABA by GAD (glutamate decarboxylase) takes place in the GABA-ergic neurons. In order for a glutamine cycle to operate in the insect nervous system, an EAAT and GS would have to be co-expressed in glial cells. GS occurs at moderate levels in the cerebral neuropile (but not in the cortex) and in the lamina neuropile and retina of the optic tract in the adult Drosophila CNS (Chase and Kankel, 1987), These regions of the CNS are rich in glial cells. drmEAAT2 is also expressed by glial cells in the fly CNS (Besson et al., 1999). Thus, it is probable that the glutamine cycle operates in the CNS of the fly. In fact, it was suggested many years ago that glial GS is involved in the inactivation of neurotransmitter glutamate in the peripheral NS at the neuromuscular synapse in cockroach and locust (Faeder and Salpeter, 1970; Botham et al., 1978) and that glial cells recycle glutamine to the neurons for glutamate and GABA synthesis (Kingan and Hildebrand, 1985). In lepidopteran skeletal muscle, which apparently lacks GS (Levenbook and Kuhn, 1962), the case for a glutamine cycle is less compelling. While there is immunocytochemical evidence that EAAT is expressed by peripheral glial cells at the lepidopteran neuromuscular synapse (Gardiner et a/., 2002), GS has not been reported at this site. High levels of GS do occur in dipteran flight muscle fibres, however, (Chase and Kankel, 1987; Dowton et al., 1988), although in flies the neuromuscular junctions lie deep within grooves in the muscle and lack a glial sheath (Dowton el al., 1988). Although it has been shown that in flies skeletal muscle GS can convert paralysing doses of exogenously applied glutamate to

NEUROTRANSMITTER TRANSPORTERS

77

glutamine (Irving et al., 1979), its involvement in processing synaptically released (i.e. neurotransmitter) glutamate in this tissue has not been determined. A high-affinity EAAT has not been localized to the sarcolemma of insect skeletal muscle. 2.2

Na 4 ,K--DEPENDENT

ASPARTATE

TRANSPORTER

Aspartate was proposed to be an excitatory transmitter at slow neuromuscular junctions in the M u s c a domestica maggot (Irving and Miller, 1980). All motor neuron terminals in Drosophila skeletal muscle, however, are glutamateimmunoreactive (Johansen et al., 1989) and it currently remains uncertain whether L-aspartate plays a significant role as an insect neurotransmitter. In mammalian neurons, the synaptic vesicle concentrating mechanism for glutamate was thought not to recognize aspartate as a transport substrate, leading to the conclusion that aspartate may not be an authentic neurotransmitter (Orrego and Villanueva, 1993; Fillenz, 1995). Recent immunogold co-localization studies reveal, however, that both aspartate and glutamate are concentrated at rat hippocampal synaptic terminals and are co-released upon K+-induced depolarization (Gundersen et al., 1998). It is possible that L-aspartate may be a co-transmitter in some glutamatergic motor neurons in insects. Certainly it is premature to conclude that aspartate is not a natural neurotransmitter at certain mammalian and insect synapses. It is noteworthy that an aspartate-selective EAAT has recently been found in the Drm'ol~hila CNS (Besson el al., 2000). Besson and colleagues first reported the occurrence of a second glutamate transporter from the genome of the fly, drmEAAT2, with a pattern of expression in the embryonic and larval nervous system, as well as in the adult eye, that differed from that of drmEAATI (Besson el al., 1999). The drmEAAT2 cDNA lacks several residues associated with highaffinity glutamate transporters, namely the histidine residue (which is replaced by tyrosine) associated with T M D 6, and the glutamate residue in the amino acid sequence LYEA lying near or within (Wahle and Stoffel, 1996) the /q-pleated TMDs 8 and 9 (Fig. 2). drmEAAT2, in addition to drmEAAT1, is a functional amino acid transporter in the fly (Besson et al., 2000). drmEAAT2 is, however, unusual in that it has a high affinity for k-aspartate but a low affinity for L-glutamate (Table 2). The canonical histidine in T M D 6 required for glutamate transport activity by EAATs (rGLT-I; Zhang et al., 1994) is apparently not required for aspartate uptake (and/or exchange) by drmEAAT2. Analysis of the Drosophila genome sequence in FlyBase revealed no other gene sequences similar to drmEAATI and drmEAAT2 (Table 2) that might represent further EAATs. It is peculiar that some glutamate transporters (drmEAAT1, apmEAAT) are reportedly expressed exclusively or primarily in neurons in some insects, while others are expressed only in glial cells (trnEAAT1, possibly drmEAAT2) in others. One plausible answer is that high-affinity glutamate transporters belonging to other families of transport

78

S. CAVENEY AND B. C. DONLY

proteins remain to be identified; another possibility is that the genome sequence data for Drosophila are as yet incomplete (although the number of gaps continues to shrink). So it remains to be shown that L-aspartate is an authentic neurotransmitter in insects and other animals. Since this amino acid may be a substrate for both glutamate and GABA synthesis, aspartate transport may be more important in neurotransmitter synthesis than directly in signalling. 2.2.1 Characteristics drmEAAT2 is the only cloned insect EAAT reported to be sensitive to kdihydrokainate (DHK) (Besson et al., 2000). drmEAATI (Seal et al., 1998) and trnEAAT1 (Donly et al., 1997) are insensitive to kainate and DHK. In mammals, only EAAT2 is kainate-sensitive (Vandenberg, 1998). Synaptosomes prepared from Drosophila CNS display high-affinity Na+/K+-dependent uptake of L-glutamate and L-aspartate that is reported to be inhibited by both kainate and N-oxalyl-L-~,fi-diaminopropionate (Ramarao et al., 1987). It appears that the transport system examined was drmEAAT2-1ike and its synaptosomal isolation confirms a neuronal expression pattern.

3

Na+/CI--dependent GABA and monoamine transporters I

Members of this family are involved in the cellular uptake of several small molecule neurotransmitters in the insect CNS, namely the amino acid F-aminobutyric acid (GABA), the indolamine serotonin (5-HT) and several phenolamines, including dopamine (DA), octopamine (OA) and the putative neurotransmitter tyramine (TA) (Osborne, 1996; N/issel, 1996). Although not identified in insects, neuronal transporters for glycine (an inhibitory transmitter in the mammalian brain and spinal column but not shown to be an insect neurotransmitter) and for proline and taurine (putative neurotransmitters in vertebrates) also belong to this family (Lill and Nelson, 1998). All of the gene products in this family have the same general structure. Sequence and hydrophobicity analyses predict that Na+/C1--dependent transporter proteins have 12 transmembrane domains (TMDs). These transporters are glycoproteins, as they possess an extended extracellular loop (EL2) between TMD 3 and TMD 4 with up to six N-glycosylation sites. Both N- and Cterminal regions of the proteins are on the cytoplasmic side of the plasma membrane. Based on molecular structure, the neurotransmitter transporter representatives of this family have been assigned to three subfamilies, the GABA (plus betaine and taurine) transporters (GATs), the amino acid (glycine and proline) transporters, and the monoamine (indolamine and catecholamine) transporters (Lill and Nelson, 1998). A fourth subfamily, the so-called 'orphan transporters', is composed of proteins that are generally larger in size than

NEUROTRANSMITTER TRANSPORTERS

79

those of the other groups and have to date not been proven to have any transport abilities. These "orphan' proteins instead may facilitate inorganic ion fluxes across the plasma membrane or act as cellular osmosensors and plasma membrane linkers to the submembranous cytoskeleton. 3.1 3.1.1

GABATRANSPORTERS Background

GABA is the principal inhibitory neurotransmitter in the nervous system of both vertebrates (Borden, 1996) and insects (Callec, 1985; Van Marie et al., 1985: Sattelle, 1990, 1992). Glutamic acid decarboxylase (GAD), the enzyme involved in GABA synthesis, is found in cell bodies and nerve processes in the insect CNS (Breer and Heilgenberg, 1985; Buchner et al., 1986: Breer et ell., 1989). Approximately 3000 GABA-ergic neurons and their axon terminals have been immunocytochemically mapped in the visual system of Drosophila CNS using antisera against GABA and G A D (Buchner et al., 1986). These GABA-ergic cells are predominantly interneurons, and include the C2 interneurons of the fly optic tract (Meyer el al., 1986). In the brain of the moth Manduca sexta approximately 18 000 interneurons are GABA-Iike immunoreactive (Homberg et al., 1987). There are also discrete clusters of cell bodies of GABA-like immunoreactive inhibitory neurons present in the thoracic and abdominal ganglia (Kingan and Hildebrand, 1985). Such cell clusters are also seen in the locust (Watson, 1988) and similar lineages of GABA-ergic neurons occur in the thoracic ganglia in many other insect orders (Witten and Truman. 1998) as well as crayfish (Mulloney and Hall, 1990). GABA-ergic motor neurons are found in hemimetabolous insects such as Schistocerca gregaria, Locusta migratoria and Periplaneta americana (Sattelle, 1990; Burrows, 1996). This type of common inhibitory motor neuron is apparently absent in holometabolous insects such as M. sexta (Witten and Truman, 1998) and D. melanogaster (Featherstone et al., 2000), No GABAlike immunoreactivity was detected in axon (motor neuron) projections exiting the ganglia along nerve roots in the hawkmoth, although immunoreactive axons were seen in the interganglionic connectives (Witten and Truman, 1998). GABA-like immunoreactive peripheral projections were also found to be lacking from five species representing the other three major orders of higher insects (Witten and Truman, 1998). In this respect the Holometabola are like the Vertebrata, where GABA-ergic terminals are found only in the CNS. lonotropic (GABAAR) and metabotropic (GABABR) receptors for GABA are widely distributed in the central and peripheral nervous systems in insects (Sieghart, 1995; Burrows, 1996: Hosie et al., 1997). Both types of GABA receptor are l\~und in the cockroach (Sattelle et al., 2000). The RDL-type of GABA receptor is the target of commercial insecticides such as dieldrin (Tanaka et al., 1984; Bloomquist, 1996). Structural mutation of these receptors

80

S. CAVENEY AND B. C. DONLY

resulting in resistance to such insecticides has been revealed (ffrench-Constant and Roush, 1991) and is commonly exhibited in insects (Georghiou, 1986). Since insecticide resistance poses a serious problem in pest management, the structure function relationship of GABA receptors has been extensively studied in order to find the mechanisms behind resistance (Sattelle, 1992; Hosie el al., 1997). The main mechanism for the inactivation of GABA at the synapse is through re-uptake into presynaptic terminals or perisynaptic glial cells (Callec, 1985). GABA uptake by arthropod nervous tissue was first demonstrated in glial cells at the lobster neuromuscular junction (Orkand and Kravitz, 1971). GABA is also taken up by the optic lobes in M u s c a and Drosophila (Campos-Ortega, 1974). In the cockroach P. americana, [3H]GABA uptake has been demonstrated in the brain (Frontali and Pierantoni, 1973) and the sixth abdominal ganglion, where autoradiography showed grains over the glial cells forming the perineurial sheath as well as over the outer (cortical) and inner (neuropile) glial cells of the ganglion (Hue et al., 1982). Few grains were found over the soma or axons of neurons. The cortical staining pattern was found to be sensitive to fl-alanine, a potent competitive inhibitor of GABA uptake by low-affinity GATs (such as mammalian GATs 3 and 4) but not by high-affinity GATs (Hue el al., 1982). Cockroach embryonic nervous tissue also displays highaffinity GABA uptake (Bermudez et al., 1988). In Schistocerca gregaria, GABA accumulates in glial cells that form a sheath around both excitatory and inhibitory neuromuscular synapses in motor units of the tibial extensor muscle (Van Made et al., 1983). Sodium-ion-dependent neurotransmitter transporters typically require high extracellular Na + levels to function efficiently. In the Lepidoptera, haemolymph levels of Na + are typically very low and this might perhaps limit the utility of such transporters in the peripheral nervous system. However, the absence of peripheral GABA-ergic neurons in Lepidoptera cannot be explained in this fashion, as the lepidopteran Na +dependent glutamate transporter trnEAATI is expressed peripherally in perijunctional glial cells (Gardiner et al., 2002; Malutan et al., 2002). 3.1.2 Structure The characterized GABA transporter cDNAs encode a set of related transporters, as evidenced by their primary sequence homology (Guastella et al., 1990; Nelson et al., 1990; Liu et al., 1993a), that in mammals are distinguishable by their dissimilar pharmacological profiles (Borden, 1996) and patterns of tissue expression (Palacin et al., 1998; Gadea and Lopez-Colome, 2001b). The first high-affinity GABA transporter cloned from an insect was from embryonic M a n d u c a sexta (masGAT; Mbungu et al., 1995). Subsequently the cDNA coding for a GABA transporter in the cabbage looper CNS (trnGAT; Gao et al., 1999), and partial sequences for some splice variants of a Drosophila GABA transporter, have been reported (Neckameyer and Cooper, 1998). The

NEUROTRANSMITTER TRANSPORTERS

81

amino acid sequences of trnGAT and masGAT are 95% identical. Hydropathy analysis of the deduced sequences of insect GATs (Mbungu et al., 1995; Gao et al., 1999) suggests the protein has 12 transmembrane domains, a structure similar to all mammalian GABA transporters cloned to date (Nelson et al., 1990; Borden et al., 1994; Clark and Amara, 1994; Rasola et al., 1995). More recent N-glycosylation analysis of rat GAT-1 (Bennett and Kanner, 1997) and glycine transporter GLYTI (Olivares el al., 1997) has provided a revised topological model that is probably applicable to all members of the Na+/ Cl--dependent transporter family (Fig. 5). 3.1.2.1 Functional domains o f G A T s . Along with significant advances in the molecular and pharmacological characterization and localization of neurotransmitter transporters, research on mammalian GATs has identified many structural components of high-affinity transporters essential for their normal function (reviewed in Miller et al., 1997; Palacin et al., 1998). Of particular

~ @ . , j leucine zipper

(MA pT~;C~

E

i'lI'0bidis

-EFWER- (GA I) _EF/yV/yXR- (MATs)

Tricyclic drug

\

~ \

_SXXS/T_

~--- PKC

PKA FIG. 5 Revised topological model for Na+/CI-dependent GABA/monoamine neurotransmitter transporters. The 12-transmembrane c~-helical segment model originally proposed now includes a 'pore-loop' structure (the former TMDI) associated with the membrane near the N-terminal end of the protein (Bennett and Kanner, 1997: Olivares et al., 1997). In this model the originally larger extracellular loop EL2 is shortened and the piece removed incorporated with the original TMD3 to form the two TMD 3 and 3*. The N-terminal region of the transporter contains most of the structural motifs shown to be involved in ion and neurotransmitter binding. See text for details.

82

S. CAVENEY AND B. C. DONLY

interest are the amino acid residues involved in substrate specificity and affinity, substrate and ion translocation, and transport inhibition. Other residues have been found to be important in protein targeting and conformation. Much of this information has been gained using site-directed mutagenesis, which allows the functional consequences of replacing a specific amino acid in the peptide with another amino acid to be analysed. Substrate bimting site(s). A tyrosine residue near TMD 3* (Yl40 in mammalian GATs. YI31 in insect GATs) is critical for GABA recognition and transport. This residue is conserved throughout the Na+/Cl--dependent transporter family and recognizes the amino group of the ligand. Mammalian GATI mutants in which phenylalanine and tryptophan are substituted for tyrosine (Y140F and Y140W) bind Na + and C1 but lack the ability to bind and transport GABA (Bismuth et al., 1997). A tryptophan residue in TMD 4 (W234 in trnGAT) is also thought to be involved in GABA binding (Kleinberger-Doron and Kanner, 1994: Soehnge et al., 1996). It is absent from mammalian (Corey et al., 1994a) or insect monoamine transporters (e.g. drmDAT or drmSERT). Finally, the characteristic aspartate residue in the pore loop that is critical to monoamine transport by other members of this family (Kitayama et al., 1992; P6rzgen el al., 2001) is replaced by glycine (G54 in trnGAT) in insect GABA transporters and the Drosophila inebriated (ine) gene product. hm-permeation site. Mammalian GAT-I has also been used to determine the potential interactions of Na + and C1- with certain amino acid residues, and consequently their role in GAT function. Three amino acids identified through mutation studies of rat GAT-1 as important in transport function are W68, R69 and C74 (Mager et al., 1996: Yu et al., 1998). These form part of a highly conserved sequence, W R F P Y L C associated with a pore-loop structure in this protein family (membrane domain I in Fig. 5) (Bennett and Kanner, 1997: Olivares et al., 1997). Mutation to W68L reduces the activity of GAT-1 by increasing its affinity for Na +, essentially 'locking' Na + on to GAT-1 (Mager el al., 1996). R69 is involved in C1 coupling (Pantanowitz el al., 1993; Jursky el al., 1994). These residues are conserved in insect GABA transporters as W59, R60 and C65 in lepidopteran GATs. Aromatic amino acids, especially tryptophan, are thought to interact with positively charged ions through their 7r electrons (Sussmann, 1991). A second conserved residue, El01 in rat GAT-I, lying within the leucine zipper motif of TMD 2, is also implicated in Na + binding. This residue (E in Fig. 5) appears to be involved in conformational changes of GAT-1 during its transport cycle (Keshet et al., 1995). Heptan leucine zipper m o t i f Many Na+/CI -dependent neurotransmitter transporters possess a characteristic L-x(6)-L-x(6)-L-x(6)-L motif in the highly conserved TMD 2. Other neutral amino acids (particularly methionine, but also isoleucine, alanine and valine) may substitute for leucine in any one or

NEUROTRANSMI-I-i-ER TRANSPORTERS

83

more of the seven-repeat positions, trnGAT and masGAT are good examples (L-x(6)-L-x(6)-M-x(6)-M). This sequence may be involved in transporter dimerization or interaction with modulatory proteins. E F W E R sequence in EL2. All cloned insect GATs contain this motif in the variable extracellular region between TMDs 3* and 4 (Fig. 5). The motif is shared with mammalian transporters for neutral amino acids (GABA, taurine, creatine, glycine and proline) and for betaine. The tryptophan residue in the EFWER sequence is missing in the monoamine transporters in this family. Although the EFWE part of this motif is seen in metabotropic glutamate receptors (Kunishima et al., 2000), it has not been implicated in ligand binding. The Drosophila inebriated gene encodes a putative transporter lacking this motif, making it unlikely that the ine product is a GABA transporter, as first proposed (Soehnge et al., 1996). Comparison of the caterpillar GAT sequences against the Drosophila genome indicates that peptide sequence CG1732 is the likely fl'uitfly GAT homologue (Table 3). Analysis of the CGI732 sequence shows it is lacking almost half the region predicted to constitute an extracellular loop between TMDs 3 and 4, including the second N-glycosylation site and the EFWER motif. However, analysis of the genomic sequence in this area shows that this motif is encoded within the proximal intron, suggesting that CG1732 does in fact conform to this structural paradigm, and that the exons in this region have been incorrectly predicted by exon scanning of the genomic sequence. This is also likely true of the N-terminus of the predicted CG1732 peptide, which is truncated when compared with the lepidopteran GATs. The true characteristics of the fruitfly GAT must await confirmation from the isolation of a complete eDNA for this locus. P K A and P K C phosphowlation sites. Neurotransmitter transporters typically have consensus phosphorylation sites at their N and/or C terminals, and along their cytoplasmic loops, suggesting that their activity may be modulated directly by protein kinase A and C phosphorylation and by protein phosphatases (Beckman and Quick, 1998). Direct tyrosine phosphorylation may also occur (Law et al., 2000). trnGAT has three putative cytoplasmic protein kinase C (PKC) phosphorylation sites, consistent with other GAT subtypes. In an effort to define the role of phosphorylation in the GABA transport process, several groups have assessed the effects of phosphorylation modulators on mammalian GATs. Cupello et al, (1993) fl~und that the V,..... of GABA uptake into rat brain synaptosomes is increased 58 74% by the PKC activators phorbol 12-myristate 13-acetate (PMA) and oleyl-acetyl glycerol (OAG). Modulation of GABA transport activity by phorbol esters, a phosphorylation activator, has also been reported. N-linked glyco.u, lation sites. Analysis of the predicted amino acid sequences of masGAT and trnGAT indicates that there are two potential N-glycosylation sites within EL2. Most mammalian GABA transporters have three

S. CAVENEY A N D B. C, DONL

84 ~.~ •

~

~

~"~

~

0",

. o ~... = ~

Z

©

< <

< < < < < < < <

,..-,

[.-. [~

~F-

E ~=

~5

[.<

©

~

~

,-n

r-..,

Z

,.8=

Z

~

~

~

0

"U.

,_1,..-, <

L)~

NEUROTRANSMITTER TRANSPORTERS

r'

f,,~ r,

~

~

85

L~ ~..~ ,_.) L~,. L,;,., , ~ f,.,~,

~j

0

4

~

~,,-.,

~u O'-

~j c~

ce-,

.~

'~..c

...:

c.-

~

=~

•,~ p .

©

..~

r,".

--

"2

©

86

S. CAVENEY AND B. C. DONLY

N-glycosylation sites within EL2 (Miller et al., 1997). The functional role of N-glycosylation has been examined in mammalian GAT-1 (Keynan et al., 1992). Incubation of GAT-1 expressing cells with tunicamycin, a specific inhibitor of N-glycosylation, caused a decrease in GABA transport that was dependent on the time of exposure to the inhibitor. 3.1.3 Distribution The tissue expression pattern of trnGAT has been examined by Northern blotting (Gao et al., 1999). Total RNA from a variety of Trichoplusia ni caterpillar tissues was hybridized with a 5' fragment of the trnGAT cDNA and a single band of about 4.4 kb was observed in the brain sample, matching the size of the largest clone isolated from the cDNA library. No bands were detectable in other tissues. As with the masGAT gene (Mbungu et al., 1995), only a single transcript was detected. This is in contrast to Drosophila, where Neckameyer and Cooper (1998) concluded, based on the heterogeneity of GAT cDNA clones isolated from a Drosophila head library, that the fruitfly GAT gene gives rise to multiple isoforms of transporter via alternative splicing. The presence of multiple transcripts was not, however, confirmed by Northern analysis. In situ hybridization of whole mounts of caterpillar CNS probed with trnGAT cDNA identified a small number of glial cells arranged in bilaterally symmetrical pairs at the neuropile/cortical layer interface in the thoracic ganglia (Fig. 6). This pattern of trnGAT expression was serially repeated in the abdominal ganglia (T. Malutan, unpublished). In the tobacco hornworm M. sexta, six neuroblast progenitors give rise to five paired post-embryonic GABA-ergic lineages and one medial unpaired lineage in the thoracic hemiganglion (Witten and Truman, 1991, 1998). The distribution of the GATpositive interface glial cells seen in T. ni partly matches the pattern of distribution of these lineage clusters and the large embryonically-derived GABA-ergic neurons that lie adjacent to them. The apparent absence of GABA-ergic motor neurons in many holometabolous insects, including Manduca (Witten and Truman, 1998) and Drosophila (Featherstone et al., 2000), may explain why GAT transcripts have not been localized by in situ methods to caterpillar neuromuscular junctions (T. Malutan, unpublished data). GAD, a marker enzyme for GABA-ergic terminals, is also restricted to the CNS in adult flies (Buchner, 1991). 3.1.4

Kinetics and pharmacology

The lepidopteran trnGAT has kinetic properties and ionic stoichiometry (Gao et al., 1999) similar to those of mammalian neuronal GABA transporter subtype GAT-1 (Guastella et al., 1990). The Km of GABA uptake by moth GATs is within the 1 20/zM range reported for high-affinity GABA uptake systems in mammals (Borden, 1996; Palacin et al., 1998). The pharmacological profile of lepidopteran GATs is distinct from those of the mammalian GAT subtypes, however. Both insect and mammalian high-affinity GABA

NEUROTRANSMITTER TRANSPORTERS

87

transporters require external Na + (Corey et al., 1994a) and a monovalent inorganic anion, normally C1- (Mabjeesh et al., 1992; Gadea and LopezColome, 2001b), to function. The CI- requirement of trnGAT is less rigid than its strict dependence on external Na +. Significant uptake activity is observed when CI is replaced by other halide anions (Gao et al., 1999). The capacity of the anions to substitute for CI- seems to be related to their ionic radii (Goncalves and Carvalho, 1994). Monovalent anions such as Br- and NO3 that have relatively small radii are reported to partially substitute for CI in driving GABA uptake by mammalian GATs. GABA uptake by trnGAT is not sustained by monovalent organic anions such as acetate and gluconate and divalent anions such as sulphate and phosphate. The activity of masGAT, on the other hand, was unaffected by acetate substitution for CI-, and only halved alter gluconate substitution (Mbungu el al., 1995). The affinity of trnGAT for C1- (Kin = 10mM at 105raM Na + and 10/~M GABA) is similar to that of mammalian GAT-I (Kin = 19 mM). The higher C1 affinity of trnGAT compared with mammalian transporters may be an adaptation to a lower [C1-] in lepidopteran blood plasma and perineurial spaces. The reported apparent stoichiometries of Na + and Cl--dependent GABA transport are either 2 Na~: 1 CI-: 1 GABA (mammalian GAT-I and GAT-3) or 3 Na+: 2 CI-: 1 GABA (canine GABA/betaine transporter BGT-I (GAT-2)). Consequently, GABA uptake is electrogenic at physiological pH. Cyclic analogues of GABA, specifically nipecotic acid, its derivative SKF 89976A and cis-3-aminocyclohexanecarboxylic acid, are potent inhibitors of Na+-dependent GABA uptake by mammalian GAT-I (Borden, 1996). They are weak inhibitors of GABA uptake by lepidopteran GATs (Mbungu el al., 1995; Gao et al., 1999). The most potent inhibitor of trnGAT tested was tra~lsaminocrotonic acid, a linear analogue of GABA constrained in an extended conformation (Gao el al., 1999). Two other linear GABA analogues, diaminobutyric acid (DABA) and fi-guanidinopropionic acid (fiGPA) are weaker inhibitors of GABA uptake by trnGAT. Betaine and fi-alanine have little effect on GABA uptake by trnGAT and masGAT, suggesting that the moth transporters fail to recognize these compounds as substrates. This is also seen in mammalian GAT-I but not mammalian GAT-2 and GAT-3. The inability of the lipophilic nipecotic acid analogue SKF 89976A to block GABA uptake by trnGAT is noteworthy. SKF 89976A is a potent blocker (Ki in the 0.[/,M range) of high-affinity GABA uptake by mammalian GAT-1. trnGAT would appear to discriminate against GABA analogues that act as GABA transport substrates in mammalian nervous tissue, suggesting that the unique pharmacology of GABA transport in the lepidopteran CNS may be a useful target for insect control. 3.1.5

Regulalion

GABA transporter expression and function in the mammalian CNS is controlled by multiple cytoplasmic signalling pathways (Beckman and Quick,

88

S. CAVENEY AND B. C. DONLY

1998; Gadea and Lopez-Colome, 2001b). GATs have multiple PKC phosphorylation consensus sites. Protein kinase C activation downregulates GABA transport by GAT-1 possibly by affecting GAT interaction with cytoplasmic trafficking proteins such as syntaxin (Beckman el al., 1998). Two inhibitors of protein-tyrosine kinase, genistein and compound 252a, downregulate wild-type GAT-I activity (Law el al., 2000). Phosphorylation of GAT-1 catalysed by tyrosine kinase causes the transporter to be removed from the cell surface (Law et al., 2000). Co-treatment of GAT-l-expressing cells with saturating concentrations of PKC activator and protein-tyrosine kinase inhibitor had an additive effect on depressing GAT-1 activity, suggesting that independent pathways for GAT-I regulation exist (Law et al., 2000). GABA transport in mammals is also influenced by ambient levels of external GABA (Bernstein and Quick, 1998). Although insect GATs contain multiple phosphorylation sites (Mbungu el al., 1995; Gao et al., 1999), GAT regulation via second messenger signalling pathways in insects remains to be investigated. GAT activity may also be influenced by neuroactive substances secreted into the synaptic cleft. GABA uptake by mammalian brain glial cells is stimulated by epinephrine and suppressed by 5-HT (Gadea and Lopez-Colome, 2001b). In the crab Cancer borealis, proctolin is co-released with GABA by motor neurons in the stomatogastric (suboesophageal) ganglion controlling the gastric mill rhythm (Blitz and Nusbaum, 1999). It is plausible that proctolin, or any other neuropeptide co-released with GABA, might modulate GAT activity through a second-messenger-mediated event involving transporter phosphorylation/dephosphorylation, but this has not been shown. A reduction in GABA neurotransmission has been implicated in the pathophysiology of a variety of neurological disorders in humans, particularly in epilepsy (Suzdak, 1993; Gadea and Lopez-Colome, 2001b). Since GABA-ergic transmission is terminated by GABA uptake into the presynaptic terminals and/or the surrounding gila by high-affinity transporters specific for this neurotransmitter, disruption of GABA transport may increase GABA-ergic function by elevating the extracellular levels of GABA (Gadea and LopezColome, 2001b). Novel pharmacological agents that disrupt GABA transport, such as the lipophilic derivatives of piperidenecarboxylic acid, tiagibine and SKF-89976A, exhibit anticonvulsant properties (Borden, 1996). Studies on GABA transport may provide next-generation insecticides to control agriculturally important insects by perturbing GABA-ergic neurotransmission. 3.1.6

Uo-localization o f E A A T and GA T in glial cells

The intimate relationship between neurons and glial cells in CNS and neuromuscular junctions in mammals suggests that glial cells play an integral modulatory role in synaptic neurotransmission (Araque et al., 1999). The synapse may be regarded as a functionally tripartite structure, or 'gliapse" (Galambos, 1961), made up of the pre- and postsynaptic nerve endings and surrounding

NEUROTRANSMI'FIER TRANSPORTERS

89

glial cell(s) (Fig. 1). According to this concept, the role of the glial cells is not simply to provide nutrients and maintain the extracellular ionic environment needed by neurons to function, but to fine-tune synaptic activity by releasing neurotransmitter molecules in a Ca:+-dependent fashion into the synaptic space (Araque et al., 1999). For example, neurotransmitter released by presynaptic neurons during synaptic transmission may bind to glial receptors and induce them to release their own neurotransmitter store, causing a feedback modulation of the presynaptic nerve terminal. Recent studies on vertebrates have shown that EAAT- and GAT-type neurotransmitter transporters are co-expressed in some glial cells and may even be co-localized in the glial cell surface. Retinal glial (Muller) cells co-express EAATI (GLAST) and GAT-3 (Derouiche and Rauen, 1995; Gadea and Lopez-Colome, 2001b) and convert both GABA and glutamate taken up to glutamine (Reichenbach et al., 1998). In the process these glial cells act as a sink for ammonia, which is neurotoxic (Poitry et al., 2000). GAT-1 is widely expressed by astroglial cells in the vertebrate cerebral cortex (Gadea and Lopez-Colome, 2001b). Many astrocytes in the vertebrate CNS also express EAAT1 or EAAT2 (GLT-1) (Rothstein et al., 1995: Derouiche and Rauen, 1995; Danbolt, 2001). Glial co-expression of glutamate and GABA transporters may be a common feature in astroglial cells. Glial cells in the insect nervous system are extremely diverse in form and function, and tripartite synapses and other types of axoglial interaction are common (Carlson, 1987: Carlson and Saint Marie, 1990). Consequently, the discovery of complex integrative roles for glial cells in the insect CNS can be anticipated. We have already discussed the glutamine cycle, in which glutamate released at glutamatergic synapses may be recycled to the neurons by associated glia as glutamine (section 2.1.6). Some insect glial cells also co-express glutamate and GABA transporters. Glutamate and GABA co-transport has been detected autoradiographically at neuromuscular synapses in locust skeletal muscle (van Marie et al., 1983, 1985), and this is probably common in many insects. The glial cells that cover neuromuscular endplates in skeletal muscle in the cabbage looper, however, express trnEAAT m R N A (Fig. 4a), but not trnGAT mRNA, in keeping with the absence of GABA-ergic motor neurons in the Lepidoptera. Flies also do not appear to express high-affinity GABA transporters outside the CNS (Neckameyer and Cooper, 1998). The situation is different in the caterpillar CNS, however. Both trnEAAT1 and t r n G A T l appear to be co-expressed by five symmetrical pairs of glial cells in each unfused ganglion, as shown in Fig. 6 for thoracic ganglion 2. The cell bodies of the glial cells co-expressing these two NTTs lie at the cortex neuropile border in a position corresponding with the "interface' glial cells described in D r o s o p h i l a (lto el al., 1995; Granderath and Klambt, 1999). Some of the many glutamate-immunoreactive motor neurons found in insect ganglia (Bicker et al., 1988: Watson and Seymour-Laurent, 1993) are known to make glutamatergic synapses with other motor neurons within the neuropile (Burrows, 1996). The fast extensor tibiae motor neuron in the locust

90

S. CAVENEY AND B. C. DONLY

c)

0

.,--= .~ ._~ ._~

[-

- ~ = ' ~ L~ ~

"-'

0

.~. ~ - - . ~

~°~ z~

~-~<

--'~"

~-,h

< ~,-., ~- p.

©

c~

¢"

©

o

NEUROTRANSMITTER TRANSPORTERS

91

metathoracic ganglion, for example, makes direct excitatory contacts with the axons of motor neurons innervating the tibial flexor rnuscles (Parker, 1994). Presumably the interface glial cells, which in the caterpillar T. ni send oligodendrocyte-like projections that fold and surround axons in the neuropile (Fig. 4d), are responsible for maintaining the extra-axonal glutamate concentrations below the synaptic threshold in this region of the ganglion. Although glutamate is also known to bind to inhibitory extrasynaptic receptors on neuronal cell bodies in the cortex of the ganglion (Burrows. 1996), antibodies to trnEAATI failed to stain the surfaces of cortical glial cells (Gardiner et al., 2002). Glial high-affinity uptake of an amino acid neurotransmitter such as glutamate or GABA. either from within the synaptic space or from extrajunctional sites, would lower its extraeellular concentration and increase its net rate of diffusion from the synaptic site of release. Consequently. glial-expressed transporters may play a critical role in ensuring a rapid attenuation in intensity and duration of post-synaptic events triggered by the discharge of neurotransmitter from a presynaptic neuron.

3.2

SEROTONIN TRANSPORTERS

Four neuroactive monoamines, serotonin (5-HT). dopamine (DA), octopamine (OA) and histamine (HA), are authenticated neurotransmitters in the insect nervous system (Brown and Nestler, 1985: Evans, 1985: Monastirioti. 1999). The status of tyramine (TA), the immediate precursor amine in OA synthesis, as a bona fide insect neurotransmitter is uncertain. Nor-epinephrine (NE), a ~hydroxylated DA derivative used widely in vertebrate systems, is present only in trace amounts in insects and is not presently considered to be a neurotransmitter in insects. These biogenic amines are stored in discrete populations of neurons in the insect CNS (N/issel, 1996; Monastirioti. 1999). 5-HT, DA and OA are slow-acting neurotransmitters that bind specific multiple metabotropic receptors (G-protein coupled receptors) on the cell surface of target neurons in the CNS, causing a rise in the concentration of second messengers such as cyclic nucleotides, Ca-'*, inositol trisphosphate (lnsP0 and nitric oxide (NO). These biogenic amines are selectively recovered frorn the synaptic cleft by neuronal high-affinity transporters located in o1" near the nerve terminals (a concept first proposed for insects by Evans (1980)), concentrated in synaptic vesicles for re-use or inactivated by N-acetylation. Considerable amounts of N-acetylated metabolites of 5-HT, DA, OA and tyramine (TA) are present in the honeybee brain (Sasaki and Nagao, 2001 ). 5-HT, DA and OA rnay also act as neurotransmitters or neuromodulators in a variety of peripheral tissues, including salivary glands, Malpighian tubules, heart and other visceral muscles. skeletal and flight muscle, and fat body, such as the firefly light organs. The Fate of these biogenic amines following their release at peripheral targets is unclear, since the presence of Na+-dependent high-affinity amine transporters

92

S. CAVENEY AND B. C. DONLY

at peripheral release sites is less well documented (Wierenga and Hollingworth, 1990; Ali and Orchard, 1996). 5-HT has been implicated in the control of arthropod behaviours as diverse as aggressiveness in the ant Formica ru~l (Kostowski et al., 1975) and the crustaceans Astacus astacus (Huber et al., 1997) and Carcinus maenas (Sneddon et al., 2000), circadian rhythms in Acheta domesticus (reviewed in Brown and Nestler, 1985) and possibly sleep in Drosophila (Shaw et al., 2000). At the cellular level, 5-HT stimulates fluid transport in the salivary glands of Calliphora ervthrocephala (Prince and Berridge, 1972) and Locusta migratoria (Baines et al., 1989; Ali and Orchard, 1996) and Malpighian tubules of Rhodnius prolixus (O'Donnell and Spring, 2000), as well as influences the transepithelial potential of accessory cells in the olfactory sensilla of Manduca sexta (Dolzer et al., 2001). 5-HT also induces the contraction of cardiac and visceral muscle in many insects (reviewed in Brown and Nestler, 1985). Neurons using the indolamine 5-HT as a neurotransmitter arose early during metazoan evolution (Hay-Schmidt, 2000). Relatively few cells containing 5-HT are present in the arthropod CNS (N~issel, 1986; Harzsch and Waloszek, 2000). The nerve cord of the larva of Drosophila melanogaster contains 84, the adult fly 106, 5-HT immunoreactive neurons (Valles and White, 1988). Most 5-HT immunoreactive cells are intraganglionic interneurons in which 5-HT is thought to serve primarily as a central neuromodulator/co-agonist rather than as a rapid-acting neurotransmitter. Serotonergic neurons in the brain branch extensively in the neuropiles of the optic tract, antennal lobe, mushroom bodies or central body of the protocerebrum (Nfissel, 1986; Burrows, 1996), where their cell bodies are usually arranged in bilaterally symmetrical pairs or clusters (N~issel, 1986). In orthopteran insects, the suboesophageal ganglion contains a pair of 5-HT containing efferent (SN2) neurons that innervate the salivary glands (Baines et al., 1989; Ali and Orchard, 1996), The unfused ventral ganglia in arthropods each contain two to four bilaterally symmetrical pairs (or clusters) of 5-HT immunoreactive neurons (Nfissel, 1986; Harzsch and Waloszek, 2000). The stomatogastric nervous system, particularly the frontal ganglion, contains many 5-HT immunoreactive neurons that innervate the foregut (Klemm et al., 1986). 3.2.1

Background

Neuronal uptake of indolamines in the insect CNS was first suggested by Klemm and Schneider (1975), who showed that nerve fibres in the isolated brain of the locust S. gregaria were able to accumulate 6-hydroxytryptamine, a strongly fluorescent analogue of 5-HT, from the bathing medium. Subsequently, high-affinity serotonin uptake was demonstrated in abdominal nerves of Rhodnius prolixus (Flanagan and Berlind, 1984; Orchard, 1989), in cultured embryonic neurons from Periplaneta americana (Bermudez and

NEUROTRANSMITTER TRANSPORTERS

93

Beadle, 1989) and in salivary neuron/gland preparations from Locusta migratoria (Ali and Orchard, 1996). The salivary glands of L. migratoria contain two Na+-dependent uptake systems for serotonin, one of high affinity (Kin = 0.74/,M) and the other of lower affinity (Km = 30#M) (All and Orchard, 1996). Because aminergic neurons are classified in part on the basis of the selectivity of their high-affinity monoamine transport systems (in particular the nature of the monoamine transporter in the plasma membrane), serotonergic neurons are distinguished fi'om neurons releasing other biogenic amines by their selective high-affinity re-uptake of 5-HT. A unique population of serotonergic neurons capable of taking up 5-HT in the larval and adult CNS of Drosophila was shown by (1) the demonstration of selective 5-HT uptake and sequestration by certain neurons in mutant flies incapable of synthesizing 5-HT (in wild type flies these cells are serotonin-immunoreactive; Valles and White, 1986, 1988), and (2) in situ localization of serotonin transporter m R N A in the somata of the same neurons (Demchyshyn et al., 1994). 3.2.2

Structure

The first cDNA cloned encoding a high-affinity, Na+/Cl -dependent monoamine transporter was that of the human nor-epinephrine transporter (hNET) (Pacholczyk et al,, 1991). The observed sequence homology between hNET and the mammalian GABA transporter GAT-1 (Guastella et al., 1990) triggered the simultaneous cloning and characterization of many other cDNAs coding for mammalian monoamine transporters in the early 1990s (reviewed in Amara and Kuhar, 1993: Masson el al., 1999). Among these were the rat dopamine (rDAT) (Kilty el al., 1991: Shimada et al., 1991) and serotonin transporters (rSERT) (Blakely et ell., 1991). These mammalian monoamine transporters share 50% to 70% amino-acid sequence similarity and contain several strongly conserved motifs (Shafqat el al., 1993). The sequence and domain homology in these mammalian transporters has allowed most of their insect counterparts to be cloned by reverse transcriptase polymerase chain reaction (RT-PCR). A partial sequence representing a Drosophihl GLYT-type homologue belonging to the mammalian GABA/monoamine transporter family was the first to be identified in a GenBank search in 1991 (Liu eta/., 1992) (Table 3). Tile Drosophila serotonin transporter drmSERT (Corey et al., 1994b: Demchyshyn et al., 1994) was the first Na+/C1--dependent neurotransmitter transporter cloned from an insect. Its protein structure is a useful reference when comparing the structures of other insect monoamine transporters belonging to the extended Na+/CI--dependent transporter subfamily. Monoamine transporters share many structural features and domains in common with functionally-distant family members, such as GATs, mammalian proline. glycine and taurine transporters, as well as orphan transporters (Lill and Nelson, 1998: Palacin ctal., 1998). The monoamine transporter subfamily of high-affinity neurotransmitter transporters includes SERTs, DATs, NETs, ETs

94

S. CAVENEY A N D B. C. DONLY

and OATs (Fig. 8). Sections 3.2 to 3.5 describe the properties of representatives of this subfamily known to be present in the insect CNS.

3.2.2.1 Functional domains. The N-terminal half of monoamine transporter proteins contains most of the characterized functional domains. For drmSERT and trnSERT these subfamily-specific domains include:

Monoamine-b#uling site(s). A canonical aspartate residue in the pore loop structure in mammalian monoamine transporters is also present in insect SERTs (Corey et al., 1994b; Demchyshyn et al., 1994), as well as DATs (P6rzgen et al., 2001; Gallant et a/., 2002) and OATs (Malutan et al., 2002). This residue is required for monoamine transport (Kitayama el al., 1992). The aspartate residue is replaced by glycine (G54 in trnGAT) in insect GATs and the Drosophila inebriated (h~e) gene product. One or more isoleucine residues in the hydrophobic T M D 3/TMD 3* section of mammalian (as well as in insect) SERTs are thought to form part of the 5-HT permeation pathway (Chen and Rudnick, 2000). These isoleucines are also thought to be involved in monoamine binding (Chen and Rudnick, 2000), as is a tryptophan residue (W249 in rat DAT) in the KVVW l/v sequence found in all cloned SERTs and DATs (Lill and Nelson, 1998). The tryptophan residue is replaced by tyrosine in GATs and the amino acid transporter subfamily. Ion permeathm site. As mentioned for GATs, the WR-X(4)-C motif crucial to Na+/C1 - binding (Mager et al., 1996) is present in all cloned insect monoamine transporters and other members of the extended family (Lill and Nelson, 1998). Heptan leuchw zipper. The leucine zipper repeat is present in T M D 2 in all insect monoamine transporters, such as the fly serotonin transporter drmSERT and the caterpillar octopamine transporter OAT. In trnOAT this motif extends from L143 to L164 (Malutan et al., 2002). Tricyclic antidepressant h~teraction site. A site within T M D 12 in the human serotonin transporter contains a phenylalanine residue (F586) that appears to be associated with the extreme sensitivity of hSERT to imipramine and desipramine inhibition relative to other SERTs (Barker and Blakely, 1996). Rat SERT, which lacks this residue, is almost 10-fold less sensitive to imipramine. Site-directed mutagenesis of F586 lowers hSERT sensitivity to tricyclic antidepressants. Equivalent mutagenesis of hNET at this residue had no comparable effect, suggesting a SERT specific role for F586 in tricyclic drug binding (Barker and Blakely, 1996). With drmSERT, which is relatively insensitive to tricyclic drugs (Table 4), the comparable position in T M D 12 is occupied by a valine residue. The comparable position in T M D 12 of human DAT and NET has a methionine residue, in caeDAT an alanine residue, in

NEUROTRANSMITTER TRANSPORTERS

95

e¢

c z

E

E

d~

c~

e~

z

fa:

~=

E~

_

z

~

.

~

~

©

~

LS~

t<

.......

~

-~

~.~

~2

i

z~

"-

F<

>-

E an

s < <

i

!

96

S. CAVENEY AND B. C. DONLY

drmDAT and trnDAT a valine residue, and in trnOAT an isoleucine residue. Nevertheless, many of these transporters are very sensitive to imipramine and desipramine inhibition (Table 4), leading to the conclusion that the role of an aromatic amino acid in this position for tricyclic drug binding in phenolamine transporters may be small. Cocaine binding site. Cocaine is a natural insecticide that suppresses feeding in M a n d u c a caterpillars and inhibits Na+-dependent monamine uptake in hemisected brains from the cockroach Blaberus (Nathanson et al., 1993). Cocaine has been shown to inhibit two cloned monoamine transporters, drmSERT (Corey et al., 1994b; Demchyshyn et al., 1994) and drmDAT (P6rzgen e t a / . , 2001). Cocaine is discussed further in section 3.4.4, since its pesticidal action is thought to be primarily due to a potentiation of octopaminergic rather than dopaminergic or serotonergic neurotransmission (Nathanson et al., 1993). High-affinity 5-HT transporters likely exist as dimeric, tetrameric or even oligomeric functional structures in cell membranes. This was recently demonstrated elegantly by the co-immunoprecipitation of two mammalian SERT protein mutants bearing different antigenic epitopes after they had been coexpressed in HeLa cells (Kilic and Rudnick, 2000). Oligomerization appears to be a standard feature of proteins belonging to either the Na+/K+-dependent family (Danbolt, 2001) or Na+/Cl--dependent families of neurotransmitter transporters.

3.2.3

Distribution

As mentioned above, the number of 5-HT-immunoreactive neurons in the arthropod nervous system is small and their neurites have a distinctive branching pattern (N/issel, 1986; Harzsch and Waloszek, 2000). In situ localization of drmSERT m R N A in whole-mount preparations of stage 15 and stage 16 Drosophila embryos (Demchyshyn et al., 1994) identified a small number of strongly staining cell bodies with a pattern of distribution virtually identical to that reported for serotonin-immunoreactive neurons in this fly (Valles and White, 1986, 1988). These cells lie in pairs ventrolateral to the neuropile of each neuromere. The suboesophageal ganglion contains three pairs of drmSERT-positive cells, with two pairs in the thoracic and abdominal ganglia, and the third pair in the terminal ganglion (Demchyshyn et al., 1994). A similar pattern of SERT m R N A expression is seen in putative serotonergic neurons in the CNS of the caterpillar T. ni (Fig. 7). The correspondence in the distribution of dnnSERT expression and 5-HT immunoreactivity suggests that drmSERT is a functional 5-HT transporter in the fly CNS (Demchyshyn et al., 1994). There is still no direct immunocytochemical evidence, however, that this monoamine transporter, or for that matter any other monoamine transporter, is localized to the axonal membrane at insect nerve terminals.

NEUROTRANSMITTER TRANSPORTERS

OAT

97

I)AT

SERI

Brain

SOG

I1

..

1-2

T3

A 1

A2-5

TAG

FIG. 7 Cellular localization of OAT, DAT and SERT mRNA in the CNS of the caterpillar Trichoplusia ni as determined by in situ hybridization. The brain, suboesophageal ganglion (SOG), thoracic ganglia (TI T3), non-fused abdominal ganglia (A1, A2 A5) and the fused terminal abdominal ganglion (TAG) contain unique sets of small numbers of neurons expressing these high-affinity monoamine transporters, trnSERT and (to a lesser extent) trnDAT are expressed in a repeating pattern in the thorax and the non-fused ganglia in the abdomen, trnOAT is peculiar in that it is apparently not expressed in the A2-A5 ganglia and in the terminal abdominal ganglion. The positions of neuron cell bodies lying dorsally in the ganglia are indicated in blue, those ventrally in red. The highest concentration of neurons that express monoamine transporter mRNAs are found in the brain and suboesophageal ganglion (A, Malutan et al. (2002): B, Gallant eta/. (2002); C, Tabita Malutan (unpublished observations)).

98

3.2.4

S. CAVENEY AND B. C. DONLY

Kinetics and pharmacology

The selective high affinity of drmSERT for 5-HT is typical of that reported for cloned mammalian SERTs (Table 4). drmSERT activity is more strongly inhibited by mazindol (K i = 27 nM) and nomifensine (Ki = 44 nM) than vertebrate SERTs (Corey el al., 1994a), but to a lesser extent by potent blockers of 5-HT uptake in mammals, such as fluoxetine, desipramine and some other antidepressant drugs (Corey et al., 1994a; Demchyshyn et al., 1994) (Table 4). drmSERT is somewhat more sensitive to inhibition by cocaine (Ki = 184nM; Corey el al., 1994b) than are vertebrate SERTs. The transport stoichiometry of monoamine uptake by mammalian SERT and other mammalian NTTs belonging to the Na+/CI -dependent neurotransmitter transporter family has been described as electroneutral with one monoamine+/Na+/CI - co-transported and one K + countertransported per cycle. Uptake was regarded as voltage-independent and not reliant on a transmembrane K + gradient (reviewed by Rudnick, 1997). The actual uptake mechanism may, however, be more complex than initially thought, as it appears that extra Na + ions may 'escape' into the cell during each transport cycle (Mager et al., 1994). Monoamine uptake may be voltage-dependent (Sonders and Amara, 1996) and coupled to a monoamine-gated ion channel (Galli et al., 1997). It is now appreciated that three types of current are associated with GABA/monoamine neurotransmitter transporters in this family: (1) one associated with a channel in the transporter that opens in the absence of transmitter (leak current); (2) a current associated with transmitter-gated channels in the absence of transport; and (3) a current produced during transmitter translocation (Galli et al., 1997). An ion-dependent 5-HT-induced current in excess of that predicted by simple stoichiometric uptake has been seen alter heterologous expression of cloned SERTs in Xenopus oocytes (drmSERT, Corey et a/., 1994b; rat SERT, Mager el al., 1994) or through analysis of SERT activity in the native state (Hirudo neurons, Bruns et al., 1993). Similar observations have been made on mammalian DATs, NETs and GATs (Beckman and Quick, 1998; Gadea and L o p e z - C o l o m e , 2001b). Since the ligand-induced currents associated with drmSERT are apparently larger than those seen in mammalian SERTs (Galli et al., 1997), the fly transporter has been used to establish the relationship between 5-HT uptake by drmSERT and membrane potential. The 5-HT-induced currents were examined by comparing the flux of radiolabelled 5-HT in individual voltage-clamped frog oocytes after injection with drmSERT RNA (Galli et al., 1997). Both 5-HT uptake and the 5-HT-induced current were Na +- and Cl--dependent and increased at negative potential. It is proposed that the ion channel in drmSERT either opens at low probability after many cycles of electrically neutral 5-HT transport (a model similar to that described for EAAT; Wadiche et al., 1995), or that there is a single event in which 5-HT moves through the same channel as the electric charge (Galli el al., 1997). This current

NEUROTRANSMITTER TRANSPORTERS

99

is ten-fold larger when Li ~ is substituted for Na + in the saline (Petersen and DeFelice, 1998), The role for the integral ion channel in insect SERT may be to serve as a feedback mechanism regulating neuronal activity through a SERT effect on membrane potential, similar to that proposed for mammalian monoamine transporters and EAATs (Beckman and Quick. 1998).

3.2.5

Re vulatioH

The regulation of mammalian SERT activity has been recently reviewed (Blakely and Bauman, 2000). SERT expression in mammals has been shown to be down-regulated by exposure to selective serotonin re-uptake inhibitors (Horschitz et al., 200 l). Although there is no corresponding information available on drmSERT, the protein structures show such close homology (Fig. 8) so that one may assume that similar mechanisms regulate SERT phosphorylation and intracellular trafficking in insects. 3.3

3.3.1

I)OPAMINE TRANSPORTERS

Backgrmmd

The vertebrate neurotransmitters epinephrine and nor-epinephrine are not demonstrable neurotransmitters in insects. Instead. dopamine (DA) is the main, if not sole, physiologically relevant catecholamine neurotransmitter in insect nervous system (Evans, 1980: Osborne, 1996). DA functions as a neurotransmitter or neuromodulator in the brain (supraoesophageal ganglion: cerebral ganglion), ventral nerve cord and in the stomatogastric nervous system (which includes the tYontal ganglion). DA-ergic neurons modulate visceral and skeletal muscle contraction (Brown and Nestler, 1985: Osborne, 1996). Salivary gland secretion in many insects is regulated by the peripheral release of DA fi-om dopaminergic nerve etadings (e.g. Locttsla mi~raloria, Orchard et a/., 1992). Aldehyde-fluorescence histochemical methods have shown that DA occurs in large amounts in the optic tract, protocerebrum and deuterocerebrum regions of the insect brain (reviewed in Klemm. 1976; Evans 1980; Brown and Nestler, 1985). Radioenzymatic methods, which are more sensitive and provide a more accurate estimate of tissue biogenic amine content, indicate that DA is present at high levels (10-30prnol) in the brain and optic lobes of the locust Schistocerca ,~regaria and in the brain of the cockroach Periphmeta americamt (Evans, 1980). Analytical chromatographic methods (primarily reverse-phase HPLC) have confirmed that DA is the most abundant catecholamine in the insect CNS. The significance of trace amounts of nor-epinephrine (NE) in the CNS remains unresolved. The DA content in the CNS of adult bees (Apis mellffera, Harris and Woodring, 1999) and moths (Acheronlia stv.v. Awad et al.. 1997) is high compared to its octopamine (OA) content. The reverse may be true in the larval stages. In Ac/ler<mtia. the amount of DA climbs from 6 pmol/brain in the wandering fifth instar caterpillar to 84 pmol/

100

S. CAVENEY AND B. C. DONLY

hDAT

lET hNET

darDAT~ ' ~ dm~AT ~nOAT tmDAT

drmSERT

hPROT

hTAUT

hCRET

dlmlNE masGAT hGATI

FIG. 8 Dendrogram/bootstrap analysis of insect Na+/Cl--dependent transporters compared with those in other animals. Sequence similarity was measured by aligning amino acid sequences using ClustalX 1.81 (Thompson el al., 1997) and an unrooted tree calculated using the neighbour joining method employed by the program. Confidence wdues for the derived tree were determined by bootstrapping the dataset using 1000 replicates and a generator seed value of 333 (ClustalX 1.81). Alignments were output in Phylip format to Treeview (1.6.5) for display. Members of each of the 4 subfamilies of biogenic amine, GABA, amino acid, and 'orphan" transporters are grouped using grey background. The accession numbers of the aligned amino acid sequences are: nematode dopamine transporter (caeDAT; Q03614), caterpillar GABA transporters (masGAT; $65673 and trnGAT; AAF70819), octopamine transporter (trnOAT; AF388173), and dopamine transporter (trnDAT; unpublished), fly dopamine transporter (drmDAT; AAF76882), serotonin transporter (drmSERT; AADI0615), and orphan transporters (drmBLOT; CAB53640 and drmlNE; AAC47292), fish dopamine transporter (darDAT; AAK52449), frog epinephrine transporter (lET; AAB67676), and human dopamine transporter (hDAT; AAAI9560), serotonin transporter (hSERT; AAA35492), nor-epinephrine transporter (hNET; P23975), GABA transporter (hGATI; CAA38484), creatine transporter (hCRET; AAC41688), taurine transporter (hTAUT: CAA79481), proline transporter (hPROT; AAB47007), and glycine transporter (hGLYTIc: AAB30785).

NEUROTRANSMI'I-I'ER TRANSPORTERS

101

brain in the adult. The amount of OA drops from 40 pmol/brain to 3 pmol/ brain over the same period (Awad et al., 1997). This approximately 14-fold shift in relative amounts of these two amines may refect the relative importance of DA-ergic and OA-ergic neural signalling in the optic tracts and/or brain of the adult compared to that of caterpillars (Linnet al., 1994). These differences may also reflect a shift in the amount of dopamine transport activity in these two regions during the metamorphosis ot" the adult CNS. Dopamine levels are also low in the CNS of the maggot of C h y m o m y z a costata (Kostal et al., 1998). DA-ergic neurons arborize widely in all major ganglionic neuropiles in the insect CNS (Osborne, 1996). Immunocytochemistry with DA antibodies showed that the orthopteran brain and ventral nerve cord (Locusta migratoria, Viellemaringe et al., 1984: Grvllus bhnaculatus, Homer et al., 1995) contains several hundred neurons with DA-like immunoreactivity. Neurons immunoreactive to tyrosine hydroxylase, a marker enzyme for DA-ergic neurons, are present in the optic tract and in the protocerebral neuropile surrounding the central and mushroom bodies of P. americana (Granholm et al., 1995). Dopaminergic terminals are also present in visceral muscle of S. ,gre~,aria (Klemm, 1972). In Drosophila, about 200 catecholamine-containing cell bodies were detected in the brain (excluding optic lobes) by glyoxylic acid-induced histofluorescence (Budnik and White, 1988). DA-containing neurons in the developing nervous system of Drosophila also have been mapped immunocytochemically (Budnik and White, 1988; Lundell and Hirsch, 1994a; Monastirioti, 1999), using antibodies against two specific marker enzymes. These neurons express the enzymes dopa decarboxylase (Beall and Hirsch, 1987; Konrad and Marsh, 1987) and tyrosine hydroxylase (kundell and Hirsch, 1994b) that are specifically involved in DA synthesis. The distribution pattern of DA-ergic neurons in dipteran and orthopteran CNS appears to be highly conserved. In the fly, each larval hemi-protocerebrum contains three clusters, and the adult protocerebrum four clusters, of between five and seven prominent DA-ergic neurons, and the suboesophageal ganglion a single medial cluster (Budnik and White, 1988). The thoracic and non-terminal abdominal ganglia in the larva contain approximately 30 DA-ergic neurons. Each ganglion has a single medial DA-ergic neuron and a lateral pair of DA-ergic neurons. Dopaminergic neurons are thought to be involved in the processing of olfactory input in the mushroom body, since dopamine Dl-like receptors are specifically expressed in MB axons (Roman and Davis, 2001). No reports of high-affinity Na+-dependent doparaine uptake in insect nervous tissue apparently exist before the 1990s, although Budnik et al. (1986) reported the accunqulation of-~[H]dopamine by specific neurons in the CNS of a Drosophila mutant deficient in dopa decarboxylase. Scavone et al. (1994) first detected a cocaine-sensitive Na+-dependent dopamine uptake system in synaptosomes prepared from the CNS of the cockroach Blaherus. Although dopamine had previously been shown to competitively inhibit

102

S. CAVENEY AND B. C. DONLY

[3HJoctopamine uptake in isolated nerve cords of Periplaneta americana (Evans, 1978), this could not be taken as direct evidence for dopamine uptake. The locust salivary nerve/gland complex has a low-affinity uptake system for DA that is Na+-independent (Ali and Orchard, 1996). Whether this represents neuronal or epithelial transport has not been established.

3.3.2

Structure

The first mammalian Na+/C1 -dependent, cocaine-sensitive, dopamine transporters were cloned and characterized a decade ago (Giros et al., 1991, 1992). Catecholamine transporters have been recently cloned from two insects, namely D. melanogasler (drmDAT; P6rzgen et al,, 2001) and the cabbage looper caterpillar Trichoplusia ni (trnDAT: Gallant et al., 2002). The only other known example of an invertebrate DAT is caeDAT from the nematode C. elegans (Jayanthi et a/., 1998). The monoamine transporter subfamily in the bootstrap tree in Fig. 8 shows the DATs cluster in two separate groups, with the invertebrate members separate from the vertebrate examples. Interestingly, the vertebrate DATs group more closely with the vertebrate NETs (and frog ET). On the basis of this and other functional observations, P6rzgen et al. (2001) have suggested vertebrate DATs and NETs may be the consequence of the duplication of a primordial catecholamine transporter represented by the invertebrate DATs. However, the recent cloning of another novel member of this subfamily from T. hi, which was found to have highest affinity for octopamine and tyramine (trnOAT), suggests that in at least some invertebrates there may also have been a unique derivation of such a primordial transporter to more specialized functions. This insect OA transporter is discussed in section 3.4. 3.3.2.1 Functional domains. Insect DATs contain the monoamine-binding site(s), the ion-permeation motif in the pore-loop, and the heptan leucine zipper in T M D 2 as described in section 3.2.2. A second zipper motif is seen in T M D 5 of drmDAT (P6rzgen eta/., 2001). Other structural features of insect DATs include the following: Phosphoo'h~tion sites. The N-terminus of the two cloned insect DATs drmDAT and trnDAT contains one to two consensus sequences for phosphorylation by protein kinase A, and one to two more on the cytoplasmic C-terminal tail. drmDAT has another in the cytoplasmic loop between TMDs 4 and 5. Protein kinase C phosphorylation sites are present on the Nterminus, on the cytoplasmic loop between TMDs i0 and 11 and there are one to two at the C-terminus (P6rzgen et a/.~ 2001: Gallant el al., 2002). A casein kinase II site is also located near the C-terminus of drmDAT (P6rzgen et a/., 2001 ).

NEUROTRANSMITTER TRANSPORTERS

103

Glycosylation sites. Several (two to six) N-linked glycosylation sites are present on the second extracellular loop (EL2) of insect, nematode and mammalian DATs. An additional one occurs on the third extracellular loop of d r m D A T (P6rzgen et ell., 2001), suggesting that insect DATs, as demonstrated for their mammalian homologues, are glycoproteins. Other mot(fis. Two cysteines eight residues apart in the second extracellular loop (EL2), crucial for correct insertion of other monoamine transporters into the plasma membrane (Wang el al., 1995), are conserved in insect DATs. DATs are also characterized by the presence of two serines in T M D 7 (Kitayama et al., 1992). This DAT motif is thought to be involved in binding the catecholamine ring of its normal substrate and possibly the phenolamine ring of OA and TA, two competitive transport substrates.

3.3.3

Distribution

DAT is expressed by a relatively small number of neurons in the brain and ventral nerve cord of Drosophila (P6rzgen el al., 2001) and Trichoplusia ~i (Gallant el al., 2002). The expression of d r m D A T in the third-instar fly larva occurs in three small groups of dopaminergic neurons in each brain lobe and in unpaired medial neurons and paired dorso-lateral neurons in the ganglia of the ventral nerve cord (P6rzgen et al., 2001). This pattern of expression is identical to that of a subset of dopa decarboxylase (drmDDC)-positive (Konrad and Marsh, 1987) and catecholamine-immunoreactive (Budnik and White, 1988) cells that represent the dopaminergic neurons in the fly CNS. Dopamine had earlier been shown to be present in the same three clusters of neurons in the brain lobes and in about 30 different neurons in the ventral ganglion (Budnik el a/., 1986). (A second set of drmDDC-positive neurons (Konrad and Marsh, 1987) are serotonin-immunoreactive (Valles and White, 1988) neurons that express drmSERT (P6rzgen el ell., 2001) in the fly. drmDDC is a synonym of AADC, or aromatic amino acid decarboxylase, an enzyme that decarboxylates both DOPA and tryptophan). DAT expression in the CNS of the cabbage looper caterpillar follows a pattern consistent with that of the distribution of DA-immunoreactive neurons in other insects. The caterpillar hemi-protocerebrum contains three prominent clusters of trnDAT mRNA-positive neurons, the deutocerebrurn two (one medial, one lateral), the tritocerebrum one medial cluster of neurons, and each segmental ganglion three to seven DAT-positive cell bodies (Fig. 7). 3.3.4

Kinetics aud pharmacolog.v

Vertebrate dopamine transporters have a high and selective affinity for their primary substrate. The Km values for DA are in the sub-micromolar to low micromolar range (Table 4). DATs generally have a much lower affinity for

104

S. CAVENEY AND B. C. DONLY

biogenic monoamines (NE, epinephrine, OA), that are structurally related to their primary uptake substrate (Pacholcyzk et al., 1991: Giros et al., 1992; Apparsundaram et al., 1997). Fly and worm DATs have a low affinity for octopamine, serotonin and histamine, as indicated by competitive inhibition of DA uptake (Jayanthi et al., 1998; P6rzgen et al., 2001) (Table 4). Insect DATs do not appear to serve double-duty as high-affinity octopamine transporters (P6rzgen et al., 2001: Gallant et al., 2002). While the affinity of drmDAT for m-tyramine is five-fold less than that for DA, its affinity for OA is about two-orders of magnitude less, even lower than its affinity for 5HT. Thus drmDAT seems unlikely to have a further role serving as a highaffinity octopamine transporter in addition to functioning as a DAT (P6rzgen et al., 2001). There are, however, no uncharacterized DAT-like sequences that might represent a Drosophila OAT present in FlyBase. Thus it has been proposed by P6rzgen et al. (2001) that OA may be cleared from the synaptic space in Drosophilu by an unrelated transporter or possibly enzymatically inactivated in the extracellular space. This is surprising, since OAT, the octopamine/tyramine transporter cloned from the caterpillar brain (see section 3.4), has a high affinity for [3H]dopamine (Kn, = 3.0 #M) as well as octopamine (Km = 2.1 #M) (Malutan et al., 2002). Monoamine transport by insect DATs requires Na + and is facilitated by anions, particularly CI . Substitution of the cations K +, Li+, choline, or Nmethylglucamine for Na * in the external saline reduces DA uptake by >95% (drmDAT, P6rzgen et a/., 2001; trnDAT, Gallant et al., 2002). Replacing saline C1- with gluconate (P6rzgen et al., 2001) or with Br-, I-, nitrate, phosphate and sulphate reduces DA uptake by insect DATs, although the phosphate anion supports transport well (Gallant et al., 2002). The pharmacological profiles of invertebrate dopamine transporters are distinct from their mammalian counterparts. Nisoxetine, a selective blocker of mammalian NETs, is a potent blocker of dopamine uptake by invertebrate DATs. Conversely, nisoxetine is a relatively weak blocker of dopamine uptake by human DAT. GBRI2909, a potent blocker of dopamine uptake by mammalian DATs is a weak blocker of invertebrate DATs. Tricyclic antidepressants such as desipramine and imipramine are relatively weak blockers of mammalian DATs but strong blockers of invertebrate DATs and mammalian NETs (Table 4). All mammalian Na+/Cl--dependent monoamine transporters are inhibited to some extent by the alkaloid cocaine. Cloned rat DAT is less sensitive to cocaine inhibition than the human NET (P6rzgen et al., 2001). Sodium-ion-dependent DA uptake by synaptosomes isolated from the nervous system of the cockroach Blaberus is less sensitive to inhibition by this non-selective blocker of monoamine uptake (ICs0 c.100#M) than synaptosomal Na+-dependent OA uptake (ICs0 c.40#M) (Scavone et al., 1994). The cocaine concentration required to reduce DA uptake by 50% by two cloned invertebrate DATs (DrosophiM, IC50 = 2.9#M; C. elegans, 1C5o approximately 5/xM) is at least one order of magnitude higher than that required to

NEUROTRANSMITTER TRANSPORTERS

105

inhibit cloned SERTs and NETs (P6rzgen et al., 2001). Although cocaine disrupts high-affinity dopamine and serotonin transport in Drosophila (Corey et al., 1994b; P6rzgen et al., 2001), its main target in insects is more likely to be the neuronal octopamine uptake system (Nathanson el al., 1993; Scavone et a/., 1994). Mazindol is a potent but relatively non-selective inhibitor of many cloned monoamine transporters, such as drmDAT, caeDAT, drmSERT, and particularly human NET (Tatsumi et al., 1997; P6rzgen el al., 2001). The inhibitory profiles of drmDAT (its high affinity for the tricyclic antidepressants desipramine and imipramine, and moderate sensitivity to cocaine) and trnDAT suggest that insect DATs are pharmacologically more closely related to mammalian NETs than to mammalian DATs or even drmSERT (Table 4). An extended series of selective and non-competitive blockers of mammalian DATs and NETs, tested for their ability to inhibit DA uptake by drmDAT (P6rzgen et al., 2001), trnDAT and trnOAT (Gallant el al., 2002) supports this idea. Desipramine, imipramine and nisoxetine, which are relatively selective inhibitors of NET and ET in vertebrates, inhibit both caterpillar DAT and OAT. Xylamine, which has been reported to block OA uptake in the cockroach CNS (Wierenga and Hollingworth, 1990), was an ineffective blocker of caterpillar DAT and OAT. drmDAT has been suggested to be a 'primordial catecholamine transporter', and the drmDAT and caeDAT genes (Jayanthi et al., 1998) may be representatives of an ancestral gene whose descendants encode vertebrate catecholamine transporters such as DAT, NET and ET (P6rzgen et al.. 2001). However, the caterpillar CNS possesses a second Na+-dependent phenolamine transporter, the octopamine transporter OAT. Apparently absent from flies and nematodes, OAT has a high transport affinity for both DA and OA (section 3.4.4). Consequently it may be premature to speculate on the ancestral catecholamine transporter before the molecular details of monoamine transporters in less advanced insect orders and other animal phyla are known. Also, pharmacological studies on intact cockroach neural tissue hint that DA and OA are transported by two pharmacologically distinct Na+-dependent systems (Wierenga and Hollingworth, 1990). 3.3.5 Regulation DAT regulation in the insect CNS has not been investigated. In contrast, the regulation of human striatal DAT has been extensively studied because of its central role in cocaine addiction and amphetamine abuse. Dopamine transport by human DAT is stimulated by dopamine binding to presynaptic dopamine D2 receptors and inhibited by D2R antagonists. The pertussis toxin-sensitive Gio signalling cascade involved in transporter up-regulation induces a recruitment of DAT protein into the presynaptic membrane of DA-ergic neurons (for references, see Mayfield and Zahniser, 2000). Presumably insect DATs, which have multiple intracellular consensus sites for protein phosphorylation,

106

S. CAVENEY AND B. C. DONLY

are similarly regulated through G-protein-coupled kinase and phosphatase second-messenger pathways. 3.4

OCTOPAMINE TRANSPORTERS

3.4.1 B a c k g r o u n d OA, the monohydroxylic analogue of NE, functions as a neurohormone, neuromodulator and neurotransmitter in insects (Evans, 1980, 1985; Orchard, 1982; Roeder, 1994; Roeder, 1999). Octopaminergic pathways are involved in the modulation of physiological processes as diverse as olfactory conditioning in the antennal lobe and mushroom bodies (Menzel and Muller, 1996; Roman and Davis, 2001), neuromuscular synaptic transmission (Evans, 1985; Orchard and Lange, 1985: Pflfiger el al., 1993) and nitric oxide release during flashing in fireflies (Trimmer et al., 2001). OA is found in all the ganglia in the CNS of the caterpillars and adult moths of M a n d u c a sexta and Spodoptera littoralis, and is present in particularly large amounts in the protocerebrum and suboesophageal ganglion (Davenport and Wright, 1986). OA is the dominant monoamine in the CNS of the caterpillar Acherontia styx, whereas DA predominates in the CNS of the adult moth (Awad et al., 1997). Populations ofoctopaminergic neurons have been identified by OA-immunoreactivity in the CNS of orthopterans (Stevenson and Sparhase-Eichmann, 1995), moths (Pflfiger et al., 1993), flies (Monastirioti et al., 1995) and bees (Kreissl et al., 1994; Menzel and Muller, 1996). Two prominent segmentallyrepeating groups of OA-ergic ceils are the so-called dorsal unpaired median (DUM) and ventral unpaired median (VUM) neurosecretory cells (Pflfiger et al., 1993). The brain in larval Drosophila lacks OA-immunoreactive neurons, although they are present in the adult brain (Monastirioti et al., 1995). Most VUM cells in the honeybee suboesophageal ganglion are OA-ergic (Kreissl el al., 1994). OA acting as a neurotransmitter is quickly removed from the synaptic space by selective plasma-membrane transporters after the initiation of the postsynaptic potential (Evans, 1978). Specific Na+-dependent OA uptake has been demonstrated in orthopteroid insects in intact nervous tissue (Evans, 1978; Scott et al., 1985; Roeder and Gewecke, 1989; Wierenga and Hollingworth, 1990), synaptosomes (Scavone el a/., 1994) and peripheral tissues such as fat body and ovary (Carlson and Evans, 1986; Wierenga and Hollingworth, 1990). Following its re-uptake by OA-ergic neurons, OA is presumably returned to synaptic vesicles for re-use or inactivated by neuronal N-acetyltransferases (Evans et al., 1980; Brodbeck et al., 1998: Sasaki et al., 2001). OA uptake has not been demonstrated in the fly CNS. Tyramine is also present in many OA-containing tissues and this has been attributed to TA being the immediate precursor in OA synthesis (Linn el al., 1994). Early studies suggested, however, that TA might also function independently as a

NEUROTRANSMITTER TRANSPORTERS

107

neurotransmitter or neuromodulator (Huddart and OIdfield, 1982; Uzzan and Dudai, 1982; Downer et al., 1993) (see section 3.4.6). Mutant female Drosophila that cannot synthesize OA are phenotypically normal except that they are unable to lay eggs (Monastirioti et al., 1996), suggesting that TA may substitute for OA as a neuromodulator.

3.4.2

Structure

A cDNA encoding a high-affinity octopamine/tyramine transporter (trnOAT) has been isolated from the CNS of the caterpillar, Triehoplusia ni (Malutan et al., 2002). This cDNA is distinct from the insect dopamine transporters discussed in section 3.3. Comparison of the amino acid sequence of trnOAT with the other members of the monoamine subfamily shows that it clusters with no other group, emerging from the main line near the same point as the invertebrate DATs (Fig. 8). Thus it appears that some invertebrates may have acquired specialized catecholamine transporting proteins. However, in the fly there are no uncharacterized DAT-like sequences in FlyBase (revision 2) that might represent a Drosophila OAT homologue. Thus it was proposed by P6rzgen et al. (2001) that in Drosophila, OA might be cleared from the synaptic space by an unrelated transporter or possibly even metabolized in the extracellular space. However, the prevalence of such a mechanism as opposed to the occurrence of specialized transporters for OA and TA, such as trnOAT, remains to be determined. An OAT homologue also appears to be absent from the nematode C. elegans (Jayanthi et al,, 1998). 3.4.2.1 Functional domains. As is characteristic of other monoamine transporters, trnOAT contains a heptan zipper motif in T M D 2, a pair each of cysteine residues and N-linked glycosylation sites in EL 2, and several potential phosphorylation sites at its cytoplasmic N-terminal, internal cytoplasm-facing, and cytoplasmic C-terminal regions. Substrate binding site(s). T M D 1 of trnOAT also contains the canonical aspartate residue required for substrate transport (Kitayama et al., 1992; P6rzgen et al., 2001). T M D 7 of trnOAT contains the two serines found in a sxXS/T motif (Fig. 5) that is thought to be involved in binding the hydroxyl group of the substrate's ring structure in other catecholamine transporters (Kitayama et al., 1992). This motif is shared with masKAAT, a K+-dependent neutral amino acid transporter cloned from tobacco hornworm midgut tissue (Castagna et al., 1998). However, masKAAT's ability to transport tyrosine (the precursor for TA and OA) was not reported. The second serine is missing from many non-catecholamine transporters in the family, such as the GABA transporters and some SERTs.

108

S. CAVENEY AND B. C. DONLY

Et'/'r ~,"r-R moN/i In insect monoamine transporters such as trnOAT (Malutan et al., 2002), drmSERT (Corey el al., 1994b, Demchyshyn et al., 1994), trnDAT (Gallant et al., 2002) and drmDAT (P6rzgen et al., 2001), a phenylalanine residue (F263, F224, F284 and F216 respectively) replaces the tryptophan residue seen in the EFW-R sequence between TMDs 3 and 4 in insect GATs. The function of this substitution is unclear, since masKAAT also lacks this tryptophan residue (Castagna et al., 1998).

3.4.3

Dis'tribution

Northern blot and RT-PCR analysis confirm that trnOAT transcripts are present in caterpillar CNS (both in brain and segmental ganglia), in adult brain and optic lobe, and also in some other tissues (Malutan et al., 2002). The cell bodies of neurons expressing trnOAT were localized by whole-mount in situ hybridization staining of the isolated intact CNS (Malutan et al, 2002). In the caterpillar brain, about 16 sets of paired neurons and a few unpaired neurons express trnOAT, while in the suboesophageal and thoracic ganglia many unpaired cells express trnOAT (Fig. 7). trnOAT was expressed by relatively few neurons in the abdominal ganglia of the T. ni caterpillar. The cellular expression pattern of OAT bears little resemblance to the pattern seen after staining the thoracic and abdominal ganglia of larval Spodoptera littoralis with neutral red, a non-specific marker of aminergic neurons (Davenport and Wright, 1986). Cells corresponding in position to the OA-ergic ventral unpaired median (VUM) motor neurons in the thoracic and abdominal ganglia of Antheraea p e m y i (Brookes and Weevers, 1988) and M. se.vta (PflOger et al., 1993) caterpillars do not stain positively for OAT, although they react to a cRNA probe for tyramine /~-hydroxylase, a marker enzyme for OA-ergic neurons (Malutan et al., 2002). It appears that OAT expression in caterpillars is restricted to neurons (interneurons?) within the central nervous system and is not expressed peripherally by VUM-type motor neurons. This situation may be peculiar to lepidopteran insects, since locust leg muscle (Roeder and Gewecke, 1989) and firefly light organ (Carlson and Evans, 1986) display high-affinity Na+-dependent OA uptake (the firefly light organ activity is sensitive to the tricyclic antidepressant imipramine). This suggests that a monoamine transporter belonging to the Na+/C1--dependent transporter family, possibly OAT, is expressed peripherally in some insects. 3.4.4

Kineticx and pharmacology

The affinity of trnOAT for its primary and alternate monoamine substrates is listed in Table 4. OA uptake by trnOAT has an absolute requirement for external Na + and, as seen for trnDAT, the relationship between the rate of monoamine uptake by trnOAT and external [Na +] is linear up to as high as 150mM Na +. Since the affinity of Na+-dependent transporters for Na + is

NEUROTRANSMITTER TRANSPORTERS

109

normally matched to its ambient in situ concentration, the concentration of free Na" in the extra-axonal fluid in the caterpillar CNS may be as high as 200 mM, similar to that present in the bee CNS (Cardinaud et al., 1994). This is far greater than the Na ÷ concentration in caterpillar haemoplasm (typically <<20raM), which may explain why N a - / K + ATPase (Emery et al., 1998), and the Na+-dependent transporters it energizes do not exist outside the nervous system in Lepidoptera. trnOAT has a relaxed ionic dependency compared to other insect monoamine transporters such as trnDAT and drmDAT. OAT is able to function at near-maximum capacity after C1- in the bathing saline is replaced by several inorganic (notably phosphate and nitrate) and organic anions (Malutan et al., 2002). A similar weak selectivity in anion dependence was noted earlier for drmSERT (Demchyshyn e t a / . , 1994). The significance of this may become clear when the anionic composition of the extra-axonal fluid in the insect CNS is better understood. Perhaps insect members of this monoamine transporter family could be more accurately described as Na+/anion than as Na*/C1--dependent. Although the molecular structure of OA is very similar to that of NE (lacking only one of two hydroxyl groups in the aromatic ring), the pharmacological properties of caterpillar OAT and mammalian NETs are quite different (Table 4). Nisoxetine, which is a strong and selective blocker of mammalian NETs and insect DATs, is a weak trnOAT antagonist. Likewise, nomifensine, desipramine and imipramine are strong blockers of insect DATs but weak blockers of OAT (Table 4). Cocaine, a broad-spectrum inhibitor of Na+-dependent monoamine uptake in insects, is reported to block OA uptake by semi-intact brains of the cockroach Blaberus (Nathanson et al., 1993), 5-HT uptake by drmSERT (Corey el al., 1994b; Demchyshyn el a/., 1994; Barker and Blakely, 1996) and DA uptake by d r m D A T (P6rzgen et al., 2001). Cocaine acts as a natural insecticide that suppresses feeding in M a m t u c a caterpillars (Nathanson et a/., 1993) and induces stereotypic behavioural responses and behavioural sensitization in Drosophiht (McClung and Hirsch, 1999). The pesticidal effects of cocaine are considered to be duc primarily to the potentiation of octopaminergic neurotransmission (Nathanson et al., 1993). This is supported by the observation that cocaine is a less effective blocker of DA uptake in the invertebrate CNS than in the mammalian CNS (Jayanthi et al., 1998: P6rzgen el al., 2001). Nevertheless, because of its relatively non-selective inhibition of monoamine transport systems, it is premature to conclude that the ability of cocaine to suppress appetite in insects is due solely to its effect on OA-ergic neurotransmission. 3.4.5

Regulation

The regulation of OAT transport in the insect CNS has not been studied. The synthesis of OA in the M a m h w a s e x t a CNS is developmentally upregulated at metamorphosis by ecdysteroids (Lehman et al., 2000a,b), suggesting that a

110

S. CAVENEY AND B. C. DONLY

co-ordinate increase in other molecular elements involved in OA-ergic neurotransmission, such as OAT expression, may occur. 3.4.6 Tvramine transport Tyramine, the biosynthetic precursor of octopamine, has been detected in the CNS of Locusta migratoria (Downer el al., 1993), Drosophila melanogaster (McClung and Hirsch~ 1999) and other insects (Osborne, 1996). The insect CNS possesses tyramine receptors thought to be associated with distinct tyraminergic systems (Roeder, 1994; Kutsukake et al., 2000). Several G-protein coupled tyramine-selective receptors have been cloned from insect brain tissue. Drosophila brain possibly expresses two tyramine receptors (octyR99AB, Saudou et al., 1990; Arakawa et al., 1990; misexpressed in the mutant hono, Kutsukake et al., 2000), while honeybee (amTyrl; Blenau et al., 2000) and locust (Vanden Broeck et al., 1995) may each express at least one tyramine receptor. P-element insertional mutation of the gene encoding the G-proteincoupled tyramine receptor octyR99AB in the fly caused a reduction in olfactory motor response, hinting that TA acts as a neurotransmitter and/or neuromodulator in olfactory signalling pathways in the insect brain (Kutsukake et al., 2000), a role shared with OA (Menzel and Muller, 1996). Although tyramine does not stimulate a locomotor response in Drosophila when administered alone, it appears to be involved in behavioural sensitization to cocaine during repeated locomotor responses to this drug (McClung and Hirsch, 1999). As mentioned earlier, cocaine blocks the selective high-affinity uptake in insects of the monoamines dopamine, octopamine and serotonin, it has been suggested that TA could be acting on presynaptic transporters or post-synaptic receptors associated with dopaminergic signalling to modulate cocaine-induced motor behaviours (P6rzgen et al., 2001). The responsiveness of octyR99AB to OA and TA depends on the cell line expressing it in ~,itro, implying that the G-protein complement of neurons expressing this receptor may influence its properties in the CNS (Reale et al., 1997). Even were tyramine to be an authentic neurotransmitter in the insect CNS, expression of a pharmacologically distinct tyramine transporter may not be needed. TA inhibits the in vitro uptake of DA and OA by trnOAT, and DA uptake by trnDAT (Table 4), suggesting that either one of these other monoamine transporters could be recruited to transport TA in putative TA-ergic neurons. As shown in Table 4, the caterpillar OAT has a higher affinity for tyramine (/¢i = 0.4#M, Malutan et al., 2002) than its primary transport substrate, while the fly DAT has a moderate affinity for tyramine ( K i --~ 23 #M, P6rzgen et al., 2001). Perhaps a few general comments on monoamine transporter nomenclature would be useful at this point. In contrast to the trivial names given to many neurotransmitter receptors (NMDA-type, AMPA-type glutamate receptors, for instance), the naming of neurotransmitter transporters is based more on

NEUROTRANSMITTER TRANSPORTERS

111

physiological context rather than on their affinity and/or selectivity for natural or synthetic ligands. The particular neurotransmitter phenolamine(s) taken up and recycled by a monoaminergic neuron depends mainly on the nature of monoamine(s) released by the neuron and by others in its immediate vicinity. For example, the frog epinephrine transporter (lET) has higher affinity for other catecholamines than epinephrine, but, because of its location in muscle, it is thought to function primarily in the re-uptake of the local neurotransmitter epinephrine (Apparsundaram et al., 1997). The cabbage looper octopamine transporter trnOAT may allow OA-ergic neurons to take up tyramine and possibly even dopamine (Table 4), in addition to its named transport substrate. The fly dopamine transporter drmDAT likely takes up tyramine, but not octopamine. The possibility that aminergic neurons use multisubstrate transporters to recover amines fl'om the synaptic space may pose a problem in amine recycling, however. Since aminergic neurons are thought to release only one type of monoamine, what is the fate of the other monoamines taken up by anainergic neurons'? Are they inactivated by selective N-methylor N-acetyltransferases or are they degraded by the monoamine oxidase pathway'?

3.5

3.5.1

ORPHAN TRANSPORTERS

Background

Many additional members of the Na+/Cl--dependent neurotransmitter transporter family have been cloned by homology-based PCR. Despite considerable effort, the organic substrates of these so-called "orphan' transporters have not been identified (Liu el ell., 1993a,b: E1 Mestikawy el al., 1994) and hence it has not been possible to assign a specific transport function to them (Shafqat et al., 1993: Masson et al., 1999). Some orphan transporters, for instance, rat NTT4/ XTI (Liu el ell., 1993b: E1 Mestikawy el al., 1994) and rat V-7-3-2 (Uhl and Hartig, 1992) are expressed in the brain. Rat NTT4/XT1 is expressed in glutamatergic and GABA-ergic regions in the cerebellum and hippocampus. Other orphan transporters, such as ROS1T and rB21, are expressed in epithelial tissues (Masson el ell., 1999). Two insect orphan transporters in this family are the product of the hlol and ira' genes, so named because functionally deficient flies exhibit Malpighian tubule cells that are distended with an enlarged apical surface (Bloated tubules Blot) or increased excitability of the larval motor neuron manifested by enhancement of the phenotypes of mutations in the Shaker (Sh) and Hyperkinetic (H/,) genes (Inebriated - lne). Although many orphan transporters may ultimately be found to have no transport activity, it is possible that the improper expression of their cDNAs ill vitro may account for the current difficulty in assigning them transport functions.

112

3.5.2

S. CAVENEY AND B. C. DONLY

Structure

Orphan transporters exhibit the classical features of the Na+/Cl--dependent neurotransmitter transporter family, as well as a number of other characteristic structural features unique to this subfamily. Rat NTT4/XT1, for example, exhibits a large extracellular loop lying between TMDs 7 and 8 and containing a potential N-glycosylation site. It also has an enlarged loop between TMDs 11 and 12, and an extended C-terminus. These additional elements, combined with the basic structure result in an increase in size for this protein to 727 amino acids. Expressed rat NTT4/XT1 protein did not transport any of the known transmitter amino acids or monoamines, or any common sugars or nucleotides tested when expressed in several cell lines (El Mestikawy et al., 1994) or in Xenopus oocytes (Liu et al., 1993b). By comparison, all the classical members of the family proven to transport substrate lack the enlarged fourth extracellular domain, and have lengths of between 520 and 670 amino acids. The first putative neuronal Na+/C1--dependent transporter cDNA in Drosophila was identified in 1991 by comparing the sequence of a conserved region from the rat GABA transporter with the GenBank database (Liu et al., 1992). The sequence identified comprises the 5' third of CG7075, a protein related to the mammalian glycine transporter rGLYT2. The CG7075 transcript was shown to be expressed in cells in the developing brain of 8h fly embryos (Liu et al., 1992). Analysis of the full FlyBase genome sequence data reveals that the Drosophila genome contains a total of approximately 16 genes showing sequence similarity to the Na+/CI -dependent transporter family (Table 3). These include the characterized family members SERT, DAT and GAT, but also many others, some of which appear to be orphan-like transporters. The sizes of the various predicted proteins vary from 524 to 1308 amino acids (Table 3). Several of these form a more tightly knit group that is similar to the K+-dependent amino acid transporter characterized in M. sexta (Castagna et al., 1998). Two others, the genes Bloated tubuh)s (Blot) and hTebriated (hm), were first identified from genetic analysis of fly mutants. Blot encodes a protein closest in structure to the mammalian orphans rRB21A and rNTT7. The Drosophila Ine transporter (CG15444) and the RosA gene product, its alternatively spliced homologue (Table 3) are most closely related to rat G A T (41% identity) and distantly related to drmSERT (Soehnge et al., 1996). drmlne is 658 amino acids long and possesses the 12 canonical TMDs characteristic of this family, but lacks the large extracellular Mop between TMDs 7 and 8 that is seen in the Blot protein. As mentioned earlier (section 3.1.2), it lacks key motifs typical of GATs. When expressed in Xenopus oocytes, the hornworm Inebriated homologue, maslne, responds to hyperosmotic stress by inducing a Ca 2+activated CI- current through stimulation of protein kinase C and inositol trisphosphate pathways (Chiu et al., 2000). masIne is thought to form a glycosylated and oligomerized plasma membrane complex of mass 170200 kDa (Chiu et al., 2000).

NEUROTRANSMIFIER TRANSPORTERS

113

Several partial or complete sequences of orphan transporters belonging to the GABA/monoamine transporter family have also been reported from the tobacco hornworm Mamtuca sexta by S. S. Gill and his colleagues at UC Riverside. A 57 kDa peptide bearing some resemblance to the glycine/proline transporter (GLYT/PROT) group of Na+/Cl--dependent transporters has been cloned from the hornworm CNS (Sandhu el al., 1998). Another 600700bp amplicon called clone EH-44 was 58% identical to the rat brain serotonin transporter, but is expressed in the midgut, Malpighian tubule and rectum (Pullikuth e¢ al., 1993). A partial sequence of the homologue of Drosophila orphan transporter sequence CG 10804 has been cloned from head cDNA of Trichoplusia ni (C. Donly, unpublished data). It is similar to mammalian orphan NTT4. 3.5.3

Distribution

Ine is expressed in the central nervous system, skeletal muscle and transporting epithelia in higher insects. In Drosophila, where lne and rosA mutants are associated with sensory and central neuronal defects, drmlne protein is found in the rhabdomere and soma regions of photoreceptors in the compound eye, as well as in the lamina and axonal projections to the medulla in the optic tract, drmlne is also present in the mechanoreceptors and axonal tracts leading from the Johnston's organ in the antenna, and in the T-tubule system of flight muscle (Chiu et al., 2000). A similar distribution is seen with maslne in the nervous system of the adult hawkmoth, In the Mamluca caterpillar it is expressed in the glial cell layer and axons in the neuropile (but not in the cortical perikarya) in ganglia and by the intersegmental connectives (axonal or glial expression?), maslne is also expressed in the T-tubule system of skeletal muscle. Many orphan transporters have a non-neuronal distribution. Ine protein is found in the basolateral plasma membrane in several ion-transporting epithelia, such as the Malpighian tubules, the hindgut and rectal complex (Chiu eta/., 2000). Ine is strongly expressed in the midgut goblet cells of the hornworm caterpillar, but not in the fruitfly maggot. As suggested by its full name, Bloated tubule, Blot is expressed in Malpighian tubules, where it appears to be involved in organizing the actin cytoskeleton at the apical cortex of the tubule cells (Johnson et al., 1999). drmBlot (CG3897) is also expressed in the ring gland and eye-antennal disc of third-instar maggots (Johnson et al., 1999). Many mammalian orphan transporters, are also non-neuronal in distribution (Uhl and Hartig, 1992). 3.5.4

Kinetics and pharmacology

By definition, few transport data are available on members of this transporter orphanage from in vitro expression studies. Examination of the phenotypes of rosa and Ine mutant flies suggest that the defect in the nervous system is an

114

S. CAVENEY AND B. C. DONLY

altered membrane K + conductance. Adult drmrosA mutants have altered lightinduced electrical oscillations in the retina (Burg el al., 1996), while drmlne mutant larvae have a potassium channel defect in their neuromuscular junctions (Stern and Ganetzky, 1992). maslne has been expressed in Xenopus oocytes and the transport characteristics of the maslne protein probed with nine radiolabelled amino acids (glu, asp, GABA, pro, lys, phe, leu, tau and creatine) and seven amines (DA, 5-HT, NE, OA, HA, TA and choline) (Chiu el al., 2000). None were taken up. maslne expression, however, elevated ligandindependent K + leakage currents in frog oocytes, comparable to that induced by m a s K A A T expression (Bossi el al., 1999). Furthermore, maslne expression modulated an endogenous osmosensitive outward-going C1- current in oocytes through its interaction with an endogenous PKC/InsP3 signalling cascade (Chiu el al., 2000). maslne is suspected of being an osmolyte-sensitive transporter that facilitates CI- fluxes through second messenger systems iu silu. It may also be involved in transporting K + from the haemolymph into the midgut epithelium in the caterpillar (but not in the fruitfly) where it may help drive nutrient uptake from the midgut lumen through an H+/K + exchanger (Chiu et al., 2000). It is possible that the Blot gene product, too, may transport osmolytes across the plasma membrane and hence may be involved in cell volume control, but this needs to be demonstrated. Other orphan transporters in this family may also be primarily involved in regulating transmembrane ion fluxes in insects.

4

N a + / C I - d e p e n d e n t t r a n s p o r t e r s II

4.1 4.1. l

CHOLINETRANSPORTERS Back~;round

Acetylcholine (ACh) is the dominant excitatory neurotransmitter in the insect CNS (reviewed in Callec, 1985; Sattelle, 1985; Sattelle and Breer, 1990). Acetylcholine release from cholinergic neurons activates both ionotropic and metabotropic postsynaptic acetylcholine receptors (AChRs). Ionotropic (nicotinic) receptors (nAChRs) are multi-subunit heteromeric complexes that undergo rapid conformational change on binding their neurotransmitter ligand. They form ion channels in the membrane when activated by ligands such as the plant alkaloid nicotine. Nicotinic acetylcholine receptors are well characterized in insects and are the target of nicotinoid insecticides such as imidacloprid (Admire) (Buckingham el al., 1997). Metabotropic receptors are single polypeptides that also undergo conformational change on binding ligand and, instead of forming ion channels, activate G-protein coupled signalling systems in the target cell. Metabotropic (muscarinic) ACh receptors

NEUROTRANSMITTER TRANSPORTERS

115

(mAChRs) are named after muscarine, a fungal agonist of this receptor type. The antennal lobe, the primary site of olfactory processing in the insect brain, is particularly rich in cholinergic activity (Waldrop and Hildebrand, 1989). Afferent fibres from moth antennal sensory neurons terminate in the glomerular neuropile of the olfactory lobe, where they make cholinergic synaptic contact with intrinsic and projection olfactory interneurons (Vickers et al., 1998). Many cholinergic sensory neurons and interneurons have been identified in the thoracic and abdominal ganglia. Afferent fibres from cereal mechanoreceptors in the cockroach Perip/anem americana make cholinergic synaptic contact in the terminal abdominal ganglion with giant interneuron dendrites (Sattelle ez al., 1985). The dorsal unpaired median (DUM) interneurons that serially repeat themselves along the abdominal ganglia are also cholinergic (Buckingham ez al., 1997). An involvement of acetylcholine in neuromuscular transmission in insects has not been shown, however (Sattelle, 1985; Sattelle and Breer, 1990). The extent to which cholinergic neurons in the insect CNS make synaptic connections in the ganglionic neuropile has been revealed by the presence of high-affinity binding sites for e~-bungarotoxin, an AChR-specific toxin. Studies with [f2Sl]bungarotoxin have shown that functional nicotinic ACh receptors occur not only on post-synaptic membranes throughout the neuropile of the sixth abdominal ganglion of P. americana, but also at extrasynaptic sites on neuronal cell bodies that lie peripherally in the cortex (Sattelle el a[., 1985). These two types of channel-forming nAChR are electrophysiologically distinguishable in neurons cultured from head and thoracic ganglia of Locusla migratoria (Tareilus et a[., 1990). Whereas neurotransmitter re-uptake is implicated in synaptic inactivation in monoaminergic and amino acid-ergic neurotransmission, cholinergic signalling is not terminated directly through acetylcholine re-uptake. Acetylcholine is first degraded into choline and acetate by extracellular cholinesterase following its release into the synaptic space, This enzyme has been greatly exploited as a target of organophosphate and carbamate insecticides. Choline and acetate are removed from the synaptic space by separate transport systems in cholinergic neurons (Yamamura and Snyder, 1972; Kuhar and Murrin, 1978) and then converted to ACh by neuronal choline acetyltransferase. High-affinity and sodium-dependent uptake of choline is a diagnostic feature of cholinergic nerve terminals and appears to be the rate-limiting step in ACh synthesis in insects (Knipper and Breer, 1986; Breer and Knipper, 1990). Choline is also a precursor in the synthesis of phosphatidyl choline, an important membrane phospholipid. Cells involved in lipid synthesis (e.g. hepatocytes in mammals, and presumably fat body cells in insects) have concentrative Na+-independent organic cationic amine transporters (OCTs) with a low-micromolar affinity for choline as a transport substrate (Sinclair el a/., 2000). The properties of choline transport in the insect CNS were first documented by Breer (1982) and Bermudez el al. (1985) (section 4.1.4).

116

4.1.2

S. CAVENEY AND B. C. DONLY

Structure

A high-affinity choline transport protein has been isolated from locust synaptosomes and reconstituted in liposomes, where it was shown to accumulate choline in a Na+-dependent, hemicholinium-3-sensitive manner (Knipper el al., 1991). Hemicholinium-3 is a potent and selective inhibitor of highaffinity choline uptake. The protein had an apparent molecular mass of 90kDa but is apparently heavily glycosylated, since treatment with endoglycosidase F (which cleaves asparagine-linked oligosaccharides) reduced this value to 65 kDa, which is approximately the deduced MW of the CHT's cloned from other animals (Table 5). Since the activity of the 90 kDa form of the protein can be measured in reconstituted liposomes, it appears that it does not need to form a heteromeric complex with other proteins in the synaptosomal membrane in order to function. Although monoclonal antibodies were raised against the locust choline transporter (Knipper et al., 1989c) and used to purify it (Knipper et al., 1991), these antibodies have apparently not been used to screen locust cDNA libraries to isolate the transporter cDNA. Attempts in the early 1990s to clone mammalian choline transporters by homology-based PCR, founded on the premise that these transporters belonged to the Na+/Cl--dependent family of GABA/monoamine transporter-like proteins, were futile. Functional expression of rat CHOTI, a cholinetransporting protein cloned by homology-based PCR (Mayser el al., 1992) revealed that it lacked the hemicholinium-3 sensitivity typical of choline uptake by cholinergic neurons (Gonzalez and Uhl, 1994). Neither did the pattern of distribution of CHOTI coincide with that of cholinergic interneurons in the mammalian brain. It is now accepted that CHOT1 cDNA encodes a creatine transporter widely distributed in non-neuronal tissues such as the intestine and adrenal gland in mammals (Gonzalez and Uhl, 1994; Schloss and Mayser, 1994). The insect homologue of this creatine transporter has not been identified. An authentic high-affinity, and apparently exclusively synaptic, choline transporter, cho-1, has recently been cloned fi-om the nematode Caenorhabdilis elegans (Okuda et al., 2000). When expressed in frog oocytes, the cho-1 cDNA produced a hemicholinium-3-sensitive, Na +- and Cl--dependent, high-affinity choline transporter. The cDNAs coding tbr hemicholinium-3-sensitive choline transport proteins (CHTs) in other animals have now been cloned. They include rat (Okuda et al., 2000), human (Okuda and Haga, 2000; Apparsundaram et al., 2000), mouse (NCBI Accession No. AAG36945), and horseshoe crab Limuhls p o l y p h e m u s (Wang et al., 2001). In the fly, a cDNA has also been sequenced as part of the Berkeley Drosophila Genome Project (NCBI accession No. AAK77253) that is highly homologous to this group of cDNAs. Based on the very high levels of sequence identity among these transporters, they appear to represent a unique family of transporters (Fig. 9). The known characteristics of these transporters are listed in Table 5 but, as of yet, details of the functional domains are not available. Membrane topology

NEUROTRANSMII-I-ER TRANSPORTERS

117

E

©

i

:z)

F, =: L.,, C ,...,

c-i

.< .< O >-.

"d "5 ~ ~ '-.--J

E

'S

z

& 2c~ L~ O C, O

"5

O ,O

.'-4

Cz.

'-m

~

c-.

S. CAVENEY AND 8. C. DONLY

118

H i

0 ~ 0

m o ~

cL~,

3,

E

ca)

~

..

d.j

~

¢,1

¢;,

~

¢".-i

©

~

<,,¢

~"c ii

i

i

.-~

q)

0

~

° °

~Nm~t

~

~

X

~g, ~,~

"~

u ,r--- ...ci

~E~.~ ..o ~ .~

NEUROTRANSMITTER TRANSPORTERS

119

models for the various CHTs are not in agreement, with some suggesting a 12 TMD configuration similar to other Na+/C1--dependent transporters (Okuda et ell., 2000) and others a 13 TMD configuration with six extracellular loops and seven intracellular ones (Wang el ell., 2001; Apparsundaram et al., 2000), The presence of 13 TMDs would be consistent with the assertion of Wang el al. (2001) that the CHTs are in fact members of the sodium-dependent glucose transporter family (Turk and Wright, 1997). An unrelated non-neuronal highaffinity choline transporter ( m O c t l / S I c 2 2 a l ) , belonging to the organic cation transporter (OCT) family, was recently described from the mouse liver (Sinclair et al., 2000). Unlike CHT, this hepatocyte choline transporter transports tetraethylammonium (TEA) and appears not to require Na + in order to be active.

4.1.3

Distribution

Sodium-ion-dependent choline uptake in the insect CNS is restricted to cholinergic neurons. Because all regions of the insect CNS contain neurons with cholinergic activity, Na+-dependent and high-affinity choline uptake is found throughout the CNS. Choline transporters appear to be restricted to the ganglionic neuropile. A monoclonal antibody raised against the locust 90 kDa synaptosomal choline transporter protein strongly labelled the glomerular neuropile of the antennal lobe and all three neuropile regions (lamina, medulla and lobula) of the optic tract (Knipper el al., 1989c), implying that some sensory neurons and/or interneurons in these sensory processing regions are cholinergic. The monoclonal antibody against locust CHT also stained areas of the neuropile in the locust mesothoracic ganglion, but the cortex, perineurial sheath and the ganglionic connectives were unstained (Knipper et al., 1989c). Extrajunctional choline transporters and/or glial choline transporters have not been demonstrated in insect ganglia. I , situ hybridization analysis of the distribution of choline transporter RNA expression in the insect CNS also remains to be done.

4.1.4

Kinetics and Pharmacology

Breer (1981, 1982) first demonstrated high-affinity choline uptake in the insect CNS using a crude synaptosomal preparation from locust ganglia. By incubating synaptosomes in [3H]choline, Breer identified a hemicholinium-3-sensitive, Na*-dependent, choline uptake system with an affinity (Kin) for choline of 0.98 IbM. A low-affinity choline transport system (Kin = 24.6 #M), apparently less Na+-dependent, was also detected in these synaptosomal preparations. Hemicholinium-3 binding has been used to show that synaptosomes prepared from locust head and thoracic ganglia are highly enriched with cholinergic

120

S. CAVENEY AND B. C. DONLY

nerve terminals (Knipper et al., 1989b). Bermudez et al. (1985) measured Na +dependent high-affinity [3H]choline uptake in neuronal cultures derived from cockroach embryos. A high-affinity component with a K,,1 of 0.5#M and a lower affinity component with a Km of 5.4#M were reported. Autoradiographs of these cultured neurons after incubation with [3H]cho[ine revealed that the majority of the axons of the neurons were labelled, but their somata were not. Saline containing l # M hemicholinium-3 reduced the uptake of 0.5 #M [3H]choline by cultured cockroach neurons by about 80% (Bermudez et al., 1985). The cloned rat choline transporter rCHT, when expressed in frog oocytes, has an affinity for choline of 2.2 #M (Okuda et al., 2000). Like other mammalian choline transporters, it is also much more strongly inhibited by hemicholinium-3 than the insect choline transport systems described above, being completely blocked by I/~M hemicholinium-3 (K~ range, l0 to 100riM) (Okuda et al., 2000). Similarly, the affinity of locust synaptosomes for [3H]hemicholinium-3 is much greater than their affinity for choline. Two hemicholinium-3 binding sites with Kjs of about 3 nM and 69 nM are reported (Knipper et al., 1992), within the range of the mammalian choline transporters (Okuda et al., 2000). This affinity is much greater than the hemicholinium-3 inhibition data would predict, suggesting that these data need to be reevaluated. [3H]choline uptake into synaptosomes is chlorpromazine sensitive (Breer, 1982). Choline transport is relatively insensitive to the metabolic inhibitors sodium azide and 2,4-dinitrophenol but sensitive to iodoacetate, an inhibitor of glycolysis (Breer, 1982; Bermudez et ell., 1985). Although it is possible that the iodoacetate effect is due to its ability to alkylate transporter SH groups (Breer, 1982), the conserved cysteine residues in CHTs are found either within the transmembrane domains or in the cytoplasmic C-terminus (Fig. 9). ATP inhibits the uptake of choline by locust synaptosomes (Knipper and Breer, 1986). Finally, monoclonal antibodies raised against the locust synaptosomal choline transporter inhibit choline uptake (Knipper et al., 1989c). Blocking choline uptake would have a different effect to blocking other NTTs, since it would suppress ACh synthesis without causing a buildup of ACh in the synaptic space. High-affinity choline transport in insects has a stringent requirement for Na + (Breer, 1983; Bermudez et al., 1985). The ion Li + is a poor substitute for Na + in driving choline uptake, reminiscent of the cation dependency of members of the Na+/K + and Na+/Cl -dependent families of neurotransmitter transporters discussed earlier. Uptake of [3H]choline by the locust synaptosomal choline transporter peaked in saline containing 150 mM NaC1, yielding a Km for Na + dependency of between 40 and 60raM (Breer, 1982). The synaptosomal transporter appears to require K + on the inside of the vesicles in order to take up choline (Breer, 1982), but it is not known whether this is due to K + countertransport during choline/Na + co-transport (i.e. whether it is an electrogenic component of the choline transporter). Elevating external [K +] lowers [3H]choline uptake in intact cells, suggesting that choline uptake may

NEUROTRANSMITTER TRANSPORTERS

121

be membrane potential sensitive (Bermudez el al., 1985). Choline uptake by the locust synaptosomal transporter also has a strong requirement for external CI . Phosphate, isothiocyanate, sulphate and acetate cannot substitute for external CI- in this anion dependency (Breer, 1983). The halide ions Brand 1- were not tested. Chloride dependence of mammalian choline transporters is well documented (Okuda et al., 2000). Br- substitutes well for CIin driving choline uptake by the cloned rat CHTI, but 1- does not (Okuda et al., 2000). 4.1.5

Regulation

Choline uptake by cholinergic nerve terminals in insects appears to be coupled to neuronal activity (Breer and Knipper, 1990). This coupling appears to be through auto- and hetero-receptors on the presynaptic membrane. The autoreceptors appear to be muscarinic M-2 subtype (Knipper and Breer, 1988). The release of acetylcholine from insect synaptosomes is also regulated by presynaptic receptors (Knipper and Breer, 1989). Sodium-ion-dependent high-affinity choline uptake in synaptosomes prepared from the head and thoracic ganglia of Locusta migraloria is downregulated by the noncompetitive agonist of M-2 receptors oxotremorine and upregulated by the M-2 antagonist, atropine. Synaptosomal choline uptake can be stimulated by octopamine (but not by serotonin or dopamine), and inhibited by D-ala-D-leu enkephalin (Breer and Knipper, 1990), suggesting that receptors are present in the cholinergic terminals. Regulation of choline uptake is apparently through the direct phosphorylation of the choline transporter by protein kinase A and protein kinase C second messenger pathways. Phosphorylation appears to increase the average occupancy time of choline transporters in the synaptosoreal membrane (Knipper et al., 1992). Several PKA and PKC phosphorylation consensus sequences occur along the cytoplasmic loops and cytoplasmic Cterminus of hCHT (Apparsundaram el al., 2000).

5 Other Na*-dependent transporters 5.1

5.1.1

HISTAMINE TRANSPORTER

Background

The synthesis of histamine (HA) in insect nervous tissue was first reported for the moth M a n d u c a sexta (Maxwell et al., 1978). Most of the histamine (>97%) in the nervous system is present in the retina and the optic lamina (Elias and Evans, 1983a; Pirvola et al., 1988). The suboesophageal and thoracic ganglia contain much of the rest. HA-immunocytochemical studies on the CNS of

122

S. CAVENEY AND B. C, DONLY

many insects, including cockroach (Pirvola et al., 1988), cricket and bee (Bornhauser and Meyer, 1997) and flies (Monastirioti, 1999), indicated that the retinal photoreceptors and the lamina interneurons stained most intensely for HA. Histamine is now confirmed as an inhibitory neurotransmitter released by most, if not all arthropod photoreceptor cells (Limulus Hart and Bate[le, 1991; Musca Hardie, 1987, 1989; locust Simmons and Hardie, 1989; Drosophila Sarthy, 1991; Balanus Callaway and Stuart, 1989; reviewed in N~issel, 1999: Stuart, 1999). Histamine released by photoreceptor cells is known to activate a histamine-gated CI conductance in the first optic neuropilc (Hardie, 1989; Nfissel, 1999). A histamine-gated chloride channel has been recently cloned from the monopolar interneurons in the lamina neuropile of Drosophila, where it is expressed exclusively (Witte et al., 2001). Histamineimmunoreactive neurons are also present in the midbrain, suboesophageal ganglion and segmental ganglia in insects (N'/issel el al., 1990; Homberg and Hildebrand, 1991; Lundquist et al., 1996). Selective glial uptake of [3H]histamine is reported to occur along the borders of the lobula and medulla regions of the optic tract in the locust Schistocerca gregaria (Elias and Evans, 1983a). This is unusual for two reasons. First, monoamine transporters are normally expressed selectively by the neurons that release a particular biogenic monoamine (see above), and second, the glial cells demonstrating histamine uptake are in regions of the locust optic lobe that contain relatively little endogenous HA (Elias and Evans, 1983a). If histamine clearance from the photoreceptor synaptic space plays a critical role in vision by disinhibiting the first-order interneurons, HA transporter activity would be expected to occur in the medullary and possibly other optic tract neuropiles. More recently, a Na+-dependent histamine uptake system has been characterized in the photoreceptor axons and terminals and associated glia in the median eye of the barnacle Bahmus nubilis (Stuart et al., 1996). Whether neuronal and glial uptake involves the identical histamine transport system is not known. As in the case of other neurotransmitters, the maintenance of the histamine pool in the presynaptic terminal involves both re-uptake of released transmitter and de novo synthesis. However, the bulk of the histamine used to replenish the transmitter supply in arthropod photoreceptors is apparently recovered from the synaptic cleft, rather than obtained by de novo synthesis from histidine in the nerve terminal (Morgan et al., 1999), emphasizing the importance of the histamine transporter in synaptic function. Arthropod photoreceptor glial cells contain a histidine uptake system that may supply histamine de novo to the photoreceptor neurons following its conversion by histidine decarboxylase (Morgan et al., 1999). 5.1.2 Molecular biology Initially it was expected that the neuronal histamine transporter would turn out to be a member of the Na+/CI -dependent GABA/monoamine transporter

NEU ROTRANSMI-I-IER TRANSPORTERS

123

family (Stuart et al., 1996). At the time of writing, however, no histamine transporters have been cloned from any animal using a RT-PCR strategy based on sequence homology with this family. An expression-cloning strategy to isolate transporter cDNAs, as pioneered by Hediger and his colleagues (Hediger et al., 1987), may need to be adopted,

5.1.3

Distribution

Although an insect histamine transporter remains to be cloned and its distribution in the insect brain mapped, this transporter is unlikely to be restricted to the optic tract. In addition to the histamine-gated chloride channel, a second type of histamine receptor, related to the G-coupled H 1 receptor of mammals, is expressed by a small number of neurons in the protocerebrum, where they terminate in the central and mushroom bodies (N'assel, 1999). These histaminergic neurons are thought to have wide-field inhibitory fnnctions distinct from GABA-ergic neurons in the brain (N~issel, 1999). In the sphinx moth Manduca sexta, strong histamine immunoreactivity was detected in 10 neurons in the median protocerebrum and one pair in the suboesophageal ganglion (Homberg and Hildebrand, 1991). These HA-immunoreactive neurons have an organization similar to that of serotonin-reactive neurons in the brain and suboesophageal ganglion. In the medial protocerebrum of the CNS of Locusm migratoria, a set of inhibitory histaminergic interneurons forms part of a photosensitive circadian oscillator (Lundquist et al., 1996).

5.1.4

Kinetics and pharmacolo~y

Histamine uptake by arthropod (barnacle) photoreceptors is partially blocked by 20 # M chlorpromazine or phenoxybenzamine, but is insensitive to cocaine or desipramine (Stuart eta[., 1996), which are relatively non-selective blockers of dopamine and octopamine uptake in insects (Wierenga and Hollingworth, 1990: Nathanson el al., 1993; Malutan et al., 2002).

6

Putative n e u r o t r a n s m i t t e r transporters

It has been suggested that the amino acids fi-alanine, taurine and glycine may act as neurotransmitters in insects. /-3-Alanine and glycine may act as agonists of the GABA receptor (Osborne, 1996), Taurine is reported to have an inhibitory mode of action distinct from that of other putative inhibitory amino acid-neurotransmitters (Hue el ell., 1981: Giles and Usherwood, 1985; Wafford and Sattelle, 1986).

124

6.1

S. CAVENEY AND B. C. DONLY

GLYCINETRANSPORTERS

Only two types of neurotransmitter-gated chloride channel (inhibitory ionotropic receptors) are found in vertebrates, one gated by glycine and the other by GABA. Invertebrates, on the other hand, have four types of neurotransmitter-gated chloride channels, namely GABAA-, glutamate-, histamine- and 5-HT-gated channels. The Drosophila genome contains genes for at least 11 ligand-gated chloride channels, one of which codes the ionotropic histamine receptor. According to Witte et al. (2001), the histamine-gated chloride channel is the arthropod equivalent of the glycine-gated chloride channel. These authors state that glycine receptors are apparently lacking from the arthropod CNS. Glycine-induced hyperpolarization of cockroach and locust neurons may be through an action on GABA receptors (Pitman, 1985). Sequence CG5549 from the orphan transporters listed in Table 3 is the closest fly homologue of a mammalian glycine transporter, with 48% amino acid sequence identity.

6.2

TAUR1NE TRANSPORTERS

Taurine (2-aminoethanesulphonic acid) is one of the most abundant amino acids in animals and serves many physiological functions (Huxtable~ 1992). Taurine-immunoreactive neurons have been detected in the brain, retina and ventral nerve cord of insects (reviewed in Bicker, 1992; Stevenson, 1999). Taurine co-localizes in insect neurons that use as chemical signals the conventional neurotransmitters acetylcholine, GABA and octopamine, and it appears that taurine may modulate the function of these neurons by regulating neurotransmitter release (Hayakawa et al., 1987; Whitton el al., 1988). Consequently, taurine is not generally regarded as an independent neurotransmitter in the insect nervous system (Stevenson, 1999). Immunocytochemical data indicate that taurine co-localizes with OA in the neurosecretory DUM cells in Periplaneta americana (Nurnberger et al., 1993) and Locusta mi,gratoria (Stevenson, 1999). Taurine-immunoreactivity is also reported in other neurosecretory OA-ergic neurons in the abdominal ganglia and identifiable GABA-ergic interneurons in the thoracic ganglia of the locust (Stevenson, 1999). OA-ergic neurons in the brain of cockroach or locust were not taurine-immunoreactive (Nurnberger el al., 1993: Stevenson, 1999). Whether taurine serves as a co-neurotransmitter in the CNS, i.e. binds to a neuronal taurine receptor in these insects, is unclear. Taurine may act locally to downregulate OA release and suppress the stress-induced elevation of haemolymph OA (Hayakawa et al., 1987; Stevenson and Sp6rhase-Eichmann. 1995). To date there is no report of a high-affinity taurine uptake system in the insect CNS. One might expect a taurine uptake system in the taurine-positive GABAergic interneurons. On the other hand, the absence of a taurine transporter in (dorsal/ventral) unpaired median neurosecretory neurons would match the

NEUROTRANSMITTER TRANSPORTERS

125

observation that these cells are the only OA-ergic neurons that lack an octopamine transporter in the Trichoplusia ni caterpillar (Malutan et al., 2002). Putative D r m G A T (CG1732; Table 3) listed in FlyBase has an amino acid sequence with closest similarity (50% sequence identity) to a fish taurine transporter.

7 7, l

Applications to insect control RELEVANCE OF INSECT NEUROPHYSIOLOGY TO PEST CONTROL

A long-standing truism is that a pest species needs to be well understood before it can be controlled by rational and species-specific means. In the 1930s the pioneer physiologist Vincent B. Wigglesworth started his lifelong experiments on Rhodnius prolixus, a blood-sucking bug that transmits Chagas" disease. He enjoyed a brilliant career as a physiologist without implementing an approach to control this tropical disease vector. But the next generation of insect physiologists and biochemists were able to put his curiosity-driven findings to practical use elsewhere, by introducing insect hormone analogues (juvanoids) for insect control. A completely new approach using transgenic symbionts has subsequently been suggested to render R. prolixus immune to the trypanosome that causes Chagas" disease (Durvascula et al., 1997). Precise knowledge of the insect nervous system per se has rarely led directly to the discovery of new pesticides. Rather, the screening and identification of insecticides has followed a shot-gun/fishing-expedition approach. Natural pesticides (such as avermectin) or synthetic organics that were discovered fortuitously to be pharmacologically active have more frequently helped us determine how the insect CNS works, rather than vice versa. The use of robotized mass screening approaches to identify blockers of new molecular targets in the insect CNS is expensive but no doubt will soon reach a commercial level of success in the agrochemical industry. Molecular genetics and informatics ('genomics' and "proteomics') will be the tools used by researchers during the next decade of research into insect control. If the current revolution in medical research is any indication, understanding of the mode of action of the hormones, neuropeptides, growth factors and other extracellular signals that regulate insect metabolism and growth (including knowledge of their receptors and signal transduction pathways, etc.) will gain ever-increasing prominence in the design of rational approaches to insect control. 7.2

N E U R O T R A N S M I T T E R TRANSPORTERS AS NEW TARGETS FOR INSECT CONTROL

At the present time, about 90% of the world-wide market in insect control is in "bard" commercial insecticides (those that act on the insect CNS and muscle)

126

S. CAVENEY AND B. C. DONLY

designed to disrupt a very small number of proteins essential to the normal functioning of the nervous system. These neurotargets include voltage-dependent sodium channels (Bloomquist, 1996: Nakahashi, 1996; Zlotkin, 1999), GABAA receptors (Mullin et ell., 1991: Hosie el al., 1997), nicotinic acetylcholine receptors (Buckingham et al., 1997), glutamate-gated chloride channels (Scott and Duce, 1985), acetylcholinesterase (Eldefrawi, 1985; Fournier el al., 1992: Zhu and Clark, 1995) and octopamine receptors (Evans, 1985: Roeder, 1999). The intense selection pressure caused by extensive field use of pesticides aimed at this small number of insecticide targets has led to widespread resistance in pest insects to, for instance, organophosphate and carbamate pesticides (Clark et al., 1995: Hosie et al., 1997; Vais et al., 1997). Many forms of 'knock-down' resistance and cross-resistance appear to result from site-specific mutations in the multiple insecticide-binding sites on this small number of overworked targets. It is self-evident that new neuronal targets need to be identified in humankind's continuing chemical war against pest insects. What new targets does the neural synapse offer for insect control? The availability of complete genome sequences for D. melanogaster and C. elegans has assisted in the discovery of related genes in numerous invertebrates and has placed approximate limits on the numbers of genes within specific gene families (Venuti and Cserjesi, 2000). It is evident that many genes expressed in the insect CNS have closely-related homologues in mammals and worms (Littleton and Ganetzky, 2000). Comparison of cDNA sequences and their translation products from the CNS of pest insects with the published genomes may help confirm their identities. More importantly, it may reveal cDNAs that are unique to the insect CNS. The products of such sequences make the most promising neural targets for insect control. This review has attempted to document and integrate recent research on the isolation and cloning of complementary DNAs encoding members of the three multigene families that encode Na+-dependent high-affinity transporters of neurotransmitters or neurotransmitter precursors in the insect CNS. In most instances these transporter cDNAs have been expressed in #7 vitro cell systems and the pharmacology and substrate kinetics of the gene products assessed. Our premise is that these proteins may provide an important new set of nerve targets for insect control strategies. In this context, it is promising to find that there is limited redundancy in the genes encoding these transporters, and also some selectivity, in that some transporters (e.g. OAT) may not be present in all insect orders. Although neurotransmitter transporters remain at present hypothetical targets for new insecticides, we suspect that they may well prove to be practical targets. Indeed, many potent psychoactive drugs ('selective re-uptake blockers') act on neurotransmitter transporters in the human brain. Consequently, they have received much attention in recent neurophysiological and molecular biological brain research. As new sites of action for insecticidal compounds, these proteins may ensure the neuromuscular junction and nerve

NEUROTRANSMITTER TRANSPORTERS

127

synapse continue as effective targets for insect control in the future. But the question remains: are neurotransmitter transporter proteins 'hard' enough targets that blocking them will produce the 'knock-down' control of pests so favoured by modern agriculture'? Obviously, the agrochemical industry is primarily interested in pursuing research that has a high likelihood of leading to new ways to "kill bugs dead'. But how may it best be demonstrated that neurotransmitter transporters offer such an approach'? '7.3

FI T U R E DIRECTIONS

One promising approach would be to examine the effects of permanently or transiently knocking out the gene coding for a particular neurotransmitter transporter. The effects of perturbing neurotransmitter transport in this way may be subtle and/or not prove instantly lethal, as suggested by neurotransmitter transporter knockout (Giros et al., 1996) or antisense knockdown experiments (Rothstein el al., 1996: Simantov et al., 1999: Rao et al.. 2001) in rodents. Mammalian systems may not be the best model for such conclusions, however, due to the greater redundancy of transporter function inherent to vertebrate systems. Furthermore, behavioural modifications ensuing from the silencing of neurotransmitter transporter genes in a pest insect could still be effective in limiting its potential for crop damage without actually being lethal. The feasibility of doing P-element mediated gene knock-out studies on D r o s o p h i l a (Preston el al., 1996), or RNAi silencing studies (Hunter, 2000) and morpholino antisense RNA knockdowns (Nasevicius and Ekker, 2000) on selected pest arthropods needs to be examined. Genetic manipulation is the most direct way to determine which of the cloned CNS proteins are absolutely essential for neurotransmission, and consequently, which, if any, represent the most attractive targets in insecticide development. However, it is important to remember that genetically altered insects can display compensatory adjustments in the expression of other genes and pathways to optimize the physiology of affected individuals. Thus for example, in the case of the mouse DAT knockout (Giros et al., 1996), adaptive changes were observed in the form of a reduction in dopamine receptor levels as well as levels of tyrosine hydroxylase, the enzyme responsible for dopamine synthesis. Such compensatory adjustments may perturb overall physiology to allow the genetic defect to be tolerated. This type of compensation mechanism would not be immediately available to an insect exposed directly to a pharmacological agent blocking transporter activity. Another approach would be to develop non-radiometric HTS ('highthroughput screening') assays to search for natural or synthetic chemicals that disrupt neurotransmitter transporter function i~t vitro. Initially, baculovirus/expression vector systems can be used in a functional HTS screen lk~r blockers of insect neurotransmitter transport. Most insect neurotransmitter co-transporters have been transiently expressed this way. A more desirable

128

S. CAVENEY AND B. C. DONLY

long-term solution for HTS, however, would be to use cell lines stably expressing the targeted insect transporter. The utilization of radiolabelled substrates used in the small-scale characterization of insect transporters may not be economically feasible in a large-scale HTS protocol (consider the problems inherent in radioactive-waste disposal), suggesting that non-radiometric approaches would need to be developed. Any blockers so discovered could then serve as lead compounds for optimization through the use of combinatorial libraries. Ultimately, the resulting compounds showing targeted specificity would have to be screened for insecticidal activity on pest species. A variety of plants and mushrooms are known to synthesize substantial amounts of non-protein amino acids and monoamine mimics capable of disrupting normal neurotransmitter activity, transport and/or metabolism. Any potent blockers present in plant extracts, or even arthropod venoms, could then be isolated and chemically characterized. The mode of action of many neuroactive compounds toxic to leaf- and seed-eating insects is not clear. 7.4

POSTGENOMIC PROSPECTS FOR INSECT PHYSIOLOGY RESEARCH

Despite the obvious validity of using certain model insects such as Drosophila to study basic cellular and genetic mechanisms, our understanding of the molecular genetics of pest insects is rudimentary. Certainly, the genomes of many pest insects will at some point be completely sequenced, leading to more 'biorational' approaches for the control of insect pests. It is ironic that funding for studies on the molecular physiology of crop plants will overshadow that of their pests in this decade. Making crop plants more pest-resistant rather than spraying ever more toxic chemicals is a politically, commercially and environmentally more acceptable approach. However, making better plants by transgenic means will still require useful transgenes, many of which will come from insect pathogens and other organisms. This is clearly a growth area for young researchers. But despite the recent introduction of transgenic crop plants modified for increased insect resistance, and the enhanced use of biocontrol methods to suppress crop pests, in practice the reality is that insect outbreaks will still have to be selectively controlled by the application of conventional pesticides for some time yet. In this regard, the challenge for the future is to discover new, more sustainable targets through which crop pests can be controlled. The introduction of pesticides selectively aimed at these novel targets would reduce the load that the broad-spectrum neurotoxic pesticides currently in use place on the environment. Have insect physiologists provided enough basic information for molecular biologists to reach these goals'? Traditional tissue-level approaches to insect physiology are on the wane. This is unfortunate since the post-genomic era will need physiologists to establish what many of the genes sequenced in pest genomes do, In practical terms one may even ask whether modern physiological research is competitive in the intense competition for limited research

NEUROTRANSMITTER TRANSPORTERS

129

dollars from research groups that use expensive high-tech molecular genetic approaches in their search for new ways to control insects. The decline in the number of insect physiologists worldwide, mostly by attrition, during the last 20 years is telling. Is insect physiology last becoming a "sunset science', along the lines of comparative anatomy and biochemical enzymology years ago'? The term 'physiology' implies a limited knowledge of the biological system under study. To gain a more complete understanding of how an insect's body functions, the discipline of insect physiology continues to transform itself into a set of narrower themes, such as neurobiology, endocrinology, immunology, molecular biology and biochemistry. As the pendulum swings today toward discrete molecular sciences it is important to remember that systems-level approaches will be required in the integrative phase to follow. The directions that insect science takes in the post-genomic era will be fascinating to watch. While the application of new genetic tools changes our approach to insect science, much of what has been learned remains to be implemented in control strategies. Only time will tell how the information presented in this review will be applied in the post-genomic era.

Acknowledgements

The authors have been supported by Agriculture and Agri-Food Canada, the Natural Sciences and Engineering Research Council of Canada and Aventis CropScience. We thank Tabita Malutan, Richard Gardiner, Heather McLean and Pam Gallant for the use of unpublished material, and Anita Caveney tk)r providing useful comments on the manuscript.

References

Ali, D. W. and Orchard, 1. (1996). The uptake and release of serotonin and dopamine associated with locust (gocusta migratoHa) salivary glands. J. Exp. Biol. 199, 699709. Amara, S. G. (ed.) (1998). Neurotransmitter transporters. Melh. Enzymol. 296, 425 436. Amara, S. G. and Kuhar, M. J. (1993). Neurotransmitter transporters: recent progress. Atom. Rev. Neurosci. 16, 73 93. Apparsundaram, S., Moore, K. R., Malone, M. D., Hartzell, H. C. and Blakely, R. (1997). Molecular cloning and characterization of an k-epinephrine transporter from sympathetic ganglia of the bullfrog Rana catesbiana. J. Neurosci. 17, 2691 2702. Apparsundaram, S., Ferguson, S. M., George, A. L. and Blakely, R. D. (20011). Molecular cloning of a human, hemicholinium-3-sensitive choline transporter. Biochem. Biophys. Res. Comm. 276, 862-867. Arakawa, S., Gocayne, J. D., McCombie, W. R., Urquhart, D. A., Hall, L. M., Fraser, C. M. and Venter, J. C. (1990). Cloning, localization, and permanent expression of a Drosophila octopamine receptor. Neuro, 2, 343-354. Araque, A., Parpura, V., Sanzgiri, R. P. and Haydon, P. G. (1999). Tripartite synapses: glia, the unacknowledged partner. Tremls Neurosci. 22, 208-215.

130

S. CAVENEY AND B. C. DONLY

Arriza, J. L., Eliasof, S., Kavanaugh, M, P. and Amara, S. G. (1997). Excitatory amino acid transporter 5, a retinal glutamate transporter coupled to a chloride conductance. Proc. NaN Acad. Sci. USA 94, 4155 4160. Attwell, D. and Mobbs, P. (1994). Neurotransmitter transporters. Curr. Opin. 3&,urobiol. 4, 353 359. Auger, C. and Attwell, D. (2000). Fast removal of synaptic glutamate by postsynaptic transporters. Neuron 28, 547 558. Awad, E. W., Amiri, M. H. and Stephen, R. S. (1997). Developmental changes in biogenic amine levels in the central nervous system and the haemolymph of the eastern death's head hawk moth, Acherontia styx (Lepidoptera: Sphingidae). Comp. Biochem. Physiol. 116C, 219 225. Baines, R. A., Tyrer, N. M. and Mason, J. C. (1989). The innervation of locust salivary glands. !. lnnervation and analysis of neurotransmitters. J. Comp. Physiol. A165, 395 405. Barker, E. L. and Blakely, R. D. (1996). Identification of a single amino acid, phenylalanine 586, that is responsible for the high affinity interactions of tricyclic antidepressants with the human serotonin transporter. Moh, c. Pharmacol. 50, 957 965. Baxter, C, F. and Torralba, G. F. (1975). y-Aminobutyric acid and glutamate decarboxylase (L-glutamate l-carboxy-lyase E,C. 4.1.1.15) in the nervous system of the cockroach, Periplaneta americana. I. Regional distribution and properties of the enzyme. BraiH Research 84, 383 397. Beal[, C. J. and Hirsch, J. (1987). Regulation of the Drosophila dopa decarboxylase gene in neuronal and glial cells. Genes am/Development l, 510 520. Beckman, M. L. and Quick, M. W. (1998). Neurotransmitter transporters: regulators of lkmction and functional regulation. J. Membrane Biol. 164, 110. Beckman, M. L., Bernstein, E. M. and Quick, M. W. (1998). Protein kinase C regulates the interaction between a GABA transporter and syntaxin I A. J. Neurosci. 18, 61036112. Bennett, E. R. and Kanner, B. 1. (1997). The membrane topology o f G A T - I , a (Na ~ + CI ) -coupled v-aminobutyric acid transporter from rat brain. J. Biol. Chem. 272, 1203 1210. Bermudez, I. and Beadle, D. J. (1989). High-affinity uptake of [3H]serotonin in cultured neurons of the cockroach Periphmeta americana (L.). Arch. Msect Biochem. Physiol. 12, 395--405. Bermudez, 1., Botham, R. P, and Beadle, D. J. (1988). High- and low-affinity uptake of amino acid transmitters in cultured neurons and muscle cells of the cockroach, Periphmeta americana. Msect Biochem. lg, 249 262. Bermudez, I.. Lees, G., Middleton, C., Botham, R. and Beadle, D. J. (1985). Choline uptake by cultured neurons from the central nervous system of embryonic cockroaches, h~sect Biochem. 15, 427-434. Bernstein, E. M. and Quick, M. W. (1998). Regulation of ?'-aminobutyric acid (GABA) transporters by extracellular GABA..I. Biol. Chem. 274, 889 895. Besson, M. T., Soustelle, L. and Birman, S. (1999). Identification and structural characterization of two genes encoding glutamate transporter homo[ogues differently expressed in the nervous system of Drosophikt melanogaster. FEBS Lett. 443, 97 104. Besson, M. T., Soustelle, L. and Birman, S. (2000). Selective high-affinity transport of aspartate by a Drosophih~ homologue of the excitatory amino-acid transporters. Curt. Biol. 10. 207-210. Bicker, G. (1992). Taurine in the insect nervous system. Cow,q). Biochem. Physiol. (" 103. 423 428.

NEUROTRANSMITTER TRANSPORTERS

131

Bicker, G., Schiller, Ottersen, O. P. and Storm-Mathisen, J. (1988). Glutamate-like immunoreactivity in identified neuronal populations of insect nervous systems. J. Neurosci. 8, 2108-2122. Bismuth, Y., Kavanaugh, M. P. and Kanner, B. 1. (1997). Tyrosine 140 of the gaminobutyric acid transporter GAT-I plays a critical role in neurotransmitter recognition. J. Biol. Chem. 272, 16 096-16 102. Blakely, R. D. and Bauman, A. L. (2000). Biogenic amine transporters. Regulation in flux. Curt. Opinion Neurohio/. 10, 328 336. Blakely, R. D., Bersom H. E., Fremeau, R. T. Jr, Carom M. G., Peek, M. M., Prince, H. K. and Bradley, C. C. (1991). Cloning and expression of a functional serotonin transporter from rat brain, Nature 354, 66 70. Blenau, W., Belfanz, S. and Baumann, A. (2000). Amtvrl: Characterization of a gene from honeybee (Apis reel/libra) brain encoding a functional tyramine receptor. J. Neurochem. 74, 900-908. Blitz, D. M. and Nusbaum, M. P. (1999). Distinct functions for cotransmitters mediating motor pattern selection. J. Neurosci. 19, 6774 6783. Bloomquist. J. R. (1996). Ion channels as targets for insecticides. Atom. Re~,. E~ltomo/. 41, 163 190. Borden. L. A. (1996). GABA transporter heterogeneity: pharmacology and cellular localization. Neurochem. Int. 29, 335 356. Borden, L. A., Dhar, T. G. M., Smith, K. E., Weinshank, R. L., Branchek, T. A. and Gluchowski, C. (1994). Tiagabine, SK&F 89976-A, C1-966, and NNC-711 are selective for the cloned GABA transporter G A T - l . Ear. J. Pharmaco/. 269, 219-224. Bornhauser, B. C. and Meyer, E. P. (1997). Histamine-like immunoreactivity in the visual system and brain of an orthopteran and a hymenopteran insect. Cell Tissue Res. 287, 211 221. Bossi. E., Centinaio, E., Castagna, M., Giovannardi, S., Vincenti, S., Sacchi, V. F. and Peres, A. (1999). lon binding and permeation through the lepidopteran amino acid transporter KAAT1 expressed in Xenopus oocytes. J. Physiol. Lond. 515, 729 742. Botham, R. P., Beadle, D. J., Hart, R. J., Potter, C. and Wilson, R. G. (1978). Synaptic vesicle depletion and glutamate uptake in a nerve-muscle preparation of the locust, Locusta migratoria L. E.xT~erientia 34, 209 210. Breer, H. (1981). Comparative studies on cholinergic activities in the central nervous system of Locuxta migratoria. .I. Contp. Phvs'iol. 141,271 275. Breer, H. (1982). Uptake of [N-Me-3H]-choline by synaptosomes fiom the central nervous system of Locusta migratoria..I. Neurohio/. 13, 107 117. Breer, H. (1983). Choline transport by synaptosomal membrane vesicles isolated from insect nervous tissue. FEBS Lett. 153, 345 348. Breer, H. and Heilgenberg, H. (1985). Neurochemistry of GABAergic activities in the central nervous system of Locu,sta nligraloria..1. Coral). Physiol. A 157, 343 354. Breer, H., M~idler, U. and Heilgenberg, H. (1989). hnmunocytochemical localization of glutamate decarboxylase in the mesothoracic ganglion of Locusta m~,ratoria. Tiss. ('ell 21, 545-558. Breer, H. and Knipper, M. (1990). Regulation of high affinity choline uptake. Y. Neurobio/. 21, 269 275. Brodbeck, D., Amherd, R., Callaerts, P., Hintermann, E., Meyer, U. A. and Affolter, M. (1998). Molecular and biochemical characterization of the aaN AT 1 (Dat) locus in Drosophihl Me/anogaster: differential expression of two gene products. DNA Cell Biol. 17, 621 633. Brookes, S. J. H. and Weevers, R. G. (1988). Unpaired median neurons in a lepidopteran larva (Antheraea pernyi). 1. Anatomy and physiology. J. K~-/). Biol. 136, 311 332.

132

S. CAVENEY AND B, C. DONLY

Brown, C. S. and Nestler, C, (1985). Catecholamines and indolalkylamines. In: Comprehensive Insect Physiology, Biochemisto, and Pharmacology, Vol. 11 (G. A. Kerkut and L. I. Gilbert, eds), pp. 436 497. Pergamon Press, Oxford. Bruns, D., Engert, F. and Lux, H-D. (1993). A fast activating presynaptic current during serotonergic transmission in identified neurons in Hirudo. Neuron 10, 559572. Buchner, E. (1991). Genes expressed in the adult brain of Drosophih~ and effects of their mutations on behavior: a survey of transmitter- and second messenger-related genes. J. Neurogenetics 7, 153-192. Buchner~ E., Buchner~ S., Crawford, G., Mason, W. T,, Salvaterra, P. M. and Sattelle, D. B. (1986). Choline acetyltransferase-like activity in the brain of Drosophila melanogaster. Cell Tissue Res. 246, 57 62. Buckingham, S. S., Lapied, B., Le Corronc, H., Grolleau, F. and Sattelle, D. B. (1997). lmidacloprid actions on insect neuronal acetylcholine receptors. J. Exp. Biol. 200, 2685 2692. Budnik, V. and White, K. (1988). Catecholamine containing neurons in Drosophila. Distribution and development. J. Comp. Neurol. 268, 400-413. Budnik, V., Martin-Morris, L. and White, K. (1986). Perturbed pattern of catecholamine containing neurons in mutant Drosophila deficient in the enzyme dopa decarboxylase. J. Neurosci. 6, 3682-3691. Burg, M. G., Sarthy, P. V., Koliantz, G. and Pak. W. L. (1993). Genetic and molecular identification of a Drosophila histidine decarboxylase gene required for photoreceptor transmitter synthesis. EMBO .I. 12, 911 919. Burg, M. C., Geng, C., Guan, Y., Koliantz, G. and Pak, W. L. (1996). Drosophila rosA gene, which when mutant causes aberrant photoreceptor oscillation, encodes a novel neurotransmitter homologue. J. Neuroff,enetics 1 l, 59 79. Burrows, M. (1996). The Neurobiology {~['an h?sect Braht. Oxford University Press, Oxford. Callaway, J. C. and Stuart, A. E. (1989). Biochemical and physiological evidence that histamine is the neurotransmitter of barnacle photoreceptors. Vis. Neurosci. 3, 311 325. Callec, J. J. (1985). Synaptic transmission in the central nervous system. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology, Vol. 5 (G. A. Kerkut and L. I. Gilbert, eds), pp. 139 179. Pergamon Press, Oxford. Campos-Ortega, J.A. (1974). Autoradiographic localization of 3H-v-aminobutyric acid uptake in the lamina ganglionaris of Musca and Drosophila. Z. Zel(/brsch. 147, 415 431. Cardinaud, B., Coles, J. A., Perrottet, P., Spencer, A. J., Osborne, M. P, and Tsacopoulos, M. (1994). The composition of the interstitial fluid in the retina of the honeybee drone: implications for the supply of substrates of energy metabolism from blood to neurons. Proc. Roy, Soc. B. 257, 49-58. Carlson, A. D. and Evans, P. D. (1986). Inactivation of octopamine in larval firefly light organs by a high-affinity uptake mechanism..I. Lvp. Biol. 122. 369 385. Carlson, S. D. ([987). The ultrastructure of arthropod neuroglia and neuropil. In: Arthropod BrahT. its Evolution, Development, Structure and Functions (A. P. Gupta, ed.), pp. 323 346. New York, Wiley. Carlson S. D. and Saint Marie, R. L. (1990). Structure and function of insect gila. Amtu. Rev. Entomol. 35, 597-621. Casado, M. Bendahan, A., Zafra, F., Danbolt, N. C., Aragon, C., Gimenez, C. and Kanner, B. (1993). Phosphorylation and modulation of brain glutamate transporters by protein kinase C. J. Biol. Chem. 268, 27 313-27 317.

NEUROTRANSMITTER TRANSPORTERS

133

Castagna, M., Shayakul, C., Trotti, D., Sacchi, V. F., Harvey, W. R. and Hediger, M, A. ([ 998). Cloning and characterization of a potassium-coupled amino acid transporter. Proc. Nail Acad. Sci. USA 95, 5395 5400. Caveney, S., McLean. H. M., Watson, I. and Starratt, A. N. (1996). Affinity of an insect Na+-dependent glutamate transporter for plant-derived cyclic substrates, lnsect Biochem. Mol. Biol. 26, 1027-1036. Chase, B. A. and Kankel, D. R. (1987). A genetic analysis of glutamatergic function m Drosol)hila. J, Neurohiol. 18, 15 41. Chaudhry, F. A., Lehre. K. P., Campagne, M., Ottersen, O. P., Danbolt, N. C. and Storm-Mathisen, J. (1995). Glutamate transporters in glial plasma membranes: highly differentiated localizations revealed by quantitative ultrastructural immunocytochemistry. Neuron 15, 711 720. Chem J.-G. and G. Rudnick (2000). Permeation and gating residues in serotonin transporter. Proc. Natl Acad. Sci. USA 97, 1044 1049. Chiu, C'.-S., Ross, L. S., Cohen, B. N., Lester, H. A. and G i l l S. S. (2000). The transporter-like protein inebriated mediates hyperosmotic stimuli through intracellular signalling. J. Exp. Biol. 203, 353l 3546. Clark, J. A. and Amara, S. G. (1994). Stable expression of a neuronal g-aminobutyric acid transporter, GAT-3, in mammalian cells demonstrates unique pharmacological properties and ion dependence. Molec. Pharmacol. 46. 550 557. Clark, J. M., Scott, J. G., Campos, F., and Bloomquist, J. R. (1995). Resistance to avermectins: extent, mechanisms, and management implications. Am~. Rev. Entomol. 40, 130.

Coco, S., Verderio, C., Trotti. D,, Rothstein, J. D., Volterra, A. and Matteoli, M. (1997). N on-synaptic localization of the glutamate transporter EAAC1 in cultured hippocampal neurons. Era'. J. Neurosci. 9, 1902 1910. Conradt, M. and Stoffel. W. (1997). Inhibition of the high-affinity brain glutamate transporter GLAST-1 via direct phosphorylation. J. Nem'ochem. 68, 1244 1251. Cook, B. J. and Holman, G. M. (1979). The action of proctolin and L-glutamate on the visceral muscle of the cockroach Leucophaea maderae. Comp. Biochem. Physiol. 64C. 183 190. Corey, J. L., Davidson, N., Lester, H. A., Brecha. N. and Quick. M. W. (1994a). Protein kinase C modulates the activity of a cloned g-aminobutyric acid transporter expressed in Xenopus oocytes via regulated subcellular redistribution of the transporter. J, Biol. Chem. 269, 14759 14767. Corey, J. L., Quick, M. W., Davidson, N., Lester, H. A. and Guastella, J. (1994b). A cocaine-sensitive Drosophila serotonin transporter: cloning, expression, and electrophysiological characterization. Proc. Natl Acad. Sci. USA 91. 1188-1192. Cupello, A., Gasparetto, B., Mainardi, P. and Vignolo, L. (1993). Effect of protein ki~ase C activators on the uptake ot" GABA by rat brain synaptosomes, k~t..l. Nem'osci. 69, 131-136. Danbolt, N. C. (2001). Glutamate uptake. Pro~,. Neurohiol. 65, 1105. Danbolt, N. C,, Pines, G. and Kanner, B. I. (1990). Purification and reconstruction of the sodium- and potassium-coupled glutamate transport glycoprotein from rat brain. Biochemistry 29, 6734 6740. Danbolt, N. C., Storm-Mathisen, J. and Kanner, B. 1. (1992). A [Na++K '] coupled ~-glutamate transporter purified from rat brain is located in glial cell processes. Nem'oscience 51, 295 310. Daniels, K. K. and Vickroy, T. W. (1999). Reversible activation of glutamate transport in rat brain gila by protein kinasc C and okadaic acid-sensitive phosphoprotein phosphatase. Nem'ochem. Rev. 24. 1017 1025.

134

S. CAVENEY AND B. C. DONLY

Davenport, A. P. and Wright, D. J. (1986). Octopamine distribution in the larvae and adults of two species of moth, Spodoptera littoralis and Mamhu'a sexta. J. Insect Phys'iol. 32, 987-993. Davis, K. E., Straff, D. J., Weinstein, E. A., Bannerman, P. G., Correale, D. M., Rothstein, J. D. and Robinson, M. B. (1998). Multiple signaling pathways regulate cell surface expression and activity of the excitatory amino acid carrier type 1 subtype of glutamate transporter in C6 glioma. J. Neurosci, 18, 2475-2485. Demchyshyn, L. L., Pristupa, Z. B., Sugamori, K. S., Barker, E. L., Blakely, R. D., Wolfgang, W. J., Forte, M. A. and Niznik, H. B. (1994). Cloning, expression, and localization of a chloride-facilitated, cocaine-sensitive serotonin transporter from Drosophila melanogaster. Proc. Natl Acad. Sci. USA 91, 5158-5162. Derouiche, A. and Rauen, T. (1995). Coincidence of L-glutamate/L-aspartate transporter (GLAST) and glutamine synthetase (GS) immunoreactions in retinal glia: evidence for coupling of GLAST and GS in neurotransmitter clearance..1. Neurosci. Res. 42, 131-143. Dolzer, J., Krannich, S., Fischer, K. and Stengl, M. (2001). Oscillations of the transepithelial potential of moth offactory sensilla are influenced by octopamine and serotonin. J. Exp. Biol. 204, 2781 2794. Donly, B. C., Richman, A., Hawkins, E., McLean, H. and Caveney, S. (1997). Molecular cloning and functional expression of an insect high-affinity Na+-depen dent glutamate transporter. Eur. J. Biochem. 248, 535-542. Donly, B. C., Jevnikar, J., McLean, H. and Caveney, S. (2000). Substrate-stereoselectivity of a high-affinity glutamate transporter cloned fi-om the CNS of the cockroach Diploptera punctata. Insect Biochem. Mol. Biol. 30, 369-376. Downer, R. G. H., Hiripi, L. and Juhos, S. (1993). Characterization of the tyraminergic system in the central nervous system of the locust, Locusta migratoria migratoides. Neurochem. Res. 18, 1245-1248. Dowton, M., Kennedy, I. R., and Wang, M. C. (1988). Localization of glutamine synthetase in fleshfly flight muscle, lnsect Biochem. 18, 717 727. Duan, S., Anderson, C. M., Stein, B. A. and Swanson, R. A. (1999). Glutamate induces rapid upregulation of astrocyte glutamate transport and cell-surface expression of GLAST. J. Neurosci. 19, 10 193-10200. Durvascula, R. V., Gumbs, A., Panackal, A., Kruglov, O,, Aksoy, S., Merrifield, R. B., Richards, F. F, and Beard, C. B. (1997). Prevention of insect-borne disease: an approach using transgenic bacteria. Proc. Natl Acad. Sci. USA 94, 3274-3278. Eldefrawi, A. T. (1985). Acetylcholinesterases and anticholinesterases. In: Comprehensive Inse~ t Physiolo,gy, Biochemistry and Pharmacology Vo[. 12 (G. A. Kerkut and L. 1. Gilbert, eds), pp. 115 130. Pergamon Press, Oxford. Elias, M. S. and Ewms, P. D. (1983a). Autoradiographic localization of )H-histamine accumulation by the visual system of the locust. Ce// Tissue Res. 238, 105 112. Elias, M. S. and Evans, P. D. (1983b). Histamine in the insect nervous system: distribution, synthesis and metabolism. J. Neurochem. 41, 562-568. El Mestikawy, S., Giros, B., Pohl, M., Hamon, M., Kingsmore, S. F., Seldin, M. F. and Carom M. G. (1994). Characterization of an atypical member of the Na+/C1 dependent transporter family: chromosomal localization and distribution in GABAergic and glutamatergic neurons in the rat brain. J. Neurochem. 62, 445-455. Emery, A. M., Billingsley, P. F., Ready, P. D. and Djamgoz, M. B. A. (1998). Insect Na+/K + ATPase. J. l , sect Physiol. 44, 197-209. Evans, P. D. (1974). An autoradiographical study of the localization of the uptake of glutamate by the peripheral nerves of the crab, Carcinus maenas. J. Cell Sci. 14, 351 367.

NEUROTRANSMITTER TRANSPORTERS

135

Evans, P. D. (1975). The uptake of k-glutamate by the central nervous system of the cockroach, Periplaneta americana. J. Exp. Biol. 62, 55 67. Evans, P. D. (1978). Octopamine: a high-affinity uptake mechanism in the nervous system of the cockroach. J. Neurochem. 30, 1015-1022. Evans, P. D. (1980). Biogenic amines in the insect nervous system. Adv. hzsect Physiol. 15, 317473. Evans, P. D. (1985). Octopamine. In: Comprehensive Insect Physiolox,y, Biochemistry amt Pharmacology, Vol. 11 (G. A. Kerkut L. 1. and Gilbert. eds), pp. 499-530. Pergamon Press, Oxford. Evans, P. D., Soderlund, D. M. and Aldrich, J. R. (1980). hi vitro N-acetylation of biogenic amines by tissues of the European corn borer, Ostrinia m~hilalis Huebner. h~scct Biochem. 10, 375 380 Faeder, I. R. and Salpeter, M. M. (1970). Glutamate uptake by a stimulated insect nerve muscle preparation. J. Cell Biol. 46, 300 307. Faeder, 1. R., Matthews, J. A. and Salpeter, M. M. (1974). [3H]Glutamate uptake at insect neuromuscular junctions: effect of chlorpromazine. Brai, Res. 80, 53 70. Fairman, W. A., Vandenberg, R, J., Arriza, J, k., Kavanaugh, M. P. and Amara, S. G. (1995). An excitatory amino-acid transporter with properties of a ligand-gated chloride channel. Nature 375, 599 603. Featherstone, D. E., Rushton, E. M.. Hilderbrand-Chae, M., Phillips, A. M., Jackson. F. R. and Broadie, K. (2{)00). Presynaptic glutamic acid decarboxylase is required for induction of the postsynaptic receptor field at a glutamatergic synapse. Neuro, 27, 71 84. ffrench-Constant, R. H. and Roush, R. T. (1991). Gene mapping and cross-resistance in cyclodiene insecticide-resistant Drosophihl mehmo~a.~ter (Mg.). Gem't. Res. S7, 121. Fillenz. M. (1995). Physiological release of excitatory amino acids. Behav. Brai, Res. 71, 51 67. Flanagan, T. R. J. and Berlind, A. (1984). Serotonin modulation of the release of sequestered [3H]serotonin from nerve terminals in an insect neurohaemal organ i, vitro. Brain Res. 306, 243-250. Fournier, D., Bride, J.-M., Hoffmann, F. and Karch, F. (1992). Acetylcholine esterase. Two types of modification confer resistance to insecticide. J. Biol. Chem. 267, 14 270 14272. Frontali. N. and Pierantoni, R. (1973). Autoradiographic localization of 3H-GABA in cockroach brain. Comp. Biochem. Physiol. 44, 1369 1372. Furuta, A., Martin, L, J., Lin, C.-L. G., Dykes-Hoberg, M. and Rothstein, J. D. (1997), Cellular and synaptic localization of the neuronal glutamate transporters excitatory amino acid transporter 3 and 4. Nez,'osciem'e 81, 1031 1042. Gabrielsson. B., Robson, T.. Morris. D. and Chung, S. H. (1986). Effects of divalent metal ions on the uptake of glntamate and GABA from synaptosomal fractions. Brai, Res. 884, 218 223. Gadea, A. and Lopez-Colome. A. M. (2001a). Glial transporters for glutamate, glycine and GABA. 1. Glutamate transporters. J. Neurosci. Res. 63, 453 460. Gadea, A. and Lopez-Colome, A. M. (2001b). Glial transporters for glutamate, glycine and GABA. 11. GABA transporters. J. Nem'osci. Res. 63. 461 468. Galambos, R. (1961). A glia-neuronal theory of brain t'unction. Proc. Nail Acad. Sci. 47, 129 136. Gallant. P., Malutan, T., McLean, H., Caveney, S. and Donly, C. (2002). Molecular and functional comparison of dopamine and octopamine transporters cloned fi'om the caterpillar CNS. in prep,

136

S. CAVENEY AND B. C. DONLY

Galli, A., Petersen, M., deBlaquiere, M., Blakely, R. D. and DeFelice, L. J. (1997). Drosophila serotonin transporters have voltage-gated uptake coupled to a serotoningated channel. J. Neurosci. 17, 3401-3411. Gao, X., Caveney, S. and Donly, C. (1999). Molecular cloning and functional comparison of a GABA transporter cloned from the CNS of the cabbage looper Trichoplusia hi, Insect Biochem. Mol. Biol. 29, 609-623. Gardiner, R. B., Ullensvang, K., Danbolt, N. C., Caveney, S. and Donly, B. C. (2002), Cellular distribution of a high-affinity glutamate transporter in the nervous system of the cabbage looper Trichoplusia ni, J. Exp. Biol. 205, 2605-26[3. Gegelashvili, G. and Schousboe, A. (1998). Cellular distribution and kinetic properties of high-affinity glutamate transporters. Brain Res. 45, 233 238. Gegelashvili, G., Civenni, G., Racagni, G., Danbolt, N. C., Schousboe, I. and Schousboe, A. (1996). Glutamate receptor agonists up-regulate glutamate transporter GLAST in astrocytes. Neuroreport 8, 261 265. Gegelashvili, G., Danbolt, N. C. and Schousboe, A. (1997). Neuronal soluble factors differentially regulate the expression of the GLTI and GLAST glutamate transporters in cultured astroglia. J. Neurochem. 69, 2612 2615. Gegelashvili. G., Dehnes, Y., Danbolt, N. C. and Schousboe, A. (2000). The highaffinity glutamate transporters GLTI, GLAST and EAAT4 are regulated via different signalling mechanisms. Neurochem. hit. 37, 162 170. Georghiou, G. P. (1986). The magnitude of the resistance problem. In: Pesticide Resistance: Strategies and Tactics jbr Management (National Academy Press, ed.), pp. 14 43. National Academy Press, Washington, D.C. Giles, D. and Usherwood, P. N. R. (1985). The effect of putative amino acid transmitters on somata isolated from neurons of the locus central nervous system. Comp. Biochem. Physiol. 80C, 231 236. Giros, B., E1 Mestikawy, S., Bertrand, L. and Caron, M. G. (1991). Cloning and functional characterization of a cocaine-sensitive dopamine transporter. FEBS Lett. 295, 149 154. Giros, B., El Mestikawy, S., Godinot, N., Zheng, K., Han, H., Yang-Feng, T. and Caron, M. G. (1992). Cloning, pharmacological characterization, and chromosome assignment of the human dopamine transporter. Molec. Pharmacol. 42, 383 390. Giros, B., Jaber, M., Jones, S. R., Wightman, R. M. and Caron M. G. (1996). Hyperlocomotion and indifference to cocaine and amphetamine in mice lacking the dopamine transporter. Ncm#'e 379, 606 612. Goncalves, P. P. and Carvalho, A. P. (1994). Effect of anions on the uptake and release of y-aminobutyric acid by isolated synaptic plasma membranes. Neurochem. Int. 25, 483-492. Gonzalez, A. M and Uhl, G. R. (1994). 'Choline/orphan VS-2-1/creatine transporter" mRNA is expressed in nervous, renal and gastrointestinal systems. Mol. Brain Res. 23, 266 270. Granderath, S. and Klambt, C. (1999). Glia development in the embryonic CNS of Drosophila. Curr. Opin. Neurobiol. 9, 531 536. Granholm, A. C. E., Price, M. L. and Owen, M. D. (1995). Tyrosine hydroxylase in the cerebral ganglia of the American cockroach (Periplaneta americana L.): an immunohistochemical study. Cell Tissue Res. 282, 49 57. Griffiths. R. (1990). Cysteine sulphinate (CSA) as an excitatory amino acid transmitter candidate in the mammalian nervous system. Progr. Neurobiol. 35, 313-323. Grunewald, M. and Kanner, B. 1. (2000). The accessibility of a novel reentrant loop of the glutamate transporter GLT-1 is restricted by its substrate. J. Biol. Chem. 275, 9684 9689,

NEU ROTRANSMI-I-IER TRANSPORTERS

137

Guastella, J., Nelson, N., Nelson, H,, Czyzyk, L., Keynan, S., Miedel, M. C., Davidson, N., Lester, H. A. and Kanner, B. I. (1990). Cloning and expression of a rat brain GABA transporter. Science 249, 1303 1306. Gundersen, V., Chaudhry, F, A., Bjaalie, J. G., Ottersen, O. P. and Storm-Mathisen, J. (1998). Synaptic vesicular localization and exocytosis of L-aspartate in excitatory nerve terminals: a quantitative immunogold analysis in rat hippocampus. J. Neurosci. 18, 6059-6070. Hardie, R. C. (1987). Is histamine a neurotransmitter in insect photoreceptors? J. Comp. Physiol. A 161,201 213. Hardie, R. C. (1989). A histamine-activated chloride channel involved in neurotransmission at a photoreceptor synapse. Nature 339, 704 706. Harris, J. W. and Woodring, J. (1999), Effects of dietary precursors to biogenic amines on the behavioural response from groups of caged worker bees (Apis mell!/i, ra) to the alarm pheromone component isopentyl acetate. Physiol. E,t. 24, 285 291. Hart, M. K. and Batelle, B.-A. (1991). Histamine: metabolism and release in the LimlduS visual system. Invest. Ophthalmol. Vis. Sci. 32, 11 51. Harzsch, S. and Waloszek, D. (2000). Serotonin-immunoreactive neurons in the ventral nerve cord of Crustacea: a character to study aspects of arthropod phylogeny. Arthr. Struct. Developm. 29, 307 322. Haugeto, ~., Ullensvang, K., Levy, L. M., Chaudhry, F. A., HonorS, T., Nielsen, M., Lehre, K. P. and Danbolt, N. C. (1996). Brain glutamate transporter proteins form homomultimers..I. Biol. Chem. 271, 27 715 27 722. Hayakawa, Y.. Bodnaryk, R. P. and Downer, R. G. (1987). Effect of taurine on excitation-induced changes in brain and hemolymph octopamine levels in Peripla,eta americaHa. Arch. Insect Biochem. Physiol. 5, 129- 137. Hay-Schmidt. A. (2000). The evolution of the serotonergic nervous system. Proc. R. Soc. Lond. B 267, 1071 1079. Hediger, M. A. (1999). Glutamate transporters in kidney and brain. Am. J. Physiol. Renal Physiol. 277, F487 F492. Hediger, M. A. and Welbourne, T. C. (1999). Introduction: Glutamate transport. metabolism, and physiological responses. Am. J. Physiol. Renal Physiol. 277, F477 F479. Hediger, M. A., Coady, M. J., Ikeda, T. S. and Wright, E. M. (1987). Expression cloning and eDNA sequencing of the first Na+/glucose cotransporter. Nature 330, 379 381. Hoffman, B. J., Mezey, E. and Brownstein, M. J. (1991). Cloning of a serotonin transporter affected by antidepressants, Science 254, 579 580. Homberg, U. and Hildebrand, J. G. (1991). Histamine immunoreactive neurons in the midbrain and suboesophageal ganglion of the sphinx moth Mcm(hu'a sexta. J. Comp. Nem'o/. 307, 647 657. Homberg, U., Kingan, T. G. and Hildebrand, J. G. (1987). lmmunocytochemistry of GA BA in the brain and suboesophageal ganglion of the sphinx moth Manch~ca sexta. Cell Tiss. Res. 248, 1 24. Horner, M., Sporhase-Eichmann, U., Helle, J., Venus, B. and Schurmann, F. (1995). The distribution of neurons immunoreactive for fi-tyrosine hydroxylase, dopamine and serotonin in the ventral nerve cord of the cricket, Grvllus himaculatus. Cell Tiss. Res. 280, 583 604. Horschitz, S., Hummerich. R. and Schloss P. (2001). Down-regulation of the rat serotonin transporter upon exposure to a selective serotonin reuptake inhibitor. NeuroReport 12, 2181 2184.

138

S. CAVENEY AND B. C. DONLY

Hosie, A. M., Aronstein, K., Satelle, D. B. and ffrench-Constant, R. H. (1997). Molecular biology of insect neuronal GABA receptors. Tremls Neurosci. 20, 578 582. Howell, P. M. R. and Evans, P. D. (1998). The characterization of presynaptic octopamine receptors modulating octopamine release from an identified neuron in the locust. J. Exp. Biol. 201, 2053 2060. Huber, R,, Smith, K., Delago, A., lsaksson, K. and Kravitz, E. A, (1997). Serotonin and aggressive motivation in crustaceans: altering the decision to retreat. Proc. Nat/ Acad. Sci. USA 94, 5939 5942. Huddart, H. and Oldfield, A. C. (1982). Spontaneous activity of foregut and hindgut visceral muscle of the locust, Locusta m(~ratoria I I. The eft}ct of biogenic amines. Comp. Biochem. Physiol. 73C, 303 31l. Hue, B., Pelhate, M. and Callec, J.-J. (1981). GABA and taurine sensitivity of cockroach giant interneuron synapses: indirect evidence of a GABA uptake mechanism. ,l. Insect Physiol. 27, 357 362. Hue, B., Gabriel, A. and Le Patezour, A. (1982), Autoradiographic localization of [3HIGABA accumulation in the sixth abdominal ganglion of the cockroach Periplaneta americana. J. h,sect Physiol. 28, 753 759. Hunter, C. P. (2000). Gene silencing: shrinking the bhtck box of RNAi. Curr. Bioh~gy 10, R137 RI40. Huxtable, R. J. (1992). Physiological actions of taurine. Physiol. Rev. 72, 101 163. Irving, S. N. and MiNer, T. A. (1980). Aspartate and glutamate as possible neurotransmitters tit the slow and fast neuromuscular junctions of the bodywall muscles of Musca domestica larvae. J. Comp. Phl,siol. 135, 299-314. Irving, S. N., Wilson, R. G. and Osborne, M. P. (1979). Studies on the L-glutamate in insect haemolymph. 11. Distribution and metabolism of radiolabelled amino acids injected into the haemolymph. Physiol. Entomol. 4, 223 230. lto, K.. Urban, J. and Technau, G. M. (1995). Distribution, classification, and development of Drosophih, glial cells in the late embryonic and early larval ventral nerve cord. Rou.v's Arch. Dev. Biol. 204, 284 307. Jackson, F. R., Newby, L. M. and Kulkarni, S. J. (1990). Drosophiht GABAergic systems: sequence and expression of glutamic acid decarboxylase. J. Neurochem. 54, 1068 1/)78. Jan, L. Y. and Jan, Y. N. (1976). L-Glutamate as an excitatory transmitter at the Drosophihl larval neuromuscular junction. J. Physiol. 262, 215-236. Jayanthi, L. D, Apparsundaram, S., Malone, M. D., Ward, E., Miller, D. M., Eppler. M. and Blakely, R. D. (1998). The Caenorhah~fitis elegans gene T23G5.5 encodes an antidepressant and cocaine-sensitive dopamine transporter. Molec. Pharmacol. 54, 601 609. Johansen, J., Halpern, M. E., Johansen, K. M. and Keshishian, H. (1989). Stereotypic morphology of glutamatergic synapses on identified muscle cells of Drosophila hirvae. J. Neurosci. 9, 71(1 725. Johnson, K., Knust, E, and Skaer, H. (1999). bloated tuhuh,s (blot) encodes a Drosophihl member of the neurotransmitter transporter family required for organization of the apical cytocortex. Dev. Biol. 212, 440 454. Jursky, F., Tamura, S., Tamura, A,, Mandiyan, S., Nelson, H. and Nelson, N. (1994). Structure, function and brain localization of neurotransmitter transporters. J. k,2",/~. Biol. 196, 283 295. Kanner, B. 1. and Bendahan, A. (1982). Binding order of substrates to the sodium and potassium ion coupled L-glutamic acid transporter from rat brain. Biochem. 21, 6327-6330.

NEUROTRANSMITTER TRANSPORTERS

139

Kavanaugh, M. P., Bendahan, A., Zerangue, N., Zhang, Y. and Kanner, B. 1. (1997). Mutation of an amino acid residue influencing potassium coupling in the glutamate transporter GLT-1 induces obligate exchange..I. Biol. C/Tern. 272, 1703-1708. Keshet, G. 1., Bendahan, A., Su, H., Mager, S., Lester, H. A. and Kanner, B. I. (1995). Glutamate-101 is critical for the function of the sodium and chloride-coupled GABA transporter GAT-1. FEBS Lett. 371, 39 42. Keynan, S., Suh. Y. J., Kanner, B. 1. and Rudnick, G. (1992). Expression of a cloned yaminobutyric acid transporter in mammalian cells. Biochemixtrr 31, 1974 1979. Kilic, F. and Rudnick G. (2000). Oligomerization of serotonin transporter and its functional consequences. Proc. Nail Acad. Sci. USA 97, 3106 3111. Kilty, J. E., korang, D. and Amara, S. G. (1991). Cloning and expression of a cocainesensitive rat dopamine transporter. Sciem'e 254, 578 579. Kingan, T. G. and Hildebrand, J, G. (1985). g-Aminobutyric acid in the central nervous system of metamorphosing and mature Mamluca se.vta. Insect Biochem. 15, 667 675. Kingan, T. G. and Hishinuma, A. (1987). Transport and metabolism of L-glutamic acid by abdominal ganglia of the hawk moth, Mandm'a sexta. Comp. Biochem. Physiol. 87, 9-14. Kitamoto, T., Wang, W. and Salvaterra, P. M. (1998). Structure and organization of the Drosophila cholinergic locus. J. Biol. Chem. 273, 2706 2713. Kitayama, S., Shimada, S., Yu, H., Markham, L. K., Donovan, D. M. and Uhl, G. R. (1992). Dopamine transporter site-directed mutations differentially alter substrate transport and cocaine binding. Proc. Natl Acad. Sci. USA 89, 7782 7785. Kleinberger-Doron, N. and Kanner, B. 1. (1994). Identitication of tryptophan residues critical for the ['unction and targeting of the y-aminobutyric acid transporter (subtype A). J. Biol. Chem. 269, 3063 3067. Klemm, N. (1972). Monoamine-containing fibres in the foregut and salivary gland of the desert locust Schistocerca y{re~aHa Forskal. ComE. Biochem. Ph.vsiol. 43A, 2(17 211. Klemm, N. (1976). Histochemistry of putative transmitter substances in the insect brain. Pro gr. Nem'ohiol. 7, 99 169. Klemm, N. and Schneider, L. (1975). Selective uptake of indolamine into nervous libers in the brain of the desert locust, Schistocerca ,~regaria Forskal (lnsecta). A fluorescence and electron microscopic investigation. Comp. Biochem. Ph)'siol. 50C, 177 182. Klemm, N., Hustert, R., Cantera, R. and Nfissel, D. R. (1986). Neurons reactive to antibodies against serotonin in the stomatogastric nervous system and the alimentary canal of locusts and crickets (Orthoptera, lnsecta). Nem'oscience 17, 247 261. Knipper, M. and Breer, H. (1986). Regulation of the high affinity choline uptake in locust synaptosomes by adenosine triphosphate. Neurosci. Left. 72, 347 351. Knipper~ M. and Breer, H. (1988). Subtypes of muscarinic receptors in insect nervous system. Comp. Biochem. Physiol. 90C, 275 280. Knipper, M. and Breer, H. (1989). Muscarinic receptors modulating actylcholine release from insect synaptosomes. Comp. Biochem. Physiol. 93C, 287 292. Knipper, M., Kahle, C. and Breer, H. (1991). Purification and reconstitution of the high affinity choline transporter. Biochhn. Bioph)'s. Acta 1065, 107 113. Knipper, M., Kahle, C. and Breer, H. (1992). Regulation of hemicholinium binding sites in isolated nerve terminals..l. Nem'obio/. 23, 163-172. Knipper, M., Krieger, J. and Breer, H. (1989b). Hemicholinum-3 binding sites in the nervous tissue of insects. Neurochem. h~i. 14, 211 215. Knipper, M., Strotmann, J., M/idler, U., Kahle, C. and Breer, H. (1989c). Monoclonal antibodies against the high affinity choline transport system. Neurochem. Int. 14~ 217 222.

140

S. CAVENEY AND B. C. DONLY

Konrad, K. D. and Marsh, J. L. (1987). Developmental expression and spatial distribution o f d o p a decarboxylase in Drosophila. Dev. Biol. 122, 172-185. Kostal, V., Noguchi, H., Shimada, K. and Hayakawa, Y. (1998). Developmental changes in dopamine levels in larvae of the fly Chymono,;a costala: comparison between wild-type and mutant-nondiapause strains..I. Insect Physiol. 44, 605-614. Kostowski, W., Tarchalska-Krynska, B. and Markowska, L. (1975). Aggressive behavior and brain serotonin and catecholamine levels in ants (Formica rul20. Pharmacol. Biochem. Behav. 3, 717 719. Krantz, D. E., Chaudhry, F. A. and Edwards, R. H. (1999). Neurotransmitter transporters. In: Neurotransmitter Release (H. Bellen, ed.), pp. 145 207. Oxford University Press, Oxford. Kreissl. S., Eichmuller, S., Bicker, G., Rapus, J. and Ecker, M. (1994). Octopamine-like immunoreactivity in the brain and suboesophageal ganglion of the honeybee. J. Comp. Neurol. 348, 583-595. Kucharski, R., Ball, E. E., Hayward, D, C. and Maleszka, R. (2000). Molecular cloning and expression analysis of a cDNA encoding a glutamate transporter in the honeybee brain. Gene 242, 399405. Kuhar, M. J. and Murrin, L. C. (1978). Sodium-dependent, high affinity choline uptake. J. Neurochem. 30, 15 21. Kunishima, N., Shimada, Y., Tsuji, T., Yamamoto, M., Kumasaka, T., Nakashini, S,, Jingami, H. and Morikawa, K. (2000). Structural basis of glutamate recognition by a dimeric metabotropic glutamate receptor. Nature 407, 971 977. Kutsukake, M., Komatsu, A., Yamamoto, D. and Ishiwa-Chigusa, S. (2000). A tyramine receptor mutation causes a defective olfactory behaviour in Drosophila me/anogaster. Gene 245, 31 ~42. Law, R. M., Stafford, A. and Quick, M. W. (2000). Functional regulation of GABA transporters by direct tyrosine phosphorylation..I. Biol. Chem. 275, 23 986-23 991. Lehman, H. K,, Murgiuc, C. M. and Hildebrand, J. G. (2000a). Characterization and developmental regulation of octopamine biosynthesis in the CNS of a moth Manduca se.vta. Insect Biochem. Molec. Biol. 30, 377 386. Lehman, H. K., Gilchrist, L. S., Klukas, K. A. and Mesce, K. A. (2000b). Steroid regulation of neurotransmitter synthesis m the CNS of Mamtuca se.vta. Steroid regulation of octopamine expression during metamorphic development of the moth, Manduca se.vta. J. Comp. Neurol. 424, 283 296. Levenbook, L. and Kuhn, J. (1962). Properties and distribution of glutamine synthetase in the southern armyworm, Spodoptera eridania. Biochh, l. Biophys. Acta 65, 219 232. Levy, L. M., Warr, O. and Attwell. D. (1998). Stoichiometry of the glial glutamate transporter GLT-I expressed inducibly in a CHO cell line selected for low endogenous Na+-dependent glutamate uptake. J. Neurosci. 18, 9620 9628. Lill, H. and Nelson, N. (1998). Homologies and family relationships among Na+/CI neurotransmitter transporters. In: Neurotransmitter tran,v)orters (S. G. Amara, ed.), Meth. En-ymol. 296, 425 436. Linn, C. E., Poole, K. R. and Roelofs, W. L. (1994). Studies on biogenic amines and their metabolites in nervous tissue and haemolymph of adult male cabbage looper moths. I. Quantitation of photoperiod changes. Comp. Biochem. Physiol. 108C, 73 85. Littletom J. T. and Ganetzky, B. (2000). Ion channels and synaptic organization: Analysis of the Drosophila genome. Neuron 26, 35-43. Liu, Q.-R., Mandiyan, S., Nelson, H. and Nelson, N. (1992). A family of genes encoding neurotransmitter transporters. Proc. Natl Acad. Sci. USA 89, 6639 6643.

NEUROTRANSMITTER TRANSPORTERS

141

Liu, Q.-R., Lopez-Corcuera, B., Madiyan, S., Nelson, H. and Nelson, N. (1993a). Molecular characterization of four pharmacologically distinct y-aminobutyric acid transporters in mouse brain..1. Biol. Chenz. 268, 2106 2112. kiu, Q.-R., Mandiyan, S., Ldpez-Corcuera, B., Nelson, H. and Nelson, N. (1993b). A rat brain eDNA encoding a neurotransmitter transporter with an unusual structure. FEBS 315, 114-118. Lummis, S. C. R., Galione, A. and Taylor, C. W. (1990). Transmembrane signalling ill insects. Atom. Rev. Entomol. 35, 345-377. Lundell, M. J. and Hirsch, J. (1994a). Temporal and spatial development of serotonin and dopamine neurons in the Drosophila CNS. Dev. Biol. 165, 385 396. kundell, M. J. and Hirsch, .I. (1994b). Regulation of the dopa decarboxylase gene during Drosophila development. Dev. Biochem. 3, 55-86. Lundquist, C. T., Baines, R. A. and Bacon, J. P. (1996). Evidence that histamine is a neurotransmitter in an insect extraocular photoreceptor pathway. J. E.vp. Biol. 199, 1973 1982. Mabjeesh, N. J., Frese, M., Rauem T., Jeserich, G. and Kanner, B. 1. (1992). Neuronal and glial y- aminobutyric acid transporters are distinct proteins. Fed. Era'. Biochem. Soc. 299(1), 99 102. McClung, C. and Hirseh, J. (1999). The trace amine tyramine is essential for sensitization to cocaine in Drosophihl. CurreJlt Biol. 9, 853 860. McLcan, H. and Caveney, S. (1993). Na+-dependent medium-affinity uptake of L-glutamate in the insect epidermis. J. Comp. Physiol. B. 163,297 306. Mager, S., Churl. M., Henry, D. J., Chavkin, C., Hoffinan, B. J., Davidson, N. and Lester, H. A. (1994). Conducting states of a mammalian serotonin transporter. Neuron 12, 845 859. Mager. S., Kleinberger-Dorom N.. Keshet, G. 1., Davidson, N., Kanner, B. 1. and Lester, H. A. (1996). lon binding and permeation at the GABA transporter GATI. d. Neurosci. 16, 5405-5414. Maleszka, R., Helliwell, P. and Kucharski, R. (200(/). Pharmacological interference with glutamate re-uptake impairs long-term memory in the honeybee, Apis me/li/i, ra. Behav. Brai~l Res. 115, 49-53. Malutan, T., McLean, H., Caveney, S. and Donly, C. (2(t02). A high-affinity octopamine transporter cloned from the central nervous system of cabbage looper Trichophtsia m . hr,'eel Biochem. Mol. Biol. 32, 343 357. Marie, H. and Attwell, D. (1999). C-terminal interactions modulate the affinity of GLAST glutamate transporters in salamander retinal glial cells. J. Phys'io/. 520, 393 397. Masson, J., Sagn6, C., Hamon, M. and El Mestikawy, S. (1999). Neurotransmitter transporters in the central nervous system. Pharmacol. Rev. 51,439~464. Maxwell, G. D., Tait. J. F. and Hildebrand, J. G. (1978). Regional synthesis of neurotransmitter candidates in the CNS of the moth Mamhlca .ve.vta. Conl/~. Bioc/lent. Phrsiol. 61C, 109 119. Mayfield, R. D., and Zahniser, N. R. (21101). Dopamine D2 receptor regulation of the dopamine transporter expressed in Xc,m)pus laeris oocytes is voltage-independent. Mol. Pharmacol. 59, 113 121. Mayscr, W., Schloss, P. and Betz, H. (1992). Primary structure and functional expression of a choline transporter expressed in the rat nervous system. FEBS Lett. 305, 31 36. Mbungu, D., Ross, L. S. and Gill, S. S. (1995). Cloning, functional expression, and pharmacology of a GABA transporter from Mcltlrh/ca .se.'cla. Arch. Biochem. Biophys. 318. 489 497.

142

S. CAVENEY AND B. C. DONLY

Meldrum, B. S. (2000). Glutamate as a neurotransmitter in the brain: Review of physiology and pathology. J. Nutrition 130, 1007S 1015S. Menzel, R. and Muller, U. (1996). Learning and memory in honeybees: from behaviour to neural substrates. Annu. Rev. Neurosei. 19, 379M04. Meyer, E.P.. Matute, C., Steit, P. and N/issel, D.R. (1986). Insect optic lobe neurons identifiable with monoclonal antibodies to GABA. His'toehem. 84, 207-216. Miller, J. W., Kleven, D. T., Domim B. A. and Fremeau, R. T. Jr (1997). Cloned sodium- (and chloride-) dependent high-affinity transporters for GABA, glycine, proline, betaine, taurine, and creatine. In: Neurotransmitter Transporters: Structure. Function, and Reguhltion (M. E. A. Reith, ed.), pp. 10l 150. Humana Press, Totowa, New Jersey. MonastiriotL M. (1999). Biogenic amine systems in the fruit fly Drosophila melanogaster. Mierosc. Res. Teeh. 45, 106-121. Monastirioti, M., Gorczyca, M., Eckert, M., Rapus, J., White, K. and Budnik, V. (1995). Octopamine immunoreactivity in the fruit fly Drosophila melanogaster..I. Comp. Neurol. 356, 275 287. Monastirioti, M., Linn, C. E. Jr. and White, K. (1996). Characterization of Drosophihl tvramim, fl-hydroxylase gene and isolation of mutant flies lacking octopamine. J. Neurosci. 16, 3900 3911. Morgan, J. R., Gebhardt, K. A. and Stuart, A. E. (1999). Uptake of precursor and synthesis in a histaminergic photoreceptor. J. Neurosei. 19, 1217-1225. Mullin, C, A., Mason, C. H., Chou, J.-C. and Linderman, J. R. (1991) Phytochemical antagonism of g-aminobutyric acid based resistance in Diabrotica. In: Molecular Mechanisms qflusectieMe Resistance (C. A. Mullin and J. G. Scott, eds), pp. 283 308. American Chemical Society Symposium Series 505, American Chemical Society, Washington D.C. Mulloney, B. and Hall, W. M. (1990). GABA-ergic neurons in the crayfish nervous system: An immunocytochemical census of the segmental ganglia and stomatogastric system. J. Comp. Neurol. 291,383 394. Munir, M., Correale, D. M. and Robinson, M. B. (2000). Substrate induced upregulation of Na+-dependent glutamate transport activity. Neuroehem. Mr. 37, 147-162. Nakahashi, T. (1996). Neuronal ion channels as the target sites of insecticides. Pharmacol. Toxieol. 78, 1 14. Nasevicius, A. and Ekker, S. C. (2000). Effective targeted gene 'knockdown" in zebrafish. Nature Genet. 26, 216-220. Nfissel, D. R. (1986). Serotonin and serotonin-immunoreactive neurons in the nervous system of insects Progr. Neurobiol. 30, 1 85. N/issel, D. R. (1996). Neuropeptides, amines and amino acids in an elementary insect ganglion: functional and chemical anatomy of the unfused abdominal ganglion. Progr. Neurobiol. 48, 325~420. N~ssel, D. R. (1999). Histamine in the brain of insects: a review. Mierosc. Res. 71~eh. 44, 121 136. Nfissel, D. R., Pivola, U. and Panula, P. (1990), Histaminelike immunoreactive neurons innervating putative neurohaemal organs and central neuropil in the thoracoabdominal ganglia of the flies Drosophila and Calliphora. J. Comp. Neurol. 297, 525-536. Nathanson. J. A., Hunnicutt, E. J., Kantham, L. and Scavone, C. (1993), Cocaine as a naturally occurring insecticide. Proc. Nat/Acad. Sei. USA 90, 9645 9648. Neckameyer, W. S. and Cooper, R. L. (1998). GABA transporters in Drosophila mehmogaster. molecular cloning, behaviour and physiology. Invert. Neurosci. 3, 279 294,

NEUROTRANSMITTER TRANSPORTERS

143

Neckameyer, W. S. and Quinn, W.G. (1989). Isolation and characterization of the gene for Drosophila tyrosine hydroxylase. Neuron 2, 1167-1175. Nelson, H., Mandiyan, S. and Nelson, N. (1990). Cloning of human brain GABA transporter. FEBS Lett. 269, 181 184. Nicholls, D. and Attwell, D. (1990). The release and uptake of excitatory amino acids. Trends Phamulcol. Sci. 11,462 -468. Nurnberger, A., Rapus, J.. Eckert, M. and Penzlin, H. (1993). Taurine-like immunoreactivity in octopaminergic neurons of the cockroach, Periploneta americana (k.). Histochenl. 100, 285 292. O'Donnell, M. J. and Spring, J. H. (2000). Modes of control of insect Malpighian tubules: synergism, antagonism, cooperation and autonomous regulation. J. hrsect Phvsiol, 46. 107 117. Okuda, T. and Haga, T. (2000). Functional characterization of the human high-affinity choline transporter. FEBS Lett. 484, 92 97. Okuda, T,, Haga, T., Kanai, Y., Endou. H., Ishihara, T. and Katsura, I. (2000). Identification and characterization of the high-affinity choline transporter. Nature Nem'osci. 3, 120 125, Olivares. L., Aragon, C., Gimenez. C. and Zafra, F. (1997). Analysis of the transmembrane topology of the glycine transporter GLYTI..I. Biol. Chem. 272, 1211 1217. Orchard, 1. (1982). Octopamine in insects: ncurotransmitter, neurohormone and neuromodulator. Can. J. Zool. 60, 659 669. Orchard, 1. (1989). Serotonergic neurohaemal tissue in Rho(h#us prolixus: synthesis. release and uptake of serotonin. J. Insect Physiol. 35, 19 27. Orchard, 1. and Lange, A. (1985). Evidence for octopaminergic modulation of an insect visceral muscle..I. Nem'ochem. 16, 171 181. Orchard, 1., Lange, A. and Brown, B. B. (1992). Tyrosine hydroxylase-like immunereactivity in the ventral nerve cord of the locust (Locusta migratoria), including neurons innervating the saliwlry gland. J. Insect Physiol. 38, 19 27. Orkand, P. M. and Kravitz, E. A. (1971). Localization of the sites of GABA uptake in lobster nerve-muscle preparations. J. Cell. Biol. 19, 75 89. Orrego, F. and Villanueva, S. (1993). The chemical nature of the main central excitatory neurotransmitter: a critical appraisal based on release studies and synaptic vesicle localization. Nem'osci. 56, 539 555. Osborne, R. H. (1996). Insect neurotransmission: neurotransmitters and their receptors. Pharmaco/. Ther. 69. 117 142. Pacholczyk, T.. Blakely, R. D. and Amara, S. D. (1991). Expression cloning of a cocaine- and antidepressant-sensitive hunaan noradrenaline transporter. Nature 350, 350 354. Palacin, M., Estevez, R., Bertran, .I. and Zorzano, A. (1998). Molecular biology of mammalian plasma membrane amino acid transporters. Physiol. Rev. 78, 969 1054. Pantanowitz. S.. Bendahan, A. and Kanner, B. 1. (I 993). Only one of the charged amino acids located m the transmembrane ~-helices of the ?'-aminobutyric acid transporter (subtype A) is essential for its activity. J. Biol. Chem. 268, 3222 3225. Parker. D. (1994). Glutamatergic transmission between antagonistic motor neurons m the locust. J. Comp. Physiol. 175A, 737 748. Pennetta, G.. Wu, M. and Bellen H. J. (I 999). Dissecting the molecular mechanisms of neurotransnritter release in Drosol~hiht. In: Nem'olra,smilter Release (H. Bellen, ed.). pp. 304 351. Oxford University Press, Oxford. Petersen. C. 1. and DeFelice, L. J. (1998). Electrophysiological characterization of the Drosophila serotonin transporter, Biol~h,vs. J. 72 (2 part 2), A407.

144

S. CAVENEY A N D B. C. DONLY

Pfltiger, H.-J., Witten, J. L. and Levine, R. B. (1993). Fate of abdominal ventral unpaired median cells during metamorphosis of the Hawkmoth, Mamtuca sextet. J. Comp. Neurol. 335, 508 522. Pines, G., Zhang, Y. and Kanner, B. 1. (1995). Glutamate 404 is involved in the substrate discrimination of GLT-1, a (Na + + K+)-coupled glutamate transporter from rat brain. J. Biol. 67win. 270, 17093-17097. Pirvola, U., Tuomisto, L., Yamatodani, A. and Panula, P. (1988). Distribution of histamine in the cockroach brain and visual system: an immunocytochemical and biochemical study. J. Comp. Neurol. 276. 514--526. Pitman, R. M. (1985). Nervous system. In: Comprehensive h~sect Physiology. Biochemistry and Pharmacology, Vol. 11 (G. A. Kerkut and L. I. Gilbert, eds), pp. 5-54. Pergamon Press, Oxford. Poitry, S., Poitry-Yamate, C., Ueberfeld, J., Macleish, P. R. and Tsacopolous, M. (2000). Mechanisms of glutamate metabolic signaling in retinal glial (Muller) cells. J. Neurosci. 20, 1809-1821. P6rzgen, P., Park, S. K., Hirsch, J., Sonders, M. S. and Amara, S. G. (2001). The antidepressant-sensitive dopamine transporter in Drosophila melanogaster." A primordial carrier for catecholamines. Molec. Pharmacol. 59, 83 95. Preston, C. R., Sved, J. A. and Engels, W. R. (1996). Flanking duplications and deletions associated with P-induced male recombination in Drosophila. Genetics 144, 1623 1638. Prince, W. T. and Berridge, M. J. (1972). The effects of 5-hydroxytryptamine and cyclic AMP on the potential profile across isolated salivary glands. J, Exp. Biol. 50, 323 333. Prokop, A. (1999). Integrating bits and pieces: synapse structure and function in Drosophila embryos, Cell Tiss. Res. 297, 169-186. Pullikuth, A. K., Ross, L. and Gill, S. S. (1993), A putative plasma membrane neurotransmitter transporter from Manduca sexta. Second Int. Symp. Molec. bisect Science, Abstract 1-29. Radice, A. D. and Lustigman, S. (1996). Cloning and characterization of cDNAs encoding putative glutamate transporters from Caenorhabtfitis elegans and Onchocerca volvuhls. Mol. Biochem. Parasitol. 80, 41-53, Ramamoorthy, S., Bauman, A. L., Moore, K. R., Tan, H., Yang-Feng, T., Chang, A. S., Ganapathy, V. and Blakeley, R. D. (1993). Antidepressant- and cocaine-sensitive human serotonin transporter: Molecular cloning, expression and chromosomal localization. Proc. Nail Acad. Sci. USA 90, 2542 2546. Ramarao, C. S.. Acharya, S. R., Krishnan, K. S. and Kenkare, U. W. (1987). High affinity uptake of L-glutamate and gamma-aminobutyric acid in Drosophila melanogaster. J. Biosci. (Bangalore) l 1, 119 136. Rao, V. L. R., Dogan, A., Todd, K. G., Bowen, K. K., Kim, B.-T., Rothstein, J. D. and Dempsey, R. J. (2001). Antisense knockout of the glial glutamate transporter GLT-I, but not the neuronal glutamate transporter EAACI, exacerbates transient focal cerebral ischemia-induced neuronal damage in the rat brain. J. Neurosci. 21~ 1876 1883. Rasola, A., Galietta, L. J. V., Barone, V., Romeo, G. and Bagnasco, S, (1995). Molecular cloning and functional characterization of a GABA/betaine transporter from human kidney. FEBS Lett. 373, 229 233. Raymond, V., Sattelle, D. B. and Lapied, B. (2000). Co-existence in DUM neurones of two GluC1 channels that differ in their picrotoxin sensitivity. NeuroReport 1 I, 26952700. Reale, V., Hannan, F., Midgley, J. M. and Evans, P. D. (1997). The expression of a cloned Drosophila octopamine/lyramine receptor in Xenopus oocytes. Brain Res. 769, 309 320.

NEUROTRANSMITTER TRANSPORTERS

145

Reichenbach, A., Skatchkov, S. N. and Reichelt, W. (19981. The retina as a model of glial function in the brain, ln: Glial Cells (P. R. Laming, E. Sykova, A. Reichenbach, G. I. 14atton and H. Bauer, eds), pp. 63-82. Cambridge University Press, Cambridge. Reith, M. E. A. (1997). Neurotransmitter transporters: Structure, Function and Regulation. Humana Press, Totowa, New Jersey. Restivo, L. L. and White, K. (1990). Molecular and genetic approaches to neurotransmitter and neuromodulator systems in Drosophila. Adv. b~seet Physiol. 22, 115-219. Rheuben, M. B. and Reese, T. S, (1978). Three-dimensional structure and membrane specializations of moth excitatory neuromuscular synapse. J. Uhras. Res. 65, 95-111. Robinson, S. R., Schousboe, A., Dringen, R., Magistretti, P., Cole, J. and Hertz, L. (1998). Metabolic trafficking between neurons and glia. In: Glial CeiLs (P. R. Laming, E, Sykova, A. Reichenbach, G, I. Hatton and H. Bauer, eds), pp. 83-106. Cambridge University Press, Cambridge. Roeder, T. (1994). Biogenic amines and their receptors in insects. ('Otnl). Biochem. P/u,siol. 107C, 1 12. Roeder, T. (1999). Octopamine in invertebrates. Progr. Neurobiol. 59. 533 561. Roeder, T. and Gewecke, M. (1989). Octopamine uptake systems in thoracic ganglia and leg muscles of Loeusta migratoria. Comp. Biochem. Phvsio/. 94C, 143 147. Roman, G. and Davis, R. L. (2001). Molecular biology and anatomy of Drosophila olfactory associative learning. BioEssays 23, 571 581. Rothstein, J. D., Martin, L., Levy, A. I., Dykes-Hoberg, M. Jim L. Wu, D., Nash, N. and Kuncl, R. W. (1995). Localization of neuronal and glial glutamate transporters. Neuron 13, 713-725. Rothstein, J. D., Dykes-Hoberg, M., Pardo, C. A., Bristol, L. A., Jim L., Kuncl, R. W., Kanai, Y., Hediger, M. A., Wang, Y., Schielke, J. P. and Welty, D. F. (1996). Knockout of glutamate transporters reveals a major role for astroglial transport in excitotoxicity and clearance of glutamate. Neuron 16, 675-686. Rudnick, G. (19971. Mechanisms of biogenic amine neurotransmitter transporters. In: Nem'otransmitter Tran,worters: Structure, FunctioH, and Regulation (M. E. A. Reith, ed.), pp. 73-100. Humana Press, Totowa, New Jersey. Salpeter, M. M. and Faeder, 1. R. (1971 ). The role of sheath cells in glutamate uptake by insect nerve muscle preparations. Progr. Brain Res. 34, 103 114. Sandhu, S. K., Ross, L. and Gill, S. S. (1998). Molecular cloning and characterization of a neurotransmitter transporter from Manduea se.\-ta. Third hit. S.vmp. Molec. bisect Science, Abstract 1 13. Sarthy, P. V. (1991). Histamine: a neurotransmitter candidate for Drosophila photoreceptors..1. Neuroehem. 57, 1757 1768. Sasaki, K. and Nagao, T. (2001). Distribution and levels of dopamine and its metabolites in brains of reproductive workers in honeybees. 3. bisect Phrsiol. 47, 12051216. Sattelle, D. B. (1985). Acetylcholine receptors. In: Comprehensive Insect Physiologr, Bioehemisn 3' and Pharmacology, Vol. 11 (G. A. Kerkut and L. I. Gilbert, eds), pp. 395 434. Oxford: Pergamon Press, Oxford. Sattelle, D. 13. (19901. GABA receptors of insects. Adv. bTseet Ph),siol. 22, 11 13. Sattelle, D. B. (1992). Receptors for L-glutamate and GABA in the nervous system of an insect (Periphowta americana). Comp. Biochem. Physiol. 103C, 429 438. Sattelle, D. B. and Breer, H. (19901. Cholinergic nerve terminals in the central nervous system of insects. J. Neuroendocrim)l. 2, 241 256. Sattelle, D. B., Harrison, J. B., Chen, H. H., Bai, D. and Takeda, M. (2000). llnmunocytochemical localization of g-aminobutyric acid receptor subunits in the head ganglia of Periplaneta americam~ using an anti-RDL C-terminal antibody. Neurosci. Lett. 289, 197 20(/.

146

S. CAVENEY AND B. C. DONLY

Sattelle, D. B., Harrow, 1. D., Hue, B., Pelhate, M., Gepner, J. I. and Hall, L. M. (1985). ot-Bungarotoxin blocks excitatory synaptic transmission between cercal sensory neurons and giant interneuron 2 of the cockroach Periplaneta americatta. J. E.vp. Biol. 107, 473 489. Saudou, F., Amlaiky, N.. Plassat, J. J.+ Borelli, E. and Hen, R. (1990). Cloning and characterization of a Drosophila tyramine receptor. EMBO J. 9, 3611-3617. Scavone, C., McKee, M. and Nathanson, J. A. (1994). Monoamine uptake in insect synaptosomal preparations. Insect Biochem. Molec. Biol. 24, 589-597. Schlag, B. D., Vondrasek, J. R., Munir, M., Kelandadze, A., Zelenaia, O. A., Rothstein, J. D. and Robinson, M. B. (1998). Regulation of the glial Na+-dependent glutamate transporters by cyclic AMP analogs and neurons. Molee. Pharnmcol. 53, 355 369. Schloss, P. and Mayser, W. (1994). The putative rat choline transporter CHOT1 transports creatinine and is highly expressed in neural and muscle cells. Bioehem. Biophys. Res. Commun. 198, 637 645. Schuldiner, S. (1997). Vesicular neurotransmitter transporters. In: NeurotransmiHer Tran,v~orters. Structure, Function and Reguk~tion (M. E. A. Reith, ed.), pp. 215240. Humana Press, Totowa, New Jersey. Scott, J. A., Johnson, T. A. and Knowles, C. (1985). Biogenic amine uptake by nerve cords fiom the American cockroach and the influence of amidines on amine uptake and release. Comp. Biochem. Physiol. 8, 43~,7. Scott, R. H. and Duce, 1. R. (1985). Effects of 21,23-dihydroavermectin Bj~, on locust (Sehistoeerea gregaria) muscles may involve several sites of action. Pestic. Sci. 16, 599 604. Seal, R. P. and Amara, S. (1999). Excitatory amino acid transporters: a family in flux. Annu. Rev. Pharmacol. Toxieol. 39, 431 456. Seal, R. P., Daniels, G. M., Wolfgang, W. J., Forte, M. and Amara, S. (1998). Identification and characterization of a cDNA encoding a neuronal glutamate transporter from Drosophila me/anogaster. Receptors and CTtanne& 6, 51 64. Seal, R. P., Leighton. B. H. and Amara, S. G. (2000). A model for the topology of excitatory amino acid transporters determined by the extracellular accessibility of substituted cysteines. Neuron 25, 695 706. Shafqat, S., Valez-Faircloth, M., Guadano-Ferraz, A. and Fremeau, R. T. (1993). Molecular characterization of neurotransmitter transporters. Mo&c. EHdoerino/. 7, 1517 1529. Shaw, P. J., Cirelli0 C,, Greenspan, R. J, and Tortoni, G. (2000). Correlates of sleep and waking in Drosophila melanogaster. Science 287, 1834-1837. Shimada, S., Kitayama, S., Lin, C.-L., Patel, A., Nanthakumar, E., Gregor, P., Kuhar, M. and Uhl, G. (1991). Cloning and expression of a cocaine-sensitive dopamine transporter complementary DNA. Science 254, 576-578. Sieghart, W. (1995). Structure and pharmacology of g-aminobutyric AcidA receptor subtypes. Pharmaeol. Rev. 43, 181 234. Simantov, R., Liu, W., Broutman+ G. and Baudry, M. (1999). Antisense knockdown of glutamate transporters alters the subfield selectivity of kainate-induced cell death in rat hippocampal slice cultures. J. Neuroehem. 73, 1828-1835. Simmons, P. J. and Hardie, R. C. (1989). Evidence that histamine is a neurotransmitter of photoreceptors in the locust ocellus. J. Exp. Biol. 138, 205 219. Sinclair, C. J., Chi, K. D, Subramanian, V., Ward, K. L. and Green, R. M. (2000). Functional expression of a high affinity mammalian hepatic choline/organic cation transporter. J. Lipid Res. 41, 1841 1848. Slotboom, D. J.+ Konings, W. N. and Lolkema, J. S. (1999). Structural features of the glutamate transporter family. Microhio/. Molee. Biol. Rev, 63, 293-307.

NEUROTRANSMITTER TRANSPORTERS

147

Sneddon, L. U.. Taylor, A. C., Huntingford, F. A. and Watson, D.G. (2000). Agonistic behaviour and biogenic amines in shore crabs Carcimts maenas. J. E.vp. Biol. 2113. 537 545. Soehnge, H., Huang, X., Becket, M., Whitley, P., Conover, D. and Stern, M. (1996). A neurotransmitter transporter encoded by the Drosophila inebriated gene. Proc. Natl Acad. Sci. USA 93, 13262-13267. Sonders, M. and Amara. S, G. (1996). Channels in transporters. Curr. Opin. Neurohiol. 6, 4631 4635. Spirodon, M., Kamm, D., Billups, B., Mobbs, P. and Attwell, D. (1998). Modulation by zinc of glutamate transporters in glial cells and cones isolated from the tiger salamander retina. J. Physiol. 506, 363-376. Stern, M. and Ganetzky, B. (1992). [dentification and characterization of i~wbriated, a gene affecting neuronal excitability in Drosophila. d. Nem'ogetwtics 8, 157 182. Stevenson, P. A. (1999). Colocalisation of taurine with transmitter-immunoreactivities in the nervous system of the migratory locust. J. Comp. Neurol. 4114, 86 96. Stevenson, P. A. and Sp6rhase-Eichmann, U. (1995), Localization of octopaminergic neurons in insects. Conq). Biochem. Physiol. I lOA, 203 215. Stuart, A. E. (1999). From fruit flies to barnacles, histamine is the neurotransmitter of arthropod photoreceptors. Neuron 22, 431M33. Stuart, A. E., Morgan, J. R.. Mekeel, H., Kempter, E. and Callaway, J. E. (1996). Selective, activity-dependent uptake of histamine into an arthropod photoreceptor. J. Neurosci. 16, 3178 3188. Sussmann, Harel, M., Frolow, F., Oefner, C., Goldman, A., Toker, L. and Silmam F. (1991). Atomic structure of acetylcholinesterase from Toqw~h; calf/brnica: a prototypic acetycholine-binding protein. Sciem'e 253, 872 879. Suzdak, P. D. (1993) Lipophilic GABA uptake inhibitors: biochemistry, pharmacology and therapeutic potential. Drug,s q/ the Future 18, 1129-1136. Takahashi, M., Billups, B., Rossi, D., Sarantis, M.. Hamann, M. and Attwell, D. (1997). The role of glutamate transporters in glutamate homeostasis in the brain. J. Exp. Biol. 200, 401M09. Takeuchi, A. and Takeucbi, N. (1964). The effect on crayfish muscle of iontopboretically applied glutamate. J. Physiol (Lond.) 170, 296-317. Tanaka, K.. Scott, J. G. and Mutsumura, F. (1984). Picrotoxinin receptor in the central nervous system of the American cockroach: its role in the action of cyclodiene-type insecticides. Pesticide Biochem. Physiol. 22, 117 127. Tareilus. E., Hanke, W. and Breer, H. ([990). Neuronal acetylcholine receptor channels from insects: a comparative electrophysiological study. J. Comp. Physiol. A 167, 521 526. Tatsumi, M. Groshan, K., Blakely. R. D. and Richelson, E. (1997). Pharmacological profiles of antidepressants and related compounds at human monoamine transporters, Eur. J. Pharmacol. 3411, 249 258. Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. and Higgins. D. G. (1997). The ClustralX Windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 24, 4876 4882. Thorens, B. (1996). Glucose transporters in the regulation of intestinal, renal and liver glucose fluxes. Am, J. Physiol. 2711 (Gastrointest. Liver Physiol. 33) G541-G553. Tomlin. E., McLean, H. and Caveney, S. (1993). Active accumulation of glutamate and aspartate by insect epidermal cells, bisect Biochem. Molec. Biol. 23, 561 569. Trimmer, B. A., Aprille, J. R., Dudzinski, D. M., Lagace, C. J., Lewis, S. M., Michel, T., Qazi, S. and Zayas, R. M. (2001). Nitric oxide and the control of firefly flashing. Sciem'e 292, 2486 2488.

148

S. CAVENEY AND B. C. DONLY

Turk, E. and Wright, E. M. (1997). Membrane topology motifs in the SGLT cotransporter family. J. Membrane Biology 159, 1 20. Uhl, G. R. and Hartig, P. R. (1992). Transporter explosion: update on uptake. Trends Neurosci. 13, 421-425. Usherwood. P. N. R. (1994). Insect glutamate receptors. Adv. Insect. Phj,siol. 24, 309341. Uzzan, A. and Dudai, Y. (1982). Aminergic receptors ill Drosophila mehmogaster: Responsiveness of adenylate cyclase to putative neurotransmitters. J. Neurochem. 38, 1542- 1550. Vais, H., Williamson, M. S., Hick, C. A.. Eldursi, N., Devonshire, A. L. and Usherwood, P. N. R. (1997). Functional analysis of a rat sodium channel carrying a mutation for insect knock-down resistance (kdr) to pyrethroids. FEBS Lett. 413, 327 332. Valles, A. M. and White, K. (1986). Development of serotonin-containing neurons in Drosophila mutants unable to synthesize serotonin. J. Neurosei. 6, 1482-1491. Valles, A. M. and White, K. (1988). Serotonin-containing neurons in Drosophila melanogaster. development and distribution. J. Comp. Neurol. 268, 414-425. Van Marle, J., Piek, T., Lind, A. and van Weeren-Kramer, J. (1983). Localization of a Na+-dependent system for glutamate in excitatory neuromuscular junction of the locust Schistoeerca gregaria. Comp. Biochem. Physiol. C 74, 191 194. Van Marie, J. V., Piek, T., Lammertse, Th., Lind, A. and Weeren-Kramer, J. V. (1985). Selectivity of the uptake of glutamate and GABA in two morphologically distinct insect neuromuscular synapses. Brain Res. 348, 107 111, Vanden Broeck, J., Vulsteke, V., Huybrechts, R. and De Loot', A. (1995). Characterization of a cloned locust tyramine receptor cDNA by functional expression in permanently transformed Drosophila $2 cells. J. Neurochem. 64, 2387 2395. Vandenberg, R. J. (1998). Molecular pharmacology and physiology of glutamate transporters in the central nervous system. ('lhT. Exp. Pharm. Physiol. 25, 393-400. Vandenberg, R. J., Mitrovic, A. D. and Johnston, G. A. R. (1998). Molecular basis for differential inhibition of glutamate transporter subtypes by zinc ions. Molec. Pharmaeol. 54, 189~ 196. Venuti, J. M. and Cserjesi, P. (2000). Cell biology for a new millennium. Trends Genet. 16, 156 157. Vickers, N. J., Christensen, T. A. and Hildebrand, J. G. (1998). Integrating behavior with neurobiology: odor-mediated moth flight and olfactory discrimination by glomerular arrays. Inter. Biol. 1,224 230. Viellemaringe, J., Duris, P., Geffard, M., Lemoal, M., Delaage, M., Bensch, C. and Girardie, J. (1984). Immunocytochemical localization of dopamine in the brain of the insect Locusta migratoria mi~ratorioides in comparison with catecholamine distribution determined by histofluorescence technique. Cell Tiss. Res. 237, 391-394. Wadiche, J. 1., Arriza, J. L. Amara, S. G. and Kavanaugh, M. P. (1995). Kinetics of a human glutamate transporter. Neuron 14, 1019-1027. Waffk~rd, K. A. and Sattelle, D.B. (1986). Effects of amino acid neurotransmitter candidates on an identified insect motoneurone. Neurosci. Letl. 63, 135 14(I. Wafford, K. A. and Sattelle, D. B. (1989). L-glutamate receptors on the cell body membrane of an identified insect motor neurone. J. Exp. Biol. 144, 449~462. WaNe, S. and Stoffel, W. (1996). Membrane topology of the high-affinity L-glutamate transporter (GLAST-1) of the central nervous system. J. CWI Biol. 135. 1867 1877. Waldrop, B. and Hildebrand, J. G. (1989). Physiology and pharmacology of acetylcholinergic response of interneurons in the antennal lobes of the moth Manduca sexta. J. Comp. Physiol. A 164, 433-441.

NEUROTRANSMITTER TRANSPORTERS

149

Wang, J. B., Moriwaki, A., and Uhl, G. R. (1995). Dopamine transporter cysteine mutants: Second extracellular loop cysteincs are required for transporter expression. J. Neurochem. 64, 1416 1419. Wang, Y.. Cao, Z., Newkirk, R. F., Ivy, M. T., and Townsel, J. G. (2001). Molecular cloning of a cDNA for a putative choline co-transporter from Limulus eDNA. Ge~e 268, 123 131. Watson, A. H. D. (1988). Antibodies against GABA and glutamate label neurons with morphologically distinct synaptic vesicles in the locust central nervous system. Neuroscience 26, 33M4. Watson, A. H. D. and Seymour-Laurent, K. J. (1993). The distribution of glutamatelike immunoreactivity in the thoracic and abdominal ganglia of the locust (Schistocerca gregaria). Cell Tissue Res. 273, 557 570. Whitton, P. S., Nicholson, R. A. and Strang, R. H. C. (1988). Metabolism of taurine by synaptosomal preparations from an insect. Biochem. Soc. Trans. 14, 609-610. Wierenga, J. M. and Hollingworth, R. M. (1990). Octopamine uptake and metabolism in the insect nervous system. J. Neurochem. 54, 479 489. Witte, 1., Schmidt, M., Gewecke. M. and Roeder, T. (2001). Identification of histaminegated chloride channels in insects, www.biologie.uni-osnabrueck.de/DZG/Neur.pdf Witten. J. L. and Trmnan, J. W. (1991). The regulation of transmitter expression in postembryonic lineages in the moth Manduca se.\-td. 1. Transmitter identification and developmental acquisition of expression..I. Nem'osci. !1, 1980-1989. Witten, J. L. and Truman, J. W. (1998). Distribution of GABA-like immunoreactivc neurons in insects suggests lineage homology. J. Comp. New'o~. 398, 515 528. Worrall, D. M. and Williams, D. C. (1994). Sodium ion-dependent transporters for neurotransmitters: a review of recent developments. Biochem. J. 297, 425 436. Wright, T. R. F. (1987). The genetics of biogenic amine metabolism, sclerotization and melanization in Drosophiht. Adv. Gene:. 24, 127 222. Yamamura, H. 1. and Snyder, S. M. (1972). Choline: high-affinity uptake by rat brain synaptosomes. Sciem'e 178, 626 628. Yu, N., Cao, Y., Mager, S., and kester, H. A. (1998). Topological localization of cysteine 74 in the GABA transporter, GAT1, and its importance in ion binding and permeation. FEBS Lett. 426, 174 178. Zerangue, N. and Kavanaugh, M. P. (1996). Interaction of L-cysteine with a human excitatory amino transporter..I. Physiol. 493(2), 419-423. Zerangue, N., Arriza. J. L.. Amara, S. G. and Kavanaugh, M. P. (1995). Differential modulation of human glutamate transporter subtypes by arachidonic acid. J. Biol. Chem. 270, 6433 6435. Zhang, Y. and Kanner, B. 1. (1999). Two serine residues of the glutamate transporter Gh-I are crucial t\~r coupling the fluxes of sodium and the neurotransmitter. Proc. Natl Acad. Sci. USA 96, 1710-1715. Zhang, Y., Pines, G. and Kanner, B. I. (1994). Histidine 326 is critical for the function of GLT-I, a (Na + + K+)-coupled glutamate transporter from rat brain. J. Biol. Chem. 269, 19 573 19 577. Zhu, K. Y. and Clark, J. M. (1995). Cloning and sequencing of a eDNA encoding acetylcholincsterase in Colorado potato beetle., Leptim)tarsa decemlim,ata (Say). bisect Biochem. Mol. Biol. 25, 1129 1138. Zlotkin, E. (1999). The insect voltage-gated sodium channel as target of insecticides. Atom. Rev. Entomol. 44, 429 456.

Sound Signalling in Orthoptera David J. Robinson ~ and Marion J. Hall Department of Biological Sciences, The Open University, Milton Keynes, MK7 6AA, UK

1 Introduction 152 2 Mating systems 154 2.1 Variation in mating systems 154 2.2 Patterns of calling 157 3 Songs and signals 159 3.1 Intensity, distance and size 160 3.2 Mechanisms of sound production 161 3.3 Pattern generation in crickets 163 3.4 Changes in sound signals with age 165 3.5 Vibratory communication 166 3.6 Sex differences 167 3.7 Temperature effects 167 3.8 The energetic costs of calling 168 4 Hearing and ears 169 4.1 Structure of hearing organs 170 4.2 Auditory receptor organs in the tibia 182 4.3 Tonotopic organization of receptor projections 186 4.4 Directional hearing 187 5 Analysis 189 5.1 Song analysis 190 5.2 Identified auditory interneurons in the nervous system 194 5.3 Symmetry and asymmetry 206 6 Information content of signals 207 6.1 Components of sound signals 209 6.2 Effects of the environment on sound signals 209 6.3 Species recognition 212 6.4 Recognition of sex 215 6.5 Mate location 215 6.6 Mate choice 217 7 Exploitation of heterospecific sounds 224 7.1 Defences against acoustically orienting predators in general 226 7.2 Defences against parasitoids 229 7.3 Defences against bats 234 7.4 Sexual versus nalural selection 241 7.5 Evolution of predator avoidance mechanisms 242 lThis review was completed while DJR was a visiting professor al the University of Nagoya, Japan. ADVANCES

IN I N S E C T P H Y S I O L O G Y

I S B N 0 12-0242~9 X

VOL

29

( ,,pir(E.,M ~ 201i2 EL~m'ier S~icme Lid

.Ill ~i~hf~ ~?f ~c],co&¢cli<m m ~ml li,~m r~.scprcd

152

D.J. ROBINSON AND M. J. HALL

8 Cooperation and competition between males 243 8.1 Spacing, aggregating and fighting 244 8.2 Choruses 247 8.3 Satellite males and silent searching 251 9 New directions 253 Acknowledgements 254 References 255

Abstract The sounds produced by orthopteran insects are very diverse. They are widely studied for the insight they give into acoustic behaviour and the biophysical aspects of sound production and hearing, as well as the transduction of sound to neural signals in the ear and the subsequent processing of information in the central nervous system. The study of sound signalling is a multidisciplinary area of research, with a strong physiological contribution. This review considers recent research in physiology and the links with related areas of acoustic work on the Orthoptera.

1

Introduction

Many orthopteran insects are noisy, a lot are large enough for physiological and neurobiological research, and some at least are of economic significance. So it is not surprising that orthopteran insects are popular with entomologists, and much of the research on insect sound has been carried out using members of this order. Their large size relative to many other insects has enabled detailed studies of the nervous system and the ear, and analysis of hearing at the level of identifiable single neurons has been possible. As information has accumulated about the process of detection and analysis of both intraspecific and interspecific signals, from laboratory studies, it has been possible to carry out more detailed analysis of the behaviour of these animals under natural conditions and this has included recording from identified interneurons in the field. Developments in the study of behavioural ecology, for example, the interest in asymmetry and its role in mate choice, and the evolution of animal signalling, have opened up new areas of research, such as the degree to which the auditory system shows asymmetry at the neuronal level, and the physiological costs of signals. The application of cladistic methods to the study of the evolution of the group has thrown up interesting problems, as the similarities between the auditory systems of members of the group at the neuronal level may not imply a common ancestry. As the process of integrating the results from physiological, behavioural and evolutionary studies gathers pace, now seems an opportune time to review recent advances in the study of the physiology of hearing and sound production in the Orthoptera. This is an extensive field of

SOUND SIGNALLING IN ORTHOPTERA

153

endeavour. In this review, we concentrate on the most recent physiological research and summarize the results and conclusions reported. In addition, we review some recent work in related areas of acoustic behaviour, physiological ecology and the evolution of acoustic systems that highlight the fact that much of the research on the acoustics of orthopteran insects is interdisciplinary. To keep this review to a manageable length, we have often provided references to other reviews that cover much of the earlier work, rather than citing earlier work ourselves. The order Orthoptera is divided into two suborders, the Ensifera, which includes the true crickets (Grylloidea) and the bushcrickets or katydids (Tettigonioidea), and the Caelifera, which includes the grasshoppers and locusts. There are fundamental differences between the ears of the Ensifera and the Caelifera, which means that ears evolved more than once in the Orthoptera. Several unrelated taxa in the Caelifera do not produce sound, and in this group the sound producing organs are not homologous. Thus there must have been multiple origins of sound production in the Caelifera (Riede, 1987: Riede et al., 1990). A recent review of the phylogeny of the Ensifera (Gwynne, 1995) suggests an evolutionary tree that satisfies multiple origins of signalling within the sub-order. He argues that both stridulation and ears have evolved independently in the Grylloidea and the Tettigonioidea, basing his argument in part on the fact that his explanation is more parsimonious than one which suggests a single origin with multiple independent losses of hearing and stridulatory ability. If there had been a single origin, then there would have been four subsequent losses of tibial ears and five losses of tegminal stridulatory organs at the family level, which seems inherently less likely. Furthermore, the raphidophorids are both deaf and mute and are the basal family in the tettigonioid clade. Thus deaf mutes are ancestral and the auditory and signalling systems that we see today in gryllids and tettigoniids evolved separately after the separation of the two groups. There is substantial evidence of homology in the neural elements that make up the auditory systems of the two groups, which might argue for a common origin, although it is possible that the auditory systems, despite evolving separately, have nevertheless incorporated the same, pre-existing neural elements within the nervous system. The homologies between the neural elements of the auditory system are described in detail in section 5.2. The sounds of orthopteran insects are very diverse and often play a crucial role in reproduction, so a substantial amount of research has been targeted on acoustic behaviour during phonotaxis, and the neural bases of sound detection, localization and recognition mechanisms that underlie that behaviour (see, for example, Huber el al., 1989; Bailey and Rentz, 1990). These aspects of sound signalling will be considered in sections 4 and 5. More recently, interest in aspects of mate choice has burgeoned (section 6.6). However, there are other uses for sound communication, such as male to male communication (section 8), as well as unintended consequences, such as being overheard by

154

D.J. ROBINSON AND M. J. HALL

an enemy or rival (section 7), and these have almost certainly exerted an influence on the evolutionary history of the group. The functions of acoustic communication and the selection pressures that have influenced the evolution of acoustic communication systems in Orthoptera have been reviewed by Greenfield (1997) and will be considered in sections 6, 7, and 8. The structure, development and evolution of insect auditory systems has been discussed by Yager (1999), and a recent review of the evolution and function of auditory systems in insects (Stumpner and von Helversen, 2001) compares both the structure of the ear and the soundactivated interneurons in Orthoptera with other insect groups. The study of auditory processing in the Ensifera has been extensive and has demonstrated some surprising similarities in the structure of the neuronal circuits in the central nervous system that are devoted to processing intraspecific signals (Mason and Schildberger, 1993). However, there appear to be differences in the organization of the circuits that are sensitive to ultrasound. The acoustic startle response that is triggered by the ultrasonic signals of predators is elicited via different central neuronal pathways in the true crickets (Gryllidae) and the bushcrickets (Tettigoniidae) (Libersat and Hoy, 1991). The mole crickets (Gryllotalpidae) are the most closely related ensiferan family to the true crickets, but the organization of the ultrasound-sensitive auditory circuitry is different from that in the true crickets (Mason et al., 1998). The songs of mole crickets contain low frequencies only (Bennet-Clark, 1989), and the ability to detect ultrasound is likely to have evolved as a predator-detection mechanism, since many mole crickets fly at night when insectivorous bats are active (Hoy, 1992a; see also section 7.3). Auditory processing will be discussed in detail in section 5.

2 Mating systems 2.1

V A R I A T I O N IN M A T I N G SYSTEMS

In Ensifera, the typical communication pattern between males and females is that males sing loudly and often continuously and the females locate the male by a phonotactic approach. Thus the males are speculative singers and selection tends to favour long and conspicuous song, despite the advertisement this provides to acoustically orienting predators (see section 7). In the tettigoniids (reviewed by Robinson, 1990), the male produces an advertisement call. Some species also produce aggressive calls during male male encounters and courtship songs immediately be(bre copulation (e.g. Meixner and Shaw, 1986). In most species, the female is silent and approaches the male to mate but, in some 'duetting' bushcrickets, the female calls in response to the male and the singing rate of the male is influenced by whether or not a response is received (Robinson, 1980). Female song is restricted to

SOUND SIGNALLING IN ORTHOPTERA

155

three unrelated groups of the Tettigoniidae: it is common in phaneropterines, found in a few species of ephippigerines and may also occur in some pseudophyllines. In some duetting species, such as Platvstolus obvius, the female still approaches the male: in others, such as Leptophyes punctatissima (Kilduff, 2000), the female remains stationary and the male approaches her: in yet others, such as Steropleurus stall and S. nohrei, either sex may perform phonotaxis in response to the other's call (Hartley, 1993). In S. stall, it has been shown that the male approaches a responding female if she fails to perform phonotaxis: the more matings a female has had, the greater the delay before she starts phonotaxis, resulting in the male doing more and more of the approach (Bateman, 2001). Bateman suggests that the costs of phonotaxis are high for the female, and/or the benefits of mating decline sharply after the first time. Among the ephippigerines, there are two patterns of stridulatory behaviour (Hartley, 1993). Some species, such as Ephippi~zer ephippi,ger, call at a high rate and females do not reply but just pert~rm phonotaxis towards the calling male. Other species, such as P. obvius, call sporadically and the female replies; this causes the male to increase his calling rate and the female then performs phonotaxis. The stationary singing female system is common among the phaneropterines. In most duetting species of this type, the male call tends to be short and produced at long intervals, with the female call even shorter; only when a female response is received by the male does he increase his calling rate. In addition, the male will only perform phonotaxis if the female's response falls within a very short time-window after his own call (Heller and yon Helversen, 1986: Robinson et al., 1986). L. punctatissima provides an extreme example of a stationary female system: the male call is only 5 -8 ms long, the female response is only 1 2ms long, and the time-window is only 20 50ms (Hartley and Robinson, 1976; Robinson, 1980: Robinson et al., 1986: Zimmermann el al., 1989). Such short songs are not good acoustic beacons, despite the fact that the "expectation" of a response within a certain time window enhances the ability of the male to detect a female, so the selection pressures that produced them must have acted in opposition to the pressure for fast and accurate phonotaxis. The accuracy of phonotaxis in phaneropterines is thus of evolutionary, as well as physiological interest. The system may have evolved in response to predation pressure (see Robinson, 1990). It should be advantageous to the male especially in the low-density populations found in many duetting species - not to sing unless a female is known to be present, because singing is energetically expensive (see section 3.8) and risks attracting the attention of a predator (see section 7). Mating patterns among the gryllids (reviewed by Boake, 1983; Zuk and Simmons, 1997) are diverse but tend to be characterized by aggression between males, often accompanied by distinctive aggressive chirps, with possession of a burrow and large body size tending to increase the likelihood of success in male interactions. Males in most species do not invest substantially in courtship

156

D.J. ROBINSON AND M. J. HALL

feeding (exceptions are the black-horned tree cricket, Oecanthus nigricornis', and Go, Ilodes supplicans: see Gwynne, 1997) or parental care. As far as mating is concerned, only male crickets sing and they may produce a number of different calls (Alexander, 1962). Males of most species send advertisement signals; the song is delivered from burrows or cracks in the ground, temporary shelters such as leaf litter, tree trunks, or in the open. Females approach calling males and, in many species, once a pair come into contact the male often produces a distinctive courtship song. Males of some species (for example in tree crickets and in the genera Miogo'l/us and Anurogrvllus) also produce a post-copulatory call, which may function in mate-guarding. Most mole crickets (Gryllotalpidae) call, tail outwards, from specially constructed burrows that increase acoustic output (Walker and Figg, 1990). Some species produce a courtship song (Forrest, 1983; Hill, 2000) and/or an aggressive call (Ulagaraj, 1976) in addition to their advertisement call. Hill (1999) has argued that the prairie mole cricket, Grvllotalpa ma/or, demonstrates classic lekking behaviour, in that the males' burrows are spatially aggregated and serve only for sexual advertisement and mating. The Haglidae contains only five extant species, with three in the genus Cvphoderris. All three species of Qvphoderri~" show similar acoustic and mating behaviour in which males call and females approach to mate (see e.g. Dodson et al., 1983; Snedden and Sakaluk, 1992: Mason, 1996). The Anostostomatidae, Stenopelmatidae and Gryllacrididae contain several relic groups of Orthoptera, including Jerusalem crickets, king crickets and wetas (the classification of these families is currently in dispute: see e.g. Johns, 1997: Gorochov, 2001; Hale and Rentz, 2001). Their mating systems are only recently becoming known (for reviews see Field and Jarman, 2001; Hale and Rentz, 2001; Mclntyre, 2001; Monteith and Field, 2001; Toms, 200l; Weissman, 2001). Most anostostomatids do not have ears (Field, 2001b), and gryllacridids appear to have poor sensitivity to airborne sound (Hale and Rentz, 2001). This suggests that intraspecific sound communication is unlikely to play an important role in these groups, although there is some evidence that stridulation may be involved in courtship in some species, such as the Australian giant king cricket Anostostoma australasiae (Monteith and Field, 2001). Instead, many species appear to rely on vibratory communication (e.g. Jerusalem crickets: Weissman, 2001). The New Zealand wetas are a notable exception to this general lack of sound communication. Hemideina crassidens, for example, not only has tympanal organs but they are tuned to the peak frequency of the conspecific song (Field, 2001b). Species of Hemideina (tree wetas) form ~harem" aggregations consisting of one male and several females within a refuge (Morgan-Richards and Gibbs, 2001: Field and Jarman, 2001). Males have greatly enlarged heads and jaws compared with females and use these in intense male-male aggression in defence of the harems (Jamieson et al., 2000; Koning and Jamieson, 2001). H. crassidens is the only species for which a complete repertoire of sounds has

SOUND SIGNALLING IN ORTHOPTERA

157

been established, but observations on other Hemideina species suggest that their acoustic behaviour is similar (Field, 2001b). H. crassidens males produce an advertisement call, an aggressive call in male-male encounters, and a 'mating" call after a failed mating attempt. Some females also produce aggressive stridulations. It has not yet been demonstrated, however, that females are attracted to male sounds in this or any other weta species, and it is not clear whether the "advertisement call" functions to attract females or repel rival males. Unlike other orthopteran species, the temporal pattern of advertisement calling varies enormously both within and between individuals, so much so that H. crassidens cannot be characterized as having a typical call type. It is even possible that each male has a unique pattern. In contrast to tree wetas, giant wetas (Deimicrida) and ground wetas (Hemiandrus) appear to show little or no territorial or aggressive behaviour, although Hemiandrus males may defend their burrows (Field and Jarman, 2001). It is not clear whether communication using sound plays any part in mating (although Hemiandrus may use vibrational signals to attract females). In at least two species of Deinacrida, males are attracted to and follow females, possibly because the latter release pheromones. The African king cricket, Lihanasi~hls vittatus, has also been shown to rely on olfaction for sexual communication; stridulation in this species is used in defence and in aggression towards other males, but appears to have no role in sexual advertisement or mating (Bateman and Toms, 1998a,b). In many grasshoppers, males produce an advertisement call, which may serve in both territorial defence and sexual communication (e.g. Snedden el al., 1998). Some grasshoppers also produce a distinctive aggressive call, used in territorial defence (e.g. Greenfield and Minckley, 1993). In some species, the female approaches the male (e.g. Minckley and Greenfield, 1995), in others, notably in many of the gomphocerine grasshoppers, the female replies to the male's call and he approaches her (e.g. yon Helversen and yon Helversen. 1997). Males may also adopt a silent searching strategy, and the ratio of time spent in acoustic advertisement compared with silent searching varies between species: when a male finds a female he may produce a courtship song (Green, 1995). The diurnal habit and often dense aggregations found in many grasshopper species allow them to make more use of visual signals than do nocturnal orthopterans; some species do not use acoustic signals at all. Although locusts produce sounds, acoustic communication does not appear to play an important role in mating, except possibly at very short distances (Keuper et al., 1985). 2.2

PATTERNS OF CALLING

Some orthopterans call mostly at night (e.g. most bushcrickets), others call mostly during the day (e.g. most grasshoppers), while others call both day and night (e.g. most crickets). Even within these broad categories, however,

158

D.J. ROBINSON AND M. J. HALL

there are enormous differences in the diel pattern of calling. For example, Fischer et al. (1996) recognized six different patterns of calling in 14 species of grassland orthopterans during daylight hours: continuous throughout the day; peaking at noon, peaking in the morning; peaking before and after noon, only in the morning; and only in the evening. They also found that temperature and humidity can have considerable effects on the pattern shown by a species. The variation in diel calling pattern among crickets has been reviewed by Zuk and Simmons (1997) and among nocturnal orthopterans by Walker (1983a). Both reviews argue that the time of calling should be expected to coincide with the period of maximum activity for females, i.e. when calling is reproductively more profitable than other activities. This is so in the mole crickets Scal)teriscus acletus and S. vicinus: females fly during the few hours just after sunset, when predators are relatively few and temperatures are high enough to allow activity, and males call at the same time (Forrest, 1983). Walker (1983a) points out that, in nocturnal species, females mature through the day, so that there is an early evening peak in availability of responsive females: weather and predation also affect female numbers. Cold weather makes the female sluggish and more vulnerable to predation, so there will come a point where the costs of performing phonotaxis outweigh the benefits. Visually hunting predators such as birds are probably the main reason why most crickets and bushcrickets mate at night. This is supported by the fact that species living in open vegetation (trees and bushes) mainly mate at night, while species in dense vegetation may mate during the day as well. Nocturnal predators can modify the pattern of female availability, for example if the predator is effective throughout the night but capture rates are low, females will benefit from the 'selfish herd' effect (Hamilton, 1971) of moving all together, i.e. during early evening. If, however, predators are not less lethal if all females move together, perhaps because they can temporarily specialize in that prey, then it is better for females to move when other females are not moving. If a predator threatens only for a short time, for example after dusk, then the evening peak of availability should be postponed until the earliest safe time. If females mainly perform phonotaxis at times when high numbers of males are calling (thereby enhancing their choice of mates), even higher proportions of males should call during the time of maximum calling, producing a concentration of calling in the temporal analogue of a lek, which Walker calls a spree. As well as availability of females, profitability of calling at any particular time is also influenced by transmission conditions (for example, daytime thermal gradients can create sound shadow zones in which calls are inaudible at distances that they easily reach at night), the male's risk from predators, and his energetic costs (Walker, 1983a). Many species are flexible for example, they sing at night unless the temperature is too cold, in which case they sing during the day (Zuk and Simmons, 1997). The importance of temperature can be seen from the fact that some warm-climate species of crickets and

SOUND SIGNALLING IN ORTHOPTERA

159

bushcrickets show a very different diel pattern of calling compared with that typical of temperate species (Walker, 1983a).

3 Songs and signals The sounds produced by orthopteran insects not only cover a wide range of audible frequencies, but extend well into the ultrasonic range. The size of most species is such that the efficiency of sound radiation is greater at higher frequencies. Size and scale effects in sound communication in insects have been reviewed by Bennet-Clark (1998). If the sound-producing structures are small relative to the wavelength of the sound that they produce, there will be a mismatch in impedance between the insect and the air (Bennet-Clark, 1989: 1995). The sound source should have a diameter that is greater than one-third of the wavelength of the sound produced to obtain a good impedance match (Bennet-Clark, 1999). High-frequency sounds are generated by structures that act as frequency multipliers (Michelsen and Larsen, 1985). A scraper, the plectrum, rubbing along a line of teeth strikes each tooth at a rate that determines the frequency of the sound produced. Thus in Grr/lus campestris, for example, where the sound producing apparatus is on the tegmina (forewings), an area of the wing, called the harp, is the primary resonator, and the harp of the right wing is driven directly from the file of teeth. The harp on the left wing resonates in phase with the harp on the right. The size of the harp determines the resonant frequency, with larger area harps having lower dominant frequencies (Simmons and Ritchie, 1996). The frequency of wing closure is 30 Hz, but on each closure a large number of teeth are struck by the plectrum and the frequency of sound produced is 4.5 to 5 kHz (Bennet-Clark, 1999). Thus the combination of a resonant harp and a repetitive driving force from the tooth strikes produces frequency multiplication. Sound production is discussed further in section 3.2. The terminology used to describe the songs of orthopterans is far from being standardized. Broughton (1963, 1976) proposed a set of terms that has been partially adopted, and Ragge and Reynolds (1998) have developed his terminology. However, it is not used universally and has not been adopted generally outside Europe. So, in our descriptions of recent research we have followed the terminology used by the original authors. The structure of the songs of orthopteran insects is very diverse and in the tettigoniids often highly specialized. The length of songs ranges from under 1 ms in duration to continuous tones lasting minutes. The frequency content ranges from below l kHz to over 100kHz, and the intensity of sound may be as high as 95dB SPL (where SPL is sound pressure level relative to 20 x 1 0 - S N m 2 (20#PA)) at l m . For a detailed account of the properties of the calls of orthopterans, readers should consult the recent review by

160

D.J. ROBINSON AND M. J. HALL

Greenfield (1997). Here, we describe some examples of atypical sound communication of significance in recent research. 3. I

INTENSITY, DISTANCE AND SIZE

Mole crickets produce very loud sounds, achieving the high volume by the use of a second, larger acoustic resonator the burrow that the insect lives in. The burrow is tuned to the same frequency as the primary resonator, the harp, and there is the potential for fine-tuning of the resonant frequency of the burrow. An advantage of the secondary resonator is that it is larger than the primary one, and so more closely matches the impedance of the air. For the mole cricket Gryllotalpa vineae this gives an estimated transduction efficiency of between 17 and 34% (Bennet-Clark, 1970) and the song is audible at a distance of several hundred metres. The bladder grasshopper, Bullacris membracioMes, signals over a very long distance probably the longest distance so far recorded for acoustic communication in insects. From measurements of the threshold of hearing of the female and the decline in sound intensity with distance of the male call, the maximum possible communication distance is estimated at 1.9km (van Staaden and R6mer, 1997). This distance could only be achieved at night: during the day the best distance that could be achieved is 100 150m. The explanation for this is probably to be found in the thermal conditions. The speed of sound in air decreases with decreasing temperature. During the day, temperature decreases with height, and refraction of sound causes sound waves to bend upwards at boundaries of layers of air at different temperatures. At the start of the night, as the ground cools, a temperature inversion occurs (Larom et al., 1997) with cooler air closer to the ground. Under these conditions, sound energy is refracted downwards and there is a channelling effect that reduces the rate of attenuation of the sound. In general, larger male orthopterans produce louder calls and the calls contain lower frequencies. In the tettigoniids, very small species generally have very high frequency calls, for example small phaneropterines where the frequencies are well into the ultrasonic range (Heller, 1984; Robinson, 1990). However, in crickets there appear to have been selection pressures to keep the frequency of the calling song low, and so the sound production mechanism of one of the smallest stridulating gryllids, Qvcloptiloides canariensis, shows adaptations that maintain a low frequency call despite the problems of the small body size. The tegmina are similar in structure, as in other crickets, with the left tegmen a mirror image of the right, but whereas tettigoniids generally stridulate with a left-over-right movement (see the review by Masaki et al., 1987) and field crickets with a right-over-left movement, this cricket has no preferred orientation. The tegmina overlap randomly at rest and during stridulation. The characteristic frequency of the song is between 5.7 and 6.1 kHz. The tegmina are minute, roughly 2.3 mm wide by 1.7 mm long, smaller than the

SOUND SIGNALLING IN ORTHOPTERA

161

mirror cell (see section 3.2) of Gryllus campestris, and the mirror cell is about 25% of the area of the whole tegmen (Dambach and Gras, 1995). It is extremely thin, with a mean thickness of 0.2 #m. For comparison, the mean membrane thickness in G. campestris' is 2.5/xm (Nocke, 1971). The mirror cell in C. canariensis functions as a resonator, in conjunction with both the lateral harp and a strip of cuticle between the mirror cell and file (Dambach and Gras, 1995). The small size of the mirror gives it a low radiating resistance and hence poor sound emission. For such small insects high sound levels could only be maintained if the frequency were higher. However, C. canariensis has not made such a shift and retains a call with a frequency range of other crickets. To maintain resonance at a particular frequency, a smaller membrane must be thinner, and it appears that the thin membrane is an adaptation to the frequency of the song. The intensity of the song of C. canariensis is low (30dB SPL at l m), a consequence of being an insect that is 5 to 7 m m long (Dambach and Gras, 1995) producing a sound with a frequency of 6 kHz and a wavelength of 55 mm. 3.2

M E C H A N I S M S OF S O U N D P R O D U C T I O N

The mechanism by which male crickets produce sound was outlined in the introduction to section 3. The file of teeth on each wing is a thickened Cu2 vein, which forms one side of a triangular wing cell, the harp (for a review of the physics of sound production in crickets see Bennet-Clark, 1999). As each tooth on one wing is struck by the plectrum on the anterior medial end of the other file, the vibrations in the Cu2 vein on each wing cause the harps to vibrate and radiate sound. The frequency of the radiated sound (/c) is given by the rate of tooth impact, known as JPFI (plectrum-to-file impact rate). The rate at which the wing moves clearly should a f f e c t / c , so as temperature affects the rate of wing movement it would be expected that f c too would be dependent upon temperature. In fact, ,/c changes only slightly with temperature. This apparent paradox had been investigated by a number of workers and has produced conflicting views. Elliott and Koch (1985) proposed an escapement model to explain the paradox, known as the 'clockwork cricket'. They envisaged the interaction between the plectrum and the file as a discontinuous process, in which the plectrum is held briefly by a tooth when it comes into contact. The vibration of the harp is in phase with the tooth strikes, so as the harp reverses its direction of movement, the plectrum is released and bounces to the next tooth where it is again held. So it is the resonance of the harp that determines .[evb and thus .[c remains roughly constant (Elliott and Koch, 1985). This interpretation has been criticized (Stephen and Hartley, 1995b), perhaps because the analogy with a clock is less than perfect. For example, if the cricket mechanism was equivalent to a clock escapement mechanism then the song should be characterized by a single frequency, which would be relatively

162

D.J. ROBINSON AND M. J. HALL

constant. In fact the spectrum of Gr),llus bimaculatus, measured by Stephen and Hartley (1995b), contains a peak at 4.75kHz and one at 5.6kHz. Their analysis of the principal frequency component of consecutive calls showed that f c was not constant, having a mean value of 4.65 kHz and a maximum percentage variation of 4.5%. These characteristics do not fit a clock escapement mechanism. The pulse-to-pulse variation in .fc in the cricket Anurogrylh~s arboreus is 1%, still too great to be explained by an escapement mechanism of a clock type (Prestwich et al., 2000). Stephen and Hartley reject the clockwork escapement mechanism, partly on the basis of the frequency evidence. That the escapement analogy might still be useful in describing sound production, providing it is not expected to have the precise functions of a clock mechanism, is an argument put forward by Prestwich et al. (2000). When the tegmina are raised while a cricket is singing there is a partially enclosed air space underneath (Bennet-Clark, 1989: Ewing, 1989). It has been suggested (Stephen and Hartley, 1995b) that singing insects can control the resonant frequency of the subalar air space and h e n c e / c . A comparison of the calls of insects in air and in a mixture of oxygen and helium gases (heliox) showed that ./c had a mean value of 4.86 kHz in air but 5.15 kHz in heliox (Stephen and Hartley, 1995b). When the insects were deafened and the measurements in air repeated, the frequency of the chirps varied widely, suggesting that in the intact insect feedback from the ear might be used to adjust the subalar space and hence J c. Measurements of f c in A. arboreus in both air and heliox also showed an upward shift (Prestwich et al., 2000). Interestingly. in two tree crickets, Oecanthus celeri, ictus and O. quadripunclatus, which do not have subalar spaces, a similar upward shift of,[c was observed in heliox (Prestwich et a/., 2000), which suggests that frequency shifts in heliox have some other explanation. If the harps of the cricket are to function as good resonators, then they need to he damped by a relatively heavy air load (BennetClark, 1989). Calculations by Prestwich et al. (2000) suggest that the frequency shifts observed in heliox are a consequence of a reduction in the system mass as a result of a change in the gas load portion of the system, heliox being 40% as dense as air. In male bushcrickets, there is a file of teeth on the lower surl:ace of the left tegmen that is struck by a plectrum on the upper surface of the right as the tegmina are rapidly opened or closed. The vein that forms the plectrum is attached to an area of the tegmina called the mirror frame, surrounding a thinned area - the mirror. The mirror and mirror frame vibrate as each tooth is struck and radiate sound, at t~ frequency determined by the resonance properties of the cuticle (Bailey, 1970; Bailey and Broughton, 1970). If the ./pJ-i is close to the resonant frequency of the vibrating elements of the tegmen, then a long song of high intensity can be produced, for example the 15 kHz song of Ruspolia nitidula, with the individual tooth impacts not being represented in the song. The song is close to a pure sine wave (Stephen and Hartley, 1995a). However, in most bushcrickets each tooth impact with the plectrum produces

SOUND SIGNALLING IN ORTHOPTERA

163

a set of oscillations that decay rapidly before the next impact, and the tooth impact transients are visible in oscillograms of the song. Sound may be produced on the opening or closing movement of the tegmina, producing one syllable per cycle of wing movement, or on both movements of the tegmina producing either similar syllables or very different syllables. This is well illustrated by a comparative study of nine tettigoniid species (Jatho el a/., 1994). When R. nitidula was induced to sing in a heliox atmosphere, the frequency of the sound fell (Stephen and Hartley, 1995a) and the song became less harmonic, with irregularities that suggest that the tuning of the sound production system had been lost. Heliox does not affect the syllable repetition rate or syllable duration, but does reduce both frequency and intensity (Hartley el al., 2000). It has been suggested that the results from the heliox experiment can be explained by postulating a role for the subalar (subtegminal) space as a load on the tegminal resonator, a proposal discussed earlier in relation to the effect of heliox on gryllid song. Thus a song consisting of a series of brief transients would be amplified and filtered by the air in the subalar space (Hartley cl a/., 2000). However, if this is correct then heliox would be expected to raise the frequency, as observed in experiments on Gampsocleis grafiosa. The explanation offered for the decrease observed in R. nitidula is that the insect is able to regulate the subalar space (Stephen and Hartley, 1995a,b), Clearly we are only part of the way to getting an explanation of the mechanism of sound production in gryllids and tettigoniids, and further work is needed in this rather controversial area. Acridids are distinct amongst animals that produce sound in that they utilize two organs simultaneously to generate sound. Each hindleg when rubbed against the forewing generates a sound, and the sounds produced by the two legs are almost always out of phase with each other. Furthermore, in some species the phase shift is not fixed and can vary such that the leading leg in one song can become the trailing leg in the next (Elsner, 1974: Eisner and Wasser, 1994). Males and females of a particular species have essentially the same song pattern and they stridulate in the same way. The file on the hind femur of the male is homologous with that of the female, although electron micrographs show that the pegs that form the file, although derived from modified bristles in the male, are derived from the surrounding bristle cup in the female (yon Helversen and yon Helversen, 1997). 3.3

PATFERN G E N E R A T I O N IN CRICKETS

In crickets and grasshoppers, stridulatory movements ot" wings or legs are generated by networks of neurons in the thoracic ganglia that produce patterns. Control of stridulatory behaviour is exerted by the brain, as the classical experiments of Huber (1960) have demonstrated, Much of the earlier work is reviewed by Huber et al. (1989). Descending neurons from the brain to

164

D.J. ROBINSON AND M. J. HALL

the thoracic ganglia were identified initially from experiments involving stimulation of fibres in the neck connectives of crickets (Bentley, 1977) and subsequently single descending neurons that could elicit stridulation were identified in the grasshopper Omocestus viridulus (see Hedwig, 1994) and the cricket GrylIus bimaculatus (see Hedwig, 1996). The metathoracic ganglion and the first three abdominal ganglia form a complex in which the species-specific stridulatory pattern is generated (Elsner, 1994). Thus the control of stridulation is separated both in terms of function and structure within the nervous system. Although electrical stimulation of the brain can induce stridulation (Wadepuhl, 1983), localization of regions of the brain involved in the control of stridulation have been identified using pharmacological stimuli. Injection of acetylcholine (ACh) into the medial dorsal neuropile of the protocerebrum reliably elicits continuous stridulation in O. viridulus and Chorthippus mollis (Ocker et al., 1996) and is also effective in crickets (Otto, 1978). Using microelectrodes filled with ACh, Hedwig and Heinrich (1997) were able to show that most locations in the protocerebrum were not involved in stridulation. However, injection of ACh in areas where the descending command neurons are known to arborize (Hedwig, 1994) elicited stridulation reliably. Injection of fluorescent latex particles at the same time as ACh, followed by histological examination (Heinrich et al., 1997) has confirmed that the brain areas in which stridulation is produced most reliably are posterior and dorsal to the central body in the region through which the command fibres pass. When the muscarinic agonist pilocarpine is injected into the same area (Heinrich et al., 1998) long-lasting stridulation can be elicited, lasting many minutes. A subsequent injection of GABA (?,-amino butyric acid) terminates the stridulation, demonstrating that there is a role for inhibitory mechanisms in the control of stridulatory behaviour. Recent work on the control of stridulation in grasshoppers has demonstrated that stridulation is not a single behavioural event, but that normal courtship involves the successive performance of three different stridulatory movements, all under the control of the brain. In O. viridulus, four types of stridulatory behaviour can be elicited by ACh stimulation, the complete sequence of courtship behaviour and three separate components. The separate components have been described as ordinary stridulation, hindleg shaking and precopulatory movements (Hedwig and Heinrich, 1997). Each appears to be under the control of a single command neuron. The picture that emerges from these results is of an area of the brain that initiates stridulation, which is then controlled through the coordinated activity of the descending command neurons. The command neurons are bilateral and elicit the stridulatory patterns in the metathoracic ganglion. At the level of the metathoracic ganglion there is clear evidence that the pattern generation is organized separately for each leg, with the pattern generated in each hemiganglion driving the ipsilateral leg (Ronacher, 1989,

SOUND SIGNALLING IN ORTHOPTERA

165

1991). Bisecting the metathoracic ganglion with a median transection in three species of grasshopper showed that in each one the phase differences between left and right legs were abolished (Heinrich and Eisner, 1997) and both legs then followed the same pattern. A difference was observed when the fast movements of Chorlhit~l~US l)~'gUltldUS w e r e compared with the slow rhythm of stridulation in O. viridullcs. In the fast movements of C. hi,,auttulus, the left and right leg patterns gradually drifted, showing that they were free-running. In contrast, the slow movements of O. virMuhts remained in synchrony. This cannot be achieved through the influence of the brain, since no information about rhythm is carried in the descending command fibres (Hedwig, 1994, 1995). Information about the slow movements is carried to the suboesophageal ganglion by ascending fibres, and within the ganglion the stridulatory rhythm of the two legs is synchronized: this synchronicity is transmitted back to the metathoracic ganglion by descending fibres (Lins and Eisner, 1995). 3.4

( ' H A N G E S IN S O U N D SIGNALS WITIt AGE

The physical structures associated with sound production are susceptible to mechanical wear and, in some individuals where the number of wing closures per second is high, there may be substantial tbrces operating on the cuticle. This will introduce variability into the song patterns in a population. It is of great interest to determine how far such variability is acted on by sexual selection or natural selection, and which properties of the song are under the greatest selection. Furthermore, as populations age, there may be shifts in the pattern of variability of the male song, consequent upon predation of a particular subset of individuals. 3.4.1

CkaJl¢es with aj~e it1 indivi~htals

Age effects on mating success have been demonstrated in one bushcricket species, Et)hil)Piger ephiplfiger, where females discriminate against the song of older males (Ritchie eta/., 1995). In this study the song of males was shown to change over time, as a consequence of changes in the stridulatory apparatus. When presented with simulated songs that reproduced features of the song of aged males and offered the choice of an unaltered song type typical of a young male, females showed a strong preference for the young male song. Degenerative changes in the file structure of bushcrickets occur as a result of abrasion of the teeth on the file or breaking of the teeth (Hartley and Stephen, 1989; Ritchie el al., 1995) and these influence song structure. The song produced by Poeci[imoll sckmidti shortly after ecdysis is a pure tone burst at a frequency ot" 36 kHz (Hartley and Stephen, 1989). After a week of singing activity, the song starts to degrade, with aharmonic components appearing in the spectrum. The syllable length increases, being almost twice as long after live weeks. However, some other studies have not shown a relationship between

166

D.J. ROBINSON AND M. J. HALL

calling properties and age, f o r example in Gwllus penn,u, Ivanicus (Ciceran et al., 1994). 3.4.2

Changes with age in populations

The most comprehensive recent study of age structure and calling song within an ensiferan population is that of Allen (2000), on the Australian bushcricket Sciarasaga qua&ata. Analysis of the song parameters showed that, relative to other males, there was a lifetime consistency in the call properties of an individual. In the population, frequency showed low variability, but there was greater variability in chirp length, interchirp duration and duty cycle. Gaps in the tooth pattern within chirps were recognizable in 13 out of 55 males at the time of collection. Of 15 collected at the start of the season and held in the laboratory, seven showed gaps by the end of the season. This may indicate the effects of wear, but the gaps are not consistently visible in syllables, with typically less than 40% of syllables in an individual's call showing a gap. However, in the population as a whole a shift was observed towards songs with longer chirps as the season progressed, even though there was no association between chirp length and adult life span. Allen suggests that the most credible explanation for this shift lies in the effects of parasitism by the fly Homotri.va alleni. Only 6% of adult males escape parasitism, and it would appear that shorter-chirp males are at greater risk. The songs of early season males collected in October and held in the laboratory were compared with a sample of end of season males collected in January. The surviving males had longer chirp lengths and also longer duty cycles. There was a high positive value for the selection differential for both chirp length and duty cycle, supporting the idea that shorter-chirp individuals are removed selectively from the population. The activities of acoustically orienting parasitoid flies are discussed in more detail in section 7.2. 3.5

VIBRATORY C O M M U N I C A T I O N

Substrate vibrations can be detected by Orthoptera, and a number of examples are known where vibration is related to, or replaces calling song. Rain-forest tettigoniids have been reported as switching to the use of vibratory signalling in areas where bat predation is a risk (Belwood and Morris, 1987; Mason et al., 1991) (see section 8.3). Myopophyllum speciosum is a pseudophylline tettigoniid from the neotropics that uses unusually high ultrasonic frequencies in its calling song, with a mean dominant carrier of 81 kHz (Morris et al., 1994). Vibrational signals with a much shorter range are used in the latter stages of pairing, and this may also be the result of a selective response to bat predation. Tettigonia cantans males perch on plant stems to sing, but through contact between the body and the stem a vibratory signal with a similar temporal pattern is produced, which may be used by the female in the latter stages of

SOUND SIGNALLING IN ORTHOPTERA

167

phonotaxis (Keuper and Kfihne, 1983). Although it has yet to be demonstrated to have biological significance, the song of the mole cricket Gryllotalpa maior has a low-frequency component that travels through the substrate and could be detectable up to 3 m away (Hill and Shadley, 1997). It is possible that this will be shown to have significance in male spacing, which is particularly important in mole crickets, where the burrow represents a sizeable investment for the male. 3.6

SEX DWFERENCES

The song of the male bushcricket Ancistrura ,igrovittata consists of a series of verses, each verse comprising several syllables forming a group and a single syllable after an interval, at the end of the verse. The final single syllable appears to act as a trigger for a female response (Dobler et al., 1994a). All these syllables have the same carrier fi'equency, with peak energy in the range 12- 17 kHz. The female song is a single syllable with maximum sound energy between 20 and 35 kHz (Dobler et al., 1994b). This sex difference in frequency spectrum is unusual and not found in other related species that have been studied. The stridu[atory apparatus of female phaneropterine bushcrickets and also some ephippigerines is not homologous with that of the male (Robinson, 1990), the plectrum being on the lower surface of the left tegmen and the file on the upper surface of the right. Once again, the diverse evolutionary origins of sound communication in orthopterans are highlighted. 3.7

TEMPI~P,A1L RE VFFE('TS

As orthopterans are thermal conformers, temperature can be expected to exert an influence on all the physiological processes associated with sound production. Ambient temperature and heat radiation have been shown to influence the temporal pattern of the male song in tettigoniids of the ephippigerine family (Jatho et al., 1992) and in Orcheliunl (see Walker. 1975). Chirp rate and chirp duration can both change in response to temperature in acoustic interactions between males (Latimer, 1981a,b). In E1)ho~pigerida taeniata the range of temperatures over which chirp-rate and alternation is altered is 19-27 C. The chirp rate increases over this range of temperatures, but then appears to reach a maximum as the rate at 3 5 C is not significantly different from that at 2 7 C (Stiedl et al., 1994). The lower temperature limit for stridulation has been measured (Stiedl and Bickmeyer, 1991) as 15- I T C in Ephippiger ephippi~er. In crickets the pulse rate of the song increases linearly with temperature, and this relationship has been demonstrated in a number of species (Ooherty and Callos, 1991), for example the trilling cricket Gryllus rubens and the chirping cricket G../ulto,i. The effect of temperature on pulse duration in G. rubens was less pronounced than in G../idlo,i. but it still increased with increasing air

168

D.J. ROBINSON AND M. J. HALL

temperature. Frequency did not show a significant link to temperature. Similar results have been reported for G. bimaculatus, where pulse period (reciprocal of pulse rate) was shown to decrease logarithmically with temperature (Doherty, 1985). 3.8

THE ENERGETIC COSTS OF CALLING

Prestwich (1994) has reviewed the energetics of acoustic signalling in insects. The energetic costs of advertisement calling in orthopterans can be very high, requiring on average a ten-fold increase in metabolic rate. In species that call rapidly and loudly, calling can use as much energy as locomotion and can form an important component of the energy budget. Repetition rate, call duration and intensity (SPL) of the call all interact to determine the total energetic cost. Efficiency of sound production (acoustic power/net metabolic power) is low, ranging from 0.05 to 6%, compared, for example, with locomotion, where efficiencies range from about 10 to 20%. Insects producing acoustic advertisement signals have also been shown to have higher mass-independent resting metabolic rates compared with closely related species that do not call (Reinhold, 1999). The energetic costs of calling can be particularly high in species where females tend to choose those males that call for longest, or at the highest rates, or the loudest (Burk, 1988; see also section 6.6). In the variable field cricket, Gryllus lineaticeps, for example, metabolic rate increases with increasing chirp rate and pulse duration and females prefer higher chirp rates (Hoback and Wagner, 1997). Courtship song can be even more expensive than advertisement song: in Acheta domesticus for example, the energetic cost of courtship calling is over twice as much per unit time as the cost of advertisement calling (Hack, 1998). In some species, there is a trade-off between locomotion and calling. Wing dimorphism, in which one morph is long-winged and capable of flight and the other is short-winged and not capable of flight, is common in the Orthoptera (Crnokrak and Roff, 2000), and in Gwllusfirmus, for example, macropterous males call less than micropterous males and as a consequence attract fewer females (Crnokrak and Roff, 1995). Crnokrak and Roff (2000) have shown that this trade-off is an energetic one mediated by lipid stores: the energy required to maintain the flight apparatus reduces the amount of energy available for calling. Male field crickets (Gryllidae) employ alternative strategies of calling or silent searching for females dependent on population density: at higher densities, the payoff in terms of finding mates is greater for silent searching, while at lower densities, the payoff is greater for advertisement calling (Hack, 1998). There may also be a trade-off between calling and energy investment in courtship feeding in some species. For example, Requena verticalis produces a large spermatophylax (Gwynne, 1990), consisting of about 20% of body weight and representing a large energy investment (spermatophore production

SOUND SIGNALLING IN ORTHOPTERA

169

may cost the male an average of 0.22 kJ g-I per day: Simmons et al., 1992). R. verticalis males call with slow syllable rates (Bailey et al., 1993b) compared with other species of bushcricket such as the coneheads of the genus Neoconocephalus, which call at much higher rates but invest negligible amounts in courtship feeding (Greenfield, 1990; Prestwich, 1994). As a result, net metabolic cost of calling (total metabolic cost less resting cost) in R. verticalis is an order of magnitude less than the net cost for e . g . N , robustus (Bailey, 1995). R. verticalis males also control the distribution of energy reserves between calling and spermatophylax production (Bailey, 1995). After mating, they cease calling while a new spermatophore is produced. There is no difference in spermatophylax size between males on high-quality diets compared with those on low-quality diets, but the former are able to resume singing more quickly and sing for longer periods than the latter, indicating that reserves are allocated preferentially to spermatophylax production. There is also evidence in some species that, in populations that are nutrient stressed, males reduce their calling activity, for example in Ephippiger ephippiger (Ritchie et al., 1998) and Gryllus lineaticeps (Wagner and Hoback, [999). In G. [hTeaticeps, this appears to be because males have less energy available for calling. E. ephippiger may call less for the same reason but. because males produce a very large spermatophylax (up to 40% of body weight: Ritchie et al., 1998), and nutrient-deprived females compete for matings with males, an alternative explanation is that this female competition simply allows males to reduce their calling levels without suffering a reduction in mating success.

4

Hearing and ears

Hearing has evolved several times in orthopteran insects and probably in different behavioural contexts. In some grasshoppers the ability to hear is much older than the ability to communicate using acoustic signalling (Riede et al., 1990), and it seems possible that hearing may initially have developed as a response to predation, providing a predator avoidance system (section 7.5). In the Ensifera, hearing probably developed from a communication system based on vibration and may have evolved in parallel with the sound production system for mate recognition and location. There have been later evolutionary changes, of course, For example, secondary loss of hearing is known in both bushcrickets (Lakes-Harlan el al., 1991) and crickets (Otte, 1990). Despite the diverse origins of the auditory systems of the Orthoptera, there are a large number of features that appear to have either a common origin or a parallel development. This commonality is particularly apparent when the neuronal pathways in the segmental ganglia are compared (see section 5.2.3). An interesting question is how hearing systems that have evolved independently turn out to look similar and function in essentially the same way.

170

D.J. ROBINSON AND M. J. HALL

Processing of auditory information is carried out differently in grasshoppers compared with crickets and bushcrickets, which have very similar processing. The fundamental difference lies in the degree of separation of the inputs from the two ears. In bushcrickets and crickets the information from the two ears is processed separately for recognition and then subsequently a comparison is made for determining the location of the source. Thus there is discrimination of quality of song on each side. Furthermore each side has a representation of the loudest sound, improving discrimination when more than one source is present, such as when many males are calling at the same time. In grasshoppers, both directional and recognition information are processed together in parallel (von Helversen and von Helversen, 1995). Information about the patterning of the sound is received in the metathoracic ganglion and there pooled before being transferred to higher processing centres. The combination of information about pattern recognition from the two ears in ascending neurons gives poor directional information. Directional information travels in a separate channel, probably formed by the two phasic ascending neurons AN I and AN2 (see section 5.2.3). Parallel processing has arisen as a result of hearing and localization being older than acoustic communication in grasshoppers, with the result that evolution has overlaid an existing auditory system with adaptations for communication. In the Ensifera, hearing and stridulation probably evolved in parallel, and so from early on there was selection for localization and the detection of species-specific song patterns in the nervous system. However, in the haglids, a group that is probably a relict of the ancestral members of the tettigonioid clade, atympanate auditory reception evolved prior to the tibial tympanal organs. Auditory processing in the nervous system therefore predated the evolution of tympana. There have been several recent reviews of aspects of sound reception and processing. The neural processing of acoustic signals has been considered by Pollack (1998) and the analysis of sound frequency in insects by Pollack and hnaizumi (1999), while Field and Matheson (1998) and Eberl (1999) have reviewed the physiology of insect chordotonal organs. Boyan (1998) has considered the development of insect auditory systems. However, there have also been a number of substantial studies published since these reviews, an indication of the interest that this research area has engendered, and these are described in this section. 4.1

S T R U C T U R E OF H E A R I N G O R G A N S

There are two parameters that are used by insects to detect sound in air. If the distance between transmitter and receiver is short, generally less than one wavelength, then a receptor that responds to the particle velocity of the incident sound can give information about the location of the source, as velocity is a vector. Close to a sound source the relative magnitude of the particle velocity is high, compared with the sound pressure. At distances

SOUND SIGNALLING IN ORTHOPTERA

171

much greater than one wavelength from a source, sound travels as a plane wave, and sound pressure is the predominant parameter. Sound pressure receptors, such as the ears of orthopteran insects, can provide directional information if they are sufficiently far apart that there are measurable differences in either time of arrival at each ear or sound intensity at each ear. Most insects are too small for such a direct measurement to be possible. In the Orthoptera, sound stimuli can act on both sides of the tympanal membranes in the ear and thus it is a pressure gradient that the ear detects. The pathways by which sound reaches the inner side of the membrane introduce a phase shift, which may involve interaural coupling. The general effect is to enhance the difference in membrane excitation between the two ears and so provide directional information. At high frequencies, such as those used by tettigoniids, interaural differences may arise from diffraction around the body, if the wavelength is less than three times the body width, The primary studies of the bioacoustics of orthopteran ears were carried out by Michelsen, starting with the Caelifera (Michelsen, 1971a,b,c) and subsequently the Ensifera (Larsen and Michelsen, 1978; Michelsen and Larsen, 1978; Seymour et al., 1978). In recent years there have been further studies that have expanded this work, but most attention has been tk~cused on the tettigomid ear (reviews by Bailey, 1990, 1993).

4.1.1

The tettigoniid ear

The hearing organs of bushcrickets are located on the tibia of the forelegs just below the joint between the lemur and tibia (Sickmann et al., 1997). Although all three pairs of legs contain tibia[ organs comprising a subgenual organ, an intermediate organ and a crista acustica, only in the foreleg is the tibial organ adapted for sound reception. The single trachea from the prothoracic spiracle passes down the leg and divides in the tibia to give two closely adhered branches (Fig. 1) separated by a central membrane. The outer walls of the trachea are modified to form the tympana, which may be exposed, as in Polysarchus denticauda (Sickmann et al., 1997) or protected within a cavity

FIG. 1 Structure and location of the tibial hearing organ in a bushcricket.

172

D.J. ROBINSON AND M. J. HALL

that communicates with the exterior by longitudinal slits, as in Ruspolia do~ferens (Fig. 1). The original morphological study of the tettigoniid tympanal organ (Schwabe, 1906) did not recognize the acoustic role played by the prothoracic spiracle and the trachea. Since the demonstration of the role of sound entry via the prothoracic spiracle (Lewis, 1974) there has been increasing study of the tettigoniid ear in an effort to understand the sound transduction mechanism. However, while it is agreed that the spiracle is the main source of sound entry (see, for example, Nocke, 1975; Hoffmann and Jatbo, 1995), there is less agreement on the transmission properties of the acoustic trachea. The acoustic trachea approximates to an exponential horn in a number of species where morphological measurements have been made and it has been shown to have a high pass characteristic, imparting a gain of 15 20dB (Hoffmann and Jatho, 1995). A horn of this type will have a broad-band transmission above the cut-off frequency, so it is not the trachea that provides the tuning of the ear. There are several candidates, notably the resonances in structures associated with the receptor organs ( L i n e t al., 1994; Hoffmann and Jatho, 1995). It has also been suggested that the individual sensilla provide frequency selectivity (Oldfield, 1985). The tympana and slits clearly have a role to play in the auditory system but what that role was remained controversial for some years (Nocke, 1975; Stephen and Bailey, 1982: Oldfield, 1984). In the bushcricket Tetti,gonia vh'idissima, the tympana are covered by cuticular covers, but if these are removed it is possible to measure the movements of the tympanum when stimulated with sound, using the method of laser vibrometry (Bangert el al., 1998). A small reflector attached to the tympanum reflects laser light and enables the velocity of the membrane to be measured using Doppler techniques. The addition of the reflector does not affect the measurements (Stumpner and Heller, 1992), but the removal of the covering cuticle to expose the tympana does arguably have some effect (Bailey and Stephen, 1984). When the phase of both of the sound inputs is the same, there is a measurable phase difference between the incident sound pressure and the tympanal velocity over the range of 2 to 40 kHz (Fig, 2). This is a consequence of the additional travel time in the trachea. The gradient of the line in Fig. 2 is related to the length of the trachea and the velocity of sound in the trachea. The mean gradient is constant throughout the frequency range, indicating that the propagation velocity is constant over this range also, at 2 6 0 i 4 0 m s -J, and there is no dispersion inside the trachea (Bangert eta/., 1998). Measurements of the phase spectra at six different points on the tympanic membrane showed that they were all effectively the same, demonstrating that the bushcricket tympanum has a single vibratory mode over the entire frequency range. This is not the case in an acridid, the locust, where the membrane of the ear has been shown to have independently vibrating areas (Michelsen. 1971b; Stephen and Bennet-Clark, 1982). Plotting velocity amplitude measurements over the surface of the tympanic membrane of Polysarcus denticauda showed that the maximum velocity is located at a

SOUND SIGNALLING IN ORTHOPTERA

i

01

173

Polysarchusdenticauda

-180 ~

-360 -540

~

-720 -

Tettigoni

.=m -900 -1080 -

i

-1260 I

I

5

10

I

I

I

15 20 25 Frequency/kHz

I

i

I

30

35

40

FIG. 2 The phase spectrum of the tympanic membrane velocity of two bushcrickets, in relation to the phase of the acting sound pressure. The line of the mean gradient is shown, and that for Tettigonia virMissima has been olivet by 180' to separate it from that of the other species, PoO'sarchus denlicauda (Bangert et al., 1998, fig. 2). Reprinted with permission from Elsevier Science, from Hearipl+ Research 115, Bangert, M., Kalmring, K., Sickmann, T., Stephen, R. O,, Jatho, M. and LakesHarlan, R. (1998). Stimulus transmission in the auditory receptor organs of the Ik)releg of bushcrickets (Tettigoniidae) I. The role of the tympana, 27 38.

position that is both ventral and distal (Bangert et al,, 1998). This is at variance with earlier measurements on two species of Tettigonia (Michelsen and Larsen, 1978), which suggested that the m a x i m u m velocity occurred at the centre of the membrane, but the t y m p a n u m of P. denticauda is a rather different structure to that of Tettigonia, being between five and ten times thicker. The vibration amplitude of the tympanic membrane in P. denticau&l increases linearly from the dorsal edge, reaching a maximum close to the ventral edge (Bangert et al., 1998). This observation suggests that the tympanum is acting like a hinged flap with the hinge along the dorsal edge (Fig. 3). The acoustic function of the hinged tympana has yet to be completely explained. From measurements made of the sensitivity of tibial organs in the midlegs and hindlegs it has been shown that the foreleg tibial organ is 50 dB more sensitive to airborne sound (Kalmring et al., 1994). Calculating the gain produced by the trachea gives a figure of 20dB (Hoffmann and Jatho, 1995), which shows that there must be a further factor involved. It has been suggested (Bangert et al., 1998) that the tympana act together as an impedance matching transformer much in the way that the ossicles of the middle ear do in mammals, matching sound vibration in air to sound vibrations in the denser tissues of the hearing

174

D.J. ROBINSON AND M, J. HALL

organ. Clearly this is a feasible suggestion, but what particular structure of the bushcricket ear requires such an impedance match? There is as yet no answer to this question. The amount of gain provided by the acoustic trachea has been linked to the size of the spiracle. In two species of Poecilimon with different sized spiracles, the species with the smaller spiracle had a tracheal gain of 5 6 dB whereas the species with the larger spiracle had a tracheal gain of 18 20 dB (Michelsen et al., 1994a). These gains are consistent with the dimensions of the acoustic trachea. However, in a comparison of seven species from four families (Heinrich et al., 1993), no differences in gain associated with the size of the spiracle were detected. This is perhaps because the acoustic conditions under which the measurements were made masked real differences (Michelsen et al,, 1994a). The size of the auditory spiracle has been shown to determine the auditory sensitivity of female tettigoniids. In Requena verticalis, half of the

4'

FIG. 3 Schematic cross-section of the tibia of Polysarchus denticauda at the distal part of the crista acustica to show the hinge operation of the tympana and the effect on the trarlslocation of the dorsal wall of the trachea. The dotted line indicates the direction of movement of the rectorial membrane and cap cell. Labels: dw, dorsal wall; hc, haemolymph channel: nmc, nerve muscle channel; s, septum (Bangert et al., 1998, fig. 9). Reprinted with permission from Elsevier Science, from Hearing Research 115, Bangert, M., Kalmring, K., Sickmann, T., Stephen, R. O., Jatho, M. and LakesHarlan, R. (1998). Stimulus transmission in the auditory receptor organs of the foreleg of bushcrickets (Tettigoniidae) 1. The role of the tympana, 27 38.

SOUND SIGNALLING IN ORTHOPTERA

175

variation in sensitivity of different females to the male call could be attributed to the variation in the size of their auditory spiracles (Bailey, 1998). Sexual dimorphism in the auditory system is unusual (Bailey and R6mer, 1991), since males usually need to hear the calls of other males just as much as females do (so that they can, for example, assess, localize and interact acoustically with them; see section 8). Where selection on male hearing sensitivity is relaxed, however, such as when male male competition is reduced, sexual dimorphism can result. It has been found, for example, in some species of the tettigoniid Kawanaphila (see Mason and Bailey, 1998). Mason and Bailey (1998) looked at the auditory systems and acoustic behaviour of K. varraga and K. miHa and compared them with K. nartee, to test whether reduced hearing in males is associated with reduced male male competition. In each species, female hearing is more sensitive than that of males by 10 dB, and male K. mirla are most sensitive at lower frequencies than females. The species also differ in the degree of dimorphism in auditory anatomy. K. mirla and K. nartee males lack some of the auditory specializations of the prothoracic tracheal system normally present in tettigoniids; these are present in K. varraga but reduced in size compared with females. K. ~zartee show very low levels of male male acoustic communication, so it is possible that reduced sensitivity is linked to reduced male male competition in this species. K. mirla and K. varraga males, in contrast, both interact acoustically with conspecific males to synchronize their calls (see section 8.2). However, K. mirla only interact in this way over distances less than 5 m, while K..varra~a interact over distances of at least 10m. Thus the nature of raale male competition in K. mirla and K. 3'arraga does not change with decreased hearing sensitivity, just the distance over which males are able to compete. Mason and Bailey argue that selection has favoured a reduction in male auditory sensitivity independently of the function of male male acoustic signalling. They suggest two possible ecological factors that may have contributed to this. First, if resources are scattered and male mobility is limited, male male communication distances would be determined by resource spacing, but competition could still be maintained within this spacing. Second, if acoustically orienting predators or parasitoids (see section 7) have influenced the evolution of song structure, communication distances in these species may now be limited by their ability to broadcast extremely high-frequency calls. In Kau'aHai~hila , a r t e e it has also been demonstrated that sexual selection acts on the size of the ear in females (Gwynne and Bailey, 1999). In foodstressed populations of K. ~lartee, there are more sexually receptive females than males. The male produces a spermatophore that is a nuptial gift to the female, and the production requires a substantial input of food energy (Simmons and Bailey, 1990). Where food is scarce, the number of males with spermatophores is reduced but, at the same time, females too are short of food. and competition for access to males is increased. In these populations, when the encounter rate with receptive females is high, males reduce their continuous

176

D.J. ROBINSON AND M. J. HALL

calling song to an occasional brief 'click'. As a result, those females with the greatest auditory sensitivity are the most likely to detect a calling male. In an analysis of winners and losers in competitions between females for males, the mean diameter of the auditory spiracle of the winner - the first female to hold on to the male genitally - was significantly larger than that of the loser. The size of the auditory spiracle was measured by two independent methods, one optical and the other using video recordings, and there were no significant differences between the methods. Using one of the methods, the mean diameter of the auditory spiracle of winners was 0.243 4- 0.005 ram, and for the losers the mean diameter was 0.220 4-0.007mm. It might be thought that these differences would be correlated with the size of the individuals. Female size has been shown to be a factor in the m o r m o n cricket (Anabrus simph, x): at high densities~ where females compete for males, larger females tend to be more successful (Gwynne, 1984). However, in K. nartee there was no significant difference in pronotum length, which is a measure that correlates well with size, between winners and losers. Similarly, there was no significant difference in mass (Gwynne and Bailey, 1999). The highest communication frequency reported for an acoustically signalling orthopteran is 105kHz (Morris et al., 1994). Haenschiella ecuadorica is a pseudophylline and the song consists of up to 18 pulses, each less than 1 ms duration, forming a single pulse train of 51 ms duration (Mason et al., 1991: Morris et al, 1994). In pseudophyllines that communicate above 5 0 k H z the acoustic spiracle is reduced, relative to body size, as are the dimensions of the acoustic trachea. The area of the spiracle is also greatly reduced relative to the tympanal slits. At this very high communication frequency, sound entry to the tympanal organ is predominantly via the slits. 4.1.2

The acridid ear

In grasshoppers the tympanic membranes of the auditory organs are located on the first abdominal segment and the two membranes are acoustically coupled via closed air sacs within the abdomen. The air sacs transmit sound from one ear to the other, acting as low-pass filters (Michelsen, 1971c), such that the ear operates as a pressure-gradient receiver at low frequencies, but is mainly a pressure receiver at high frequencies. These biophysical findings of Michelsen, and later work by Miller (1977) were confirmed by R6mer (1976), using neurophysiological techniques. Furthermore, muscle movements in the abdomen can distort the membranes (Meyer and Hedwig, 1995) and changes in air pressure in the tracheal system can alter the volume of the air sacs (Michelsen, 1971c: Meyer and Elsner, 1995). The nature of the link between changes in the tracheal pressure and movement of the tympanum has been investigated in the migratory locust (Meyer and Hedwig, 1995). Simultaneous measurements of the movement of the tympanum and the tracheal pressure were made using a laser interferometer directed at the

SOUND SIGNALLING IN ORTHOPTERA

177

t y m p a n u m and a digital manometer connected via a sealed tube to the mesothoracic spiracle, whose valves had been fixed into the open position. The manometer recorded the variation in tracheal pressure, relative to ambient air pressure. The movements of the tympanal membrane correspond with changes in the tracheal pressure (Fig. 4). Plotting membrane displacement against the pressure at the t y m p a n u m showed that the relationship was only linear over the range +10 to 20 # m at best and the curve flattened out at greater displacements. At increasing displacement, the force on the membrane had to be greater to produce an equivalent deflection the compliance of the membrane decreases rapidly with increasing displacement. The consequence of these properties is that even small displacements of the tympanic membrane resulting from pressure changes produced by breathing movements would be expected to produce modulation of the auditory information and so influence processing. The receptors are attached directly to the inner surface of the t y m p a n u m and the membrane shows a complex vibration pattern with different modes of vibration at the different receptor attachment sites. The vibrations have been analysed using laser vibrometry (Michelsen, 1971b) and demonstrated visually (Stephen and Bennet-Clark, 1982). The receptor attachment sites on the tympanic membrane are shown in Fig. 5. There are four distinct groups that form Miller's organ. Of these the a-, b- and c-cells are sensitive to lower frequencies than the d-cells. However, more recent analysis of the central projections of the auditory receptor cells suggests that the a and b cells are not distinct types and should be grouped together (Jacobs el al., 1999).

Tracheal systqn: pressure change

Tympanal membrane: displacement /~ /'~ A /~

_

1 1500 Pa

- (0 /~Out] 80/-/Ill

2s FIG. 4 Simultaneous measurements of tracheal pressure and tympanal membrane displacement in a normally ventilating locust, showing the close correlation between the pressure and the displacement of the membrane. Outward displacement of the membrane is shown as an upward deflection (Meyer and Hedwig, 1995, fig. IB). Reprinted with permission from the Company of Biologists.

178

D.J. ROBINSON AND M. J. HALL

Ventral

Dorsal 'Ganglion' of MULler'sorgan

z#///tt

¢

e

...

.:

.

.Air . . . .sac cuticle

Folded body

Proces* of

~

- - ~

#

-~

_..

-

~al~,~(

/ e l e v a t e d

A :'=~i,._ /=.)-.-_.'~i~/

.

Thin membrane

/

• --

Fusiform body .

re --

..... , muller s organ

oldora sa

/ Auditory

/.o

c cuticle "

o

/•..-..

-:

,

0-2 mm

.."

'. : x .

Thick membrane

' ..........

~ j ~

body

•

/ Elevated p r o c e s s :. " . . : -:. "

/

FeXody 1~

[

Thin m e m b r a n e

Pyriform vesicle

FIG. 5 Drawings of the tympanal sclerites and Mtiller's organ to show the trajectories of the four groups of sense cells. The points of attachment to the sclerites are shown by small circles and stippling, the approximate length of the scolopale cells is indicated by the length of the arrows and the position of the cell bodies is indicated by the letters A to D. The upper view is from about 10 above the plane of the tympanum. The lower view is normal to the tympanum (Stephen and BennetClark, [982, fig. 4). Reprinted with permission from the Company of Biologists. The vibrational sensitivity of the locust t y m p a n u m has been analysed for the different a t t a c h m e n t sites of the receptors (Meyer and Hedwig, 1995). F r e q u e n c y - m o d u l a t e d s o u n d pulses with a b a n d spread of 1 - 2 5 k H z were used to stimulate the m e m b r a n e . The resulting vibrations, measured with a laser interferometer, were analysed using Fast F o u r i e r T r a n s f o r m ( F F T ) analysis. The shapes of the resulting spectra showed that there were substantial

SOUND SIGNALLING IN ORTHOPTERA

179

differences in vibration between the elevated process (see Fig. 5) and the pyriform vesicle. Between 2 and 9 k H z the vibration patterns at both sites were similar but at higher frequencies differences emerged. At 16 kHz, there was a difference of 26 dB in the amplitude of the oscillations at the two sites, with the pyriform vesicle showing the stronger response. The maximum response from the membrane at the elevated process was observed in the range 3.5-10kHz, wilh peaks at 5 k H z and 8kHz. These figures correspond well with the tuning curves of low frequency auditory receptor cells (a-cells) (Michelsen, 1971 b; R6mer, 1976). The membrane at the pyriform vesicle shows maximum vibrations over the range 3.5 23kHz, with a peak in the range 13 18 kHz. Once again, this corresponds well with previously measured tuning curves (Michelsen, 1971b; R6mer, 1976). Changes in the response of the locust ear to high frequencies have been investigated, following selective damage to the d-cell dendrites, A laser (Fuhr et ell., 1999) was used to ablate a small area of the dendrites or their attachment to the tympanum. High energy pulses from a UV laser could produce a damage area as small as 5 ttra x 5 ttm. Threshold curves were produced for both the damaged and intact ears over a frequency range of 3 28 kHz, and these showed a pronounced asymmetry, with the difference in threshold in the damaged ear being over 20dB in some individuals. The asymmetry was far more pronounced at the higher frequencies, as would be expected if the damage was to the d-cells. However, it is worth noting that in control experiments where measurements of thresholds were made in intact animals, two individuals showed an asymmetry of over 15 dB at the higher frequency end of the range tested. This technique will surely find future applications in work on receptor mechanisms. The t y m p a n u m of the acridid ear is sensitive to lower frequency sounds at the position of the d-cells. In the grasshopper Chorthipt)us biglttluhts, the membrane vibratiork is greatest in the range 6 2 0 k H z and in the locust, although the greatest sensitivity is in the range 13-18kHz, the membrane is also sensitive to sound in the range 2 9 kHz (Meyer and Eisner, 1995). In both this species and the locust, the membrane response to low frequencies may be greater at the site c f d-cell attachment than at the other sites. However, as described above, the d-cells are much less sensitive at the lower frequencies. So there is a mismatch between the low frequency properties of the membrane and the frequency response of the d-cells. There is as yet no explanation for this observation, although it is probable that the frequency response of the ear is not solely related to movement of the tympanum and that the way in which Mtiller's organ responds may also have an influence (Stephen and BennetClark, 1982; Brekkow and Sippel, 1985). When the t y m p a n u m vibrates, Mtiller's organ vibrates as well, and the driving forces that act on it produce a twisting movement, described by Stephen and Bennet-Clark (1982} as being similar to a Lissajous' figure. So the vibration of the different areas of the tympanum, the orientation of the groups of receptor cells and the damping

180

D.J. ROBINSON AND M, J. HALL

of the membrane by Mfiller's organ, together with its complex vibration, all combine to give the response spectra of the single receptor cells. The conclusion of studies on the effect of ventilation on hearing in the locust (Meyer and Hedwig, 1995) and in Chorthippus biguttulus (Meyer and Eisner, 1995) is that the changes in tracheal pressure that result from ventilation will alter auditory sensitivity and the frequency spectrum of the response of the ear. The low-frequency components of incident sound can be reduced by about 15 dB SPL, while high frequencies can be enhanced by about 7dB SPL. At incident sound pressure levels of 50 or 55 dB SPL in the locust, the auditory response to 4 kHz has been shown to be masked by the membrane displacement (Meyer and Hedwig, 1995). It is possible that the changes that the ventilatory movements make in the response of the ear would reduce the effectiveness of frequency analysis in the auditory pathways, a factor that might explain the use of broad-band sounds for communication in grasshoppers. The leg movements of locusts have also been shown to produce responses in the tympanic nerve, as slow tympanic membrane displacements are associated with leg movements. If these occur concurrently with the high-frequency oscillations of the membrane produced by sound stimulation, then the two responses in the tympanic nerve interfere strongly (Lang and Elsner, 1994). Taken together with the effects of ventilatory movements, it is clear that the acridid nervous system contends with an auditory input that may be heavily modulated by other forms of mechanical stimulation.

4. 1.3

The go, Hid eat"

Like tettigoniids, gryllids have hearing organs on the tibia of the forelegs, with two tympana and auditory trachea that open to the exterior via auditory spiracles. Functionally, there are some differences. The small, anterior tympanum is not functional in hearing and has yet to have a role assigned to it. The larger posterior tympanum is fully functional and, in Go, llus campestris, vibrates in response to sound stimuli over a frequency range of at least 1 to 30 kHz (Larsen and Michelsen, 1978). The prothoracic spiracle provides access for sound, as in tettigoniids, but the acoustic tracheae from each side are separated by a very thin septum and there is good acoustic communication across from one trachea to the other. This gives four possible routes of sound input into each ear. Sound reaches the external face of the posterior tympanum and the internal via the prothoracic spiracle and the trachea. In addition, sound entering the opposite ear and the opposite spiracle travels across the thin septum separating the two acoustic tracheae, providing two more sound inputs. Thus, the tympanal vibration is the sum of all four inputs. The velocity and phase angle of the posterior tympanum is highly dependent upon the input from the ipsilateral spiracle, but less dependent upon the input from the

SOUND SIGNALLING IN ORTHOPTERA

181

contralateral spiracle. The input from the contralateral tympanum has only a small effect (Larsen and Michelsen, 1978; Michelsen et al., 1994b). During stridulation, the acoustic spiracle remains open, although crickets have the ability to close the spiracle, and the tympanic membrane responds to the pattern of the stridulatory song. There is no evidence that the responsiveness of the tympanum is altered during stridulation (Poulet and Hedwig, 2001). 4.1.4

The haglid ear

The haglids are most closely related to the tettigoniids, and the tympana and the acoustic trachea are similar, although the tympana are more rudimentary, as they are less well differentiated from the surrounding cuticle. There are two morphologically distinct regions of the tympanum, with an area of the dorsal margin being thicker than the surrounding membrane. This is similar to the tympana of some tettigoniids and stenopelmatids. The vibration characteristics of the tympanum of Cyphoderris monstrosa showed that the tuning did not match the low-frequency spectrum of the auditory nerve (Mason et al., 1999) with its best frequency of 2 kHz (Mason, 1991), but rather it matched the responses of a broadly tuned receptor group. So the ear of the haglids appears to be functionally different from that of other members of the Ensifera, in that there is a group of low frequency receptors that predominate in the attditory nerve and respond independently of the tympanal movements. Atympanate receptors for low-frequency sound have been described in Bullacris membracioides, a representative of a relatively ancient group of Orthoptera, where it is suggested that there was an evolutionary development from stretch organs to auditory receptors (van Staaden and R6mer, 1998). It seems likely that a similar evolutionary change has taken place in the haglids. C. monstrosa has essentially a two-channel system, with only one channel being associated with the tympana. 4.1.5

Changes with age in indivi~htals

The auditory system in crickets has been shown to exhibit maturational changes during the first few days of adult life. By the time of the final moult all the receptor cells are morphologically complete and fully developed (Ball and Young, 1974: Ball and Hill, 1978). The posterior tympanum continues to develop after the moult is complete and the membrane thickness decreases from around 20/~m to 5 # m (Young and Ball, 1974; Ball and Cowan, 1978). This maturation process results in a 5-8 dB improvement in the sensitivity of the ear at most frequencies, and improved directionality ( P o p o v e t al., 1994). Above 10 kHz, the decrease in threshold for phonotaxis is probably attributable to changes in the central processes, but there is no hard evidence available.

182

4.2

D.J. ROBINSON AND M. J. HALL AUDITORY RECEPTOR ORGANS IN THE TIBIA

All three pairs of legs in the Tettigoniidae have complex tibial organs. The arrangement of the receptor and accessory cells is essentially the same in all three pairs of legs (Schumacher, 1979; Lin et al., 1994). The post-embryonic development and functional morphology of the organs has been studied in detail (R6ssler, 1992a,b) in the genus Ephippiger. Investigation of the physiology of the complex tibial organs in the mid and hind legs of the bushcricket Gampsocleis gratiosa shows that the receptor cells of the complex tibial organs are mostly sensitive vibratory receptors with a frequency response that ranges from 200Hz to over 3 k H z (Kahnring et al., 1994). The receptor cells of the intermediate organ and crista acustica in these legs do not respond to airborne sound, as the structural components necessary for receiving sound are not present and the leg trachea does not appear to conduct sound to them. The three components of the tibial organ of the foreleg of tettigoniids contain both vibration and sound receptors. The subgenual organ is extremely sensitive to substrate-borne vibratory signals, and the receptors of the intermediate organ also respond to vibration. However, the receptors are probably bimodal, responding to sound signals as well. The scolopidial cells of the crista acustica are sensitive auditory receptors. The arrangement of the scolopidial cells of the crista acustica has always suggested that frequency analysis takes place in the ear. The cells are attached to the anterior trachea on the dorsal side, and are arranged in a row with the largest cells at the proximal end. The cells decrease in size towards the distal end of the row. The cap of each scolopidial cell is connected to a tectorial membrane that lies over the whole of the crista acustica and the intermediate organ. In the tettigoniid Ruspolia d([]erens the proximal scolopidia are tilted (Fig. 6), but the majority are aligned vertically (Figs 7 and 8). The arrangement of the crista acustica appears to be similar in Gampsocleis" gratiosa (see L i n e t al., 1993), two species of Decticus (see Kalmring et al., 1993), Mygalopsis mark± (see Kalmring et al., 1995) and Psorodonotus illyricus (see Kalmring et al., 1993). However, the number of scolopidia differs from one species to another, as do their physiological properties. There are 33 4- 1 scolopidia in the crista acustica of G. gratiosa in a length of 700 4- 30/zm whereas M. mark± has 2 3 ± 1 in a length of 5 0 0 ± 3 0 # m . However it is noteworthy that, despite the difference in length of the crista in the two species, if it is divided into four equally large sections, then in each species there are 17-20% of the scolopidia in each of the first two proximal sections. The third section has 21-22%, and the most distal one has 42-45% (Kalmring et al., 1995). In the phaneropterine bushcricket Polysarchus denticauda, there are about 50 scolopidia (Kalmring et al., 1996), 30% more than in other species studied and, although the proportions in the first three sections are similar, the fourth, most distal section has proportionally more, with some of the scolopidia in pairs or even triplets (Sickmann et al., 1997). By

183

SOUND SIGNALLING IN ORTHOPTERA Tn sC Cc Sd

mT

20 FIG. 6 The second scolopidial organ (Sd) from the proximal end of the crista acustica of Ruspolia di[l~'rens showing the inclination of the organ and the tectorial membrane (roT) stretched across the leg at an angle (compare with Fig. 1). Cc = cap cell: E - epidermal cells; sC = cell body; Tn = tympanic nerve.

Sd ~

FIG. 7

Con

A single scolopidial organ (Sd) from the distal end of the crista acustica of

Ruspolia di[li'rens. Bs = supporting band: Ccn = cap cell nucleus: Cs = supporting column.

c o m p a r i s o n , Tettigonia cantans, which has a crista acustica o f similar length to P. denticauda, has 30 to 37 s c o l o p i d a (Zhantiev and K o r s u n o v s k a y a , 1978) c o m p a r e d with 49 + 2. Both the n u m b e r o f s c o l o p i d i a a n d the p a i r i n g o f some o f them m a k e P. denticauda an unusual tettigoniid, b u t the significance o f these o b s e r v a t i o n s awaits e x p l o r a t i o n . 4.2.1

To,otopic organization O/'sense cells

A t o n o t o p i c o r g a n i z a t i o n o f the sense cells in the crista acustica was first suggested by Z h a n t i e v (1971) for Decticus verrucivorus. A p a r t i c u l a r p r o b l e m for p h y s i o l o g i s t s has been to d e t e r m i n e whether each r e c e p t o r cell itself has a frequency response d e t e r m i n e d by its internal structure or w h e t h e r the frequency response is d e t e r m i n e d by its p l a c e m e n t on the trachea. Initial

184

D.J. ROBINSON AND M. J. HALL

D

Ta

Cm

Tp

I

150p

I

...............~ t t /

FIG. 8 A reconstruction, from serial sections, of the acoustic trachea in the region of the tympanic organ of the right foreleg of Ruspolia d(ffi, rens. The posterior trachea is shown with the outer (posterior) wall removed. Cm = central membrane; Fv = first vertical scolopidial organ; S = position of tympanic slit in tibia; Ta = anterior branch of acoustic trachea; Tp = posterior branch of acoustic trachea. The line above the trachea shows the orientation of the scolopidial organs from the most proximal (P) to the most distal (D).

work on this problem (Oldfield, 1982, 1984, 1985) suggested that each receptor cell had a centre frequency of response that was a property of the cell itself. The evidence for this came from experiments in which removal of both tympana had no effect on the tuning of auditory receptors in the crista acustica (Oldfield, 1985). Recording from the cap cells of the scolopida disrupted either the tympanum or the dorsal cuticle, and this could have affected the responses of the cells. However, correlating the threshold curves of all the receptor cells in the crista acustica with the structure of the ear suggests that the frequency response of the cells could be a property of their position in the crista acustica (Linet al., 1993; Kalmring et al., 1995). In G. gratiosa, there is a clear relationship between the position of each scolopidia on the crista acustica and the characteristics of its threshold curve. The characteristic frequencies are shown in Table I. The scolopidia with the lowest frequency response are at the proximal end of the crista acustica, where the large cap cells make direct contact with the tectorial membrane (similar to Fig. 6). Moving distally, the tectorial membrane covers the smaller cap cells of the scolopidia in a V shape, restricting the area between the dorsal wall of the trachea and the membrane (similar to Fig. 7). This change in structure may be associated with the higher frequency threshold curve of these cells (Lin et al., 1993). It should be noted that these studies did not include complete staining of receptor cells, as has

SOUND SIGNALLING IN ORTHOPTERA

185

been achieved recently in Pholidoptera griseoaptera, where single cells were stained to show both their soma position in the crista acustica and their central branching pattern in the prothoracic ganglion (St6lting and Stumpner, 1998). Evidence that the scolopidia do not, in themselves, have an intrinsic frequency response comes from the comparison of the response to low frequencies of two tettigoniids (Kalmring et al., 1995). G. gratiosa has a broad band calling song with a fundamental frequency of 3.8 kHz and peaks at 7.6kHz and 30kHz. In contrast, Mygalopsis marki has a narrow band song with a lYequency range of 9 to 25 kHz. The threshold curves of the whole auditory organ of each species show that G. gratiosa is, as might be expected, much more sensitive to low frequencies below 5 kHz than M. marki. The important question is whether this difference is a property of the scolopidia themselves. The similarity between the structure of the crista acustica in these two species was mentioned above. Examination of the acoustic trachea reveals differences that account for the difference in response of the scolopidia (Kalmring et al., 1995). The transmission frequencies of an exponential horn depend upon the cross-sectional area and the flaring constant. Treating the acoustic trachea as an exponential horn gave a lower cut-off frequency (fo) of 3.05 kHz in G. gratiosa and 7.2kHz in M. marki. Thus the correspondence between the frequency spectrum of the conspecific song and the audible range of the ear is not brought about by the presence or absence of particular receptor cell types but by the structure of the sound-conducting system of the ears and a shift in the cut-off frequency of the amplifier - the acoustic trachea. Single cell recordings from auditory receptor fibres in the cricket Teh, o~ryllus oceanicus have revealed that, unlike tettigoniids, where there is a single characteristic frequency (CF) for each fibre, there are additional sensitivity peaks at other frequencies (Imaizumi and Pollack, 1999). Furthermore, the CFs are not distributed over a broad range of frequencies but instead fall into three groups, with the largest group centring close to the dominant frequency of the species song, 4.5 kHz (Balakrishnan and Pollack, 1996).

TABLE 1 The frequency response of individual scolopidia in the crista acustica of Gampsocleis gratiosa. Data from Lin el al. (1993) Scolopidia in the crista acustica (cell number) Proximal

Distal

Characteristic frequency (kHz) 1 to7 8to 11 12 to 16 17 to 23 24 to 31

4 5,4 7. 5-7,6-9.6 12,9 6 7,6 9.6 12,6 20 7 18,9 18, 10 18, 12- 25 16-20, 16 25, 16 30 Characteristic frequencies above 18kHz and up to 40kHz

186

4.3

4.3.1

D.J. ROBINSON AND M. J. HALL T O N O T O P I C O R G A N I Z A T I O N OF RECEPTOR PROJECTIONS

Tonotopic organization (?fprimarv af/i'rents in the prothoracic ~ganglion

Within the prothoracic ganglion there is a tonotopic organization in the branching pattern of the primary afferents, with low-frequency receptors projecting into anterior portions of the neuropile. Higher frequency receptors have projections into more posterior parts of the neuropile. Cells tuned to the frequencies of the conspecific song appear to have a larger projection area than other receptors. In Tettigonia viridissima the broad spectrum song has several peaks, including one at 20 kHz - a frequency that has been shown also to be important in directional hearing (Rheinlaender and R6mer, 1980). The receptor tuned to 20 kHz has the largest projection area within the auditory neuropile (R6mer, 1983) and, presumably, the largest number of synaptic contacts, although no link has yet been shown between absolute number of synapses and area of projection. However, staining of complete receptors in Pholidoptera gri~'eoaptera has shown no evidence for single receptor cells tuned to the frequencies of the conspecific song having larger projection areas (St61ting and Stumpner, 1998), although the general pattern of projections is similar to other bushcrickets. Proximal cells in the crista acustica project anteriorly in the auditory neuropile, medial cells project ventrally and posteriorly and distal cells project to the more dorsal regions. The termination areas of the auditory receptors overlap, so there is at least a possibility that auditory interneurons do not distinguish between different receptors with different frequency responses if they have terminations in the same area. This would mean a loss of frequency information, which does not seem entirely rational, even though auditory internenrons might thus gain a greater dynamic range. In the haglid Qvphoderris monstrosa, the auditory receptors fall into two groups, a low-frequency tuned group and a broad frequency tuned group. The projections of the two groups differ in that the low-tuned receptors have a bifurcation near the terminal aborization, whereas the axons of the broadly tuned receptors do not (Mason et al., 1999). This arrangement is quite different from that of other Ensifera, being simpler and less sophisticated, and it supports the idea (see section 4.1.4) that there are two separate hearing channels with independent evolutionary origins. The best frequency of the auditory nerve is 2 kHz, but the species song is an almost pure 12 kHz tone (Mason, 1991). The mismatch between the two frequencies is marked and only the broad frequency tuned receptors respond to the species song. 4.3.2

Tonotopic organization oj'primarv qfferents in acridids

Afferent neurons from the sense cells of the tympanic organ of grasshoppers (see section 4.1.2) enter the metathoracic ganglion complex at the first abdominal neuromere. Branches project to the neuropile of each thoracic

SOUND SIGNALLING IN ORTHOPTERA

187

ganglion, with the projections of the four different receptor groups separated geographically within the ganglia. In the metathoracic ganglion, the highfrequency afferents project posteriorly to the other three groups of lower frequency (R6mer, 1985), each of which projects to a distinct but more anterior region. There appears to be a gradient of decreasing sensitivity to low frequencies, both away from the midline and anteriorly (Halex et al., 1988; Bickmeyer et al., 1992). The higher-frequency afferents have a greater anterior projection into the prothoracic ganglion than most of the lowerfrequency ones, many of which do not project beyond the mesothoracic ganglion, while others only reach to the metathoracic ganglion. Although there has been substantial work on the central auditory pathways in grasshoppers (see, for example Boyan, 1999; Prier and Boyan, 2000), nobody has yet demonstrated that there is a tonotopic organization of projections within each of the four groups of receptor cells so, for the present, the tettigoniids are the group in which there is the most clearly defined tonotopic arrangement of receptor projections (see section 5.2.5 for further discussion of the central projection of receptor cells in acridids). A comparison of the projection of receptors in tympanic organs in insects can be found in the recent review by Stumpner and yon Helversen (2001). 4.4

D|RECTIONAL HEARING

Where the body size of an animal is smaller than the wavelengths of sounds of biological significance, the sound pressure difference between the external surface of the tympana is too small to permit directional hearing. However, in the Orthoptera, directional hearing is possible as a consequence of the fact that sound arrives at both the external and the internal faces of the tympana. The phase and pressure differences between the external and internal signals produce differences between the two ears, sufficient to provide directional information. The directional dependence of tympanal vibration has been measured in gryllids (Michelsen et al., 1994b) and acridids (Michelsen and Rohrseitz, 1995). Laser vibrometry measurements of the tympanal vibrations in acridids show that the vibrations are very much dependent upon the direction of the incident sound. The acridid ear is described as a two-sound-input model (Michelsen and Rohrseitz, 1995), because sound reaches the external surface directly or the internal surface via the air sacs (see section 4.1.2). The amplitude of the sound reaching the internal surface of the tympanum in Schistocerca ,~re~aria decreases with frequency. At 5 k H z the pressure difference across the membrane provides directionality. At 12kHz, where there is a much smaller transfer of sound through the air sacs, there is still a significant directional effect as diffraction effects are greater at the higher frequency (Michelsen and Rohrseitz, 1995). The directional cues provided by amplitude and phase angle are not as great in small acridids, Chorthippus b~uttulus is three to four times

188

D.J. ROBINSON AND M. J. HALL

smaller in length and diameter of thorax and abdomen than S. gregaria. The amount of delay in the sound transmitted between the two ears is therefore much less, and directionality is poor at 5 kHz (Michelsen and Rohrseitz, 1995). This effect would not be the same in small crickets, as the acoustic trachea provides a dramatic phase shift, although only at a restricted frequency range, providing crickets with excellent directionality at the dominant frequency of the species song. The finding of poor directionality in C. biguttulus at 5 kHz does not match behavioural observations which show that males have excellent lateralization. For example, the orientation turn during phonotaxis by the male is made on the basis of whether the sound source is on the left or the right and such distinction can be made with 100% accuracy when the sound amplitude differences are as little as 1 to 2 dB (yon Helversen and Rheinlaender, 1988; von Helversen, 1998). The results obtained by Michelsen and Rohrseitz (1995) have been challenged by Schul et al. (1999), who suggest that the phase delay between the external sound signal and that reaching the internal side of the tympanum is approximately 4 0 above that measured by Michelsen and Rohrseitz. Thus the internal pathway produces a considerably larger delay, resulting in good directional hearing at low frequencies and matching the behavioural observations. Using the "biological microphone' experimental design (Rheinlaender and R6mer, 1986) in which recordings are made in the field from identified nerve cells, the directional hearing ability of C. biguttulus has been examined in the natural habitat (Gilbert and Elsner, 2000). In sparse vegetation the overall pattern of directionality was similar to measurements made in the laboratory under free-field conditions, with only a small degradation. Correct lateralization would be expected at all angles of incidence between 20 and 160° in the laboratory, figures that match those obtained in behavioural studies. Over gravel, directionality was much more degraded and, over dense vegetation, there were several angles of incidence of sound where the difference between the two ears would have been less than 1.5dB, no longer providing good lateralization of the sound source, particularly if the animal is horizontal. These field neurophysiological observations have yet to be confirmed by field behavioural studies. Gryllids use relatively low frequencies for communication, and the differences in sound intensity at the two ears as a consequence of diffraction are small. At 4-5 kHz the presence of the cricket distorts the sound field such that there is a l - 2 d B difference in intensity at tympana and thoracic spiracles, compared with the undistorted field (Kleindienst, 1978). However the sound pressure and phase differences due to the four inputs to the cricket ear (see section 4.1.3) provide directional information. There is a delay imposed by the medial septum upon information from the contralateral side, which contributes to the pressure gradient across the tympanum. With a sound source at 90 <~to the cricket, the difference in pressure between the two ears may be up to 15 dB

SOUND SIGNALLING IN ORTHOPTERA

189

(Michelsen et al., 1994b). Central processing enhances directionality (see section 5.2.2). At high frequencies, diffraction effects are more marked and tettigoniids, which generally communicate in the range of 10kHz up to 102kHz, are able to gain directional information in a similar way to acridids. The gain produced by the exponential horn shape of the acoustic trachea has been measured as 10 15dB in some species (Heinrich et al., 1993), so the primary source of tympanal stimulation is via the spiracle. In experiments with Gampsocleis gratiosa, blockage of the spiracle markedly influenced the auditory threshold and directionality of the ipsilateral auditory neurons, but had very little effect on contralateral auditory neurons, while blockage of the tympanal slits had only a small effect on directional hearing (Shen, 1993). It has been pointed out, however (Michelsen et al., 1994b), that blocking sound inputs can change the radiation impedance of the tympanum and so influence the results observed. A change in membrane impedance was observed when the ipsilateral spiracle of Gry//us bimaculatus was blocked (Michelsen el a/., 1994b). The role of the sound entering via the tympanal slits in directional hearing has been demonstrated by Bailey and Stephen (1978), and it is clear that the slits do have a function in hearing. The air cavity, formed by the cuticular cover over the tympanum that forms the slits, may have resonance properties that, under certain conditions and at certain frequencies, modify the pressure across the tympanic membrane. Some tettigoniids even possess structures by the slits that may act in a similar way to the pinna of mammals (Bailey, 1993).

5 Analysis There are four key pieces of information that most orthopterans glean from acoustic signals: the amplitude modulation pattern, the frequency, the intensity of the sound and the location of the sound source. For example, males of the grasshopper species Chorthippus biguttulus produce a calling song to which the female replies. The male uses the female song as a signal for phonotaxis, and so retrieves from the song the information that the caller is a conspecific. When the female sings, the male turns towards her, on the basis of a simple decision about whether to turn left or right. After turning, the male sings again, and then, when the female replies, makes a new decision (yon Helversen, 1972; yon Helversen and yon Helversen, 1983). When males were given a choice between two speakers playing female calls, it was found that they selected the speaker with the loudest sound, even if the difference in intensity between the two sounds was only 1 2 dB (von Helversen and Rheinlaender, 1988). The acoustic cues that orthopterans use and the neural mechanisms by which the information is analysed are the subject of much research, and there has been considerable recent advance in our understanding. In addition to the purely physiological interest in the mechanisms, there is now sufficient information

190

D.J. ROBINSON AND M. J. HALL

available for us to start to work out the evolutionary stages in the development of auditory processing in the nervous system.

5. [

SONG ANALYSIS

The environment of most orthopterans is a noisy one and their communication systems would be expected to show adaptations that favour signal transmission and signal detection. The use of signals for species recognition implies a level of sophistication in the analysis of signals, and yet the signals must survive degradation in the natural environment without losing their information content. Where the signals are used for location of potential mates, the signals must maintain their directional information, even in the complex environment that a meadow or hedgerow presents. The use of the fundamental or carrier frequency of the song to convey information can provide a measure of resistance to masking (R6mer and Bailey, 1998). This may be enhanced if parallel selection has matched the best frequency of the receiver and the carrier frequency of the song (Endler, 1992), and in most Orthoptera there is such close matching, although in some species there is a mismatch (Bailey and R6mer, 1991). A recent study of a southern Australian sagine tettigoniid, Sciarasaga quadrata, illustrates the selection pressures that may operate on a species in maximizing the effectiveness of acoustic communication (R6mer and Bailey, 1998). The carrier frequency of the call is 5 kHz, which is very low by comparison with other tettigoniids. The intensity of the song is also lower than bushcrickets of comparable size, being 60 dB SPL at 1 m. The sensitivity of the ear to sound in the range 3-70 kHz was measured by analysing the discharge in the tympanal fibres and the omega neuron. At the carrier frequency of the male call (5 kHz), the ear is approximately 20 dB less sensitive than it is between 10 and 20 kHz. The difference implies that the ear and call are not matched. It also implies that if there are other species in the same environment utilizing the frequency band between 10 and 20kHz, the response of the ear to their calls might mask the 5 kHz calls of conspecifics quite effectively. R6mer and Bailey found three such species with long-duration calls, Mygalopsis pauperculus, Hemisaga denticulata and Metaballus litus. Field recordings suggested that the calls of S. quadrata would be masked by the calls of other species. This was demonstrated neurophysiologically by using a portable preparation of the omega cell in the field, first described by Rheinlaender and R6mer (1986). The preparation was firstly positioned with conspecific males at distances of l m and 6m and a microphone 5cm away. The record from the omega neuron (Fig. 9A) shows that both males are detected. A second measurement was made with a conspecific male at a distance of 4 m and a male Mygalopsis paupercuhls at 6 m. The continuous song of M. pauperculus masks the response in the omega cell to the conspecific (Fig. 9B). Moving the male M. pauperculus

SOUND SIGNALLING IN ORTHOPTERA

191

closer (3.5 m) increased the masking effect. In this case (Fig. 9C), another heterospecific was also singing, from slightly further away. In section 4.1.2 it was suggested that in grasshoppers such as Chorthippus higuttulus, the changes that the ventilatory movements make in the response of the ear would reduce the effectiveness of frequency analysis in the auditory pathways and that this observation (Meyer and Eisner, 1995; Meyer and Hedwig, 1995) might explain the use of broad-band sounds for communication in grasshoppers. Poor spectral resolution may preclude the use of carrier frequencies centred in frequency bands that have low noise levels, and so a combination of a broad-band signal and analysis in the time domain might be expected to be a solution found in grasshoppers. The temporal pattern of the song of C. bi,~uttulus is species-specific, and recognition prevents hybridization with species of gomphocerine grasshoppers (von Helversen and von Helversen, 1975: von Helversen and yon Helversen, 1994). The song consists of a number of repetitions of stereotyped subunits (Elsner, 1974), each lasting about 80 ms. The duration of the song of females under natural conditions has been measured as 1.18-~ 0.23s, with a range of 0.8-1.5s (yon Helversen and yon ttelversen, 1975). HoweveL experiments have demonstrated that under laboratory conditions males can recognize the songs of conspecific females if the duration is as low as 250 ms (Ronacher and Krahe, 1998), Is it possible that longer duration signals might improve recognition in the field, where the levels of background noise are likely to be much higher than in the laboratory

~li|tl|ll|ltll

~llll~ll

I~ll~lHlll

A I I I I J I I I I I I ~ I I I I I I W l l WlTIII~IlIIE t t l ! l l l l

I~ t ~

"1t

C iw,,,,-,,~,ln,~,,,,,,F.,~MI . m . , . . = . . ,q u=~ ,., ,., nol i p,l .m .m .n P.p.~.u ~

tllllll.t|Jllll|llll llllllllNII l l l l l I ] l l l

&

I IIII

i i i i i i ll~ ] 1 1 I II i~ II1~11~ I I + 11~ 1tli

I.mimll

. . . . . . W. l.l m l l

|~n~lnmlnl

iL I [ l l , l

lllUll~llU~lla l~|IIll~l

,-. I[W!:

FIG. 9 Masking of the response to conspecific male calls of Sciarasaga quach'ata by heterospecific bushcricket species in the feld, using the response from the omega cell as a measure (upper trace) and a microphone by the preparation to record the calls (lower trace). A The response in the omega cell is shown to the calls of two conspecific males, one 1 m away and one (arrowed) 6 m away in the same direction. B Masking of the call of a conspecific male (arrowed) 4 m away by a male Mygalopsis paupercu/us 6 m away. C. Increased masking when the male M. pauperculus is 3,5 m away. Another heterospecific is calling at a distance of around 5 m. (R6mer and Bailey, 1998, fig. 5). Reprinted with permission from the Company of Biologists.

192

D.J. ROBINSON AND M. J. HALL

(Gilbert and Elsner, 2000)? Under natural conditions males produce calling songs. A female who is willing to mate replies with her own song, providing the male with a phonotactic cue (von Helversen, 1997). The male makes a turn towards the female and moves in that direction before singing again. The female response produces another turn. The male traces out a zigzag course towards the female because he does not turn through an angle that corresponds with the direction of the sound source. Rather, he lateralizes the sound (Rheinlaender, 1984). The ability of C. biguttulus males to recognize female songs amidst background noise has been tested under laboratory conditions (Ronacher et al., 2000). In the laboratory, the responses of freely moving males to songs containing differing numbers of subunits, and therefore of differing durations, were scored on the basis of the number of turns made as a percentage of the total number of presentations of the song. With a stimulus intensity of 45dB or 50dB and no masking noise, a male responded to 100% of stimulus songs, whether they contained three subunits or 12 subunits. As the intensity of the masking sound was increased (Fig. 10) the response percentage fell. However, the longer song was more resistant to masking than the short one.

Taking the 50% response level and comparing long and short songs showed that there was a 2-3 dB shift (AdB) in the response curve. An analysis of this shift for all animals tested suggested that 1 dB of added noise led to a mean value of 14.8% decrease in response to the 12-subunit signal and a 13.2% decrease in response to the three-subunit song. Additional tests were made

100, =

::]~------o..

g

',

_~ 25

' \

45 a

,,

",

'b,, "

50 55 noise level [dB SPL]

=11:45.8

"

60

FIG. l0 Response of a male Chorthippus biguttulus to short (three subunits) and long (12 subunits) female songs under masking noise broadcast from above. The percentage of turning responses is shown relative to the total number of stimulus presentations. The solid symbols represent the responses to stimuli with an intensity of 45 dB and the open symbols the response to stimuli with an intensity of 50 dB. The total number of stimuli tested was 541 (Ronacher et al., 2000, fig. 2a). Reprinted with permission from Springer-Verlag GmbH & Co. KG.

SOUND SIGNALLING IN ORTHOPTERA

193

using four-subunit and five-subunit songs, compared with the responses to 12subunit songs. The mean values obtained for +dB are shown in Fig. l l, and the slope of the regression line crosses the zero line at 441 ins duration. This suggests that song lengths greater than this value do not improve significantly the detection of the signal against background noise. A further feature of long signals in C. biguttulus is the opportunity for the male to make a correcting turn immediately after he has turned towards a female. A short song will have ended before the turning response of the male (the open-loop state), but a long song continues until after the male has completed his turn and he can therefore make a further correcting turn (closed-loop state). In a comparison of the turns performed by males in response to short and long songs, only one correcting turn was observed to the three-subunit song (n = 101L whereas 55 were observed in the responses to the 12-subunit song (n = 129) (Ronacher et al., 2000). The fact that females produce songs that are at least twice as long as the duration needed by the male for recognition under noisy conditions would appear to be the result of selection for a more a direct and faster phonotactic response, in that the male can perform two turns in response to one burst of female song. The length of the song does not appear to be an aid to lateralization. Males have been shown to be equally effective at making correct turns to three-subunit songs as to 12-subunit ones (Ronacher and Krahe, 1997; Ronacher et al., 2000).

,

oo c O

3

4

5 subunits vs. 1 2 subunits

4, m

3. O 143 co nn

2. intersection at 441 ms

1

<1

0 -1

2O0

360 .......

460

'

560

signal duration [ms] FIG. 11 The influence of signal duration upon noise tolerance. The values of &dB are calculated relative to the 12-subunit song (Ronacher et al., 2000, fig. 3b). Reprinted with permission from Springer-Verlag GmbH & Co. KG.

194 5.2

D.J. ROBINSON AND M. J. HALL IDENTIFIED A U D I T O R Y I N T E R N E U R O N S IN T H E NERVOUS SYSTEM

The auditory receptors project into the prothoracic ganglion (crickets and bushcrickets) and the abdominal ganglion (grasshoppers) and their aborizations are entirely ipsilateral. The projections are primarily to the medial ventral association centre (mVAC} of the ganglia. Many local interneurons are small and remain uncharacterized, with most of the information we have coming from studies on a few large interneurons, for example the omega neuron. Ascending neurons generally have dendrites in the ganglion on the soma side, and an ascending axon contralateral to the soma, that ascends to the brain. In comparing the identified and characterized neurons in orthopterans it is clear that there is substantial homology between the groups, for example in the T-cell. Furthermore, first order interneurons in the grasshopper have been shown to be serially homologous along the central nervous system, which suggests that the tympanal receptors are a specialized part of the general chordotonal system, with a similar arrangement of projections from homologous receptors to homologous interneurons (Boyan, 1993). This developing picture of homologies between groups and serially within individuals is one of the exciting outcomes of recent neurophysiological work, which is now able to feed into discussion of the evolutionary relationships within the Orthoptera (van Staaden and R6mer, 1998) and wider relationships with other insect groups (see, for example, Yager, 1999: Prier and Boyan, 2000). 5.2.1

The T-cell

T-shaped interneurons iT-cells) have been identified in many groups of orthopteran insects. Since the first one was identified in the tettigoniid Gampsock, is buergeri (see Suga and Katsuki, 1961), their presence has been described subsequently in other tettigoniids (Rheinlaender el al., 1972; Schul, 1997: Faure and Hoy, 2000b), haglids (Mason and Schildberger, 1993), gryllids (Atkins and Pollack, 1987), acridids (Ramer, 1985) and mole crickets (Mason el al., 1998). The anatomy of the T-cell was described by Kalmring et al. (1979), who identified it as part of a giant fibre system, and Oldfield and Hill (1983) who localized the cell body in the prothoracic ganglion. Information from the tympanic nerve in the foreleg reaches the prothoracic ganglion where the cell body of the T-cell passes information both anteriorally and posteriorally in the ventral nerve cord. The T-cells are paired in the prothoracic ganglion, and it has been shown that responses to auditory stimuli in the ipsilateral tympanic nerve exert a strong inhibitory influence on activity in the contralateral T-cell (Rheinlaender e l al., 1972). The T-cell might be expected to play a role in the detection and localization of the conspecific song. However, studies comparing the tuning curve of the T-cell with the spectral content on the species song have given equivocal results. For example, in the zaprochiline Kawanaphi& nartee, Bailey and R6mer (1991) found that

SOUND SIGNALLING IN ORTHOPTERA

195

the T-cell in both sexes had a best frequency of 15 kHz below the peak spectral frequency of the calling song of 50 kHz. This does not preclude the T-cells from a role in detecting the conspecific song; it suggests that other roles, such as predator avoidance, may have shaped the tuning curve during evolution. The most comprehensive study of the T-cell in a tettigoniid is that of Faure and Hoy (2000a,b,c) who have studied the neuroethology of the cell in

Neoconocephalus ensiger. A typical tuning curve for a T-cell is shown in Fig. 12 and is remarkably similar to the tuning curves measured by other workers for different species of tettigoniid (Suga, 1966: Rheinlaender and R6mer, 1980). The tuning curves for male and females are also similar, but the T-cell of females is slightly more sensitive to audio frequencies than in males. The best threshold for females was about 5 dB lower than in males, and the best frequency was about 5 kHz lower. In both sexes, the best frequency of the T-cell was 20 kHz or higher (Faure and Hoy, 2000b). From this it appears that the female will respond better to the conspecific song than males. In tests using a simulation of the calling song, the female T-cell responded better than the male (Faure and Hoy, 2000c). The calling song of N. enslaver has a peak frequency at about 14 kHz with a range of 9 to 25 kHz, but there is appreciable sound energy in the ultrasonic range up to 40 kHz (Faure and Hoy, 2000d). The ability of the T-cell to follow the temporal patterning of sound was not the same at all frequencies. At a frequency of 40 kHz, with the sound source at 9 0 , the T-cell encoded the temporal pattern of the sound pulses over a wide

8O 70 60 0

.xz

40 3O I

0 FIG.

12 The

mean

'

!

!

I[)

20

'

I

'

I

'

I

'

I

I

30 40 50 {,0 70 t,'requenc~ (kl Iz)

excitatory

threshold

tuning

'

I

'

I

'

|

8{t 91) 100 curve for

the

T-cell

of

Neocom~cephah~s ens~er, with the loudspeaker situated at 9 0 to the long axis of the body. Note the higher degree of variability in audiosound (7 17kHz) thresholds (Faure and Hoy, 2000b, fig. 2A). Reprinted with permission from the Company of Biologists.

196

D.J. ROBINSON AND M. J. HALL

range of repetition rates with almost one-to-one coding up to a repetition rate of 30 Hz. However, at a frequency of 15 kHz, which is within the maximum peak energy band of the calling song, accurate coding of the temporal pattern was degraded above a repetition rate of 5 Hz (Faure and Hoy, 2000b). The latency of the response in the T-cell also varied with frequency, being shorter at 4 0 k H z than at lower frequencies. Both of these findings would support the view that the T-cell is able to provide the insect with predator-avoidance information. Searching bats increase the repetition rate of their locating pulses when approaching an object, and decrease their calling intensity. The T-cell with its short latency of response and an ability to follow fast repetition rates would enable an insect to detect a bat that was preparing to attack (Fullard, 1984). The responses of the T-cells of N. ensiger to simulated bat calls have been compared with the responses to a simulation of the conspecific song (Faure and Hoy, 2000c). The syllable repetition rate of a calling N. ensiger at 25"C is 14.25Hz (Faure and Hoy, 2000d), and simulated bat calls used the same repetition rate. Two types of bat call were simulated, one (Bat 10) with 80kHz to 30kHz FM sweeps of 10ms duration, and a second (Bat 30) with sweeps of 30 ms duration, the duration corresponding with the duration of syllables in the calling song of N. ensiger. The way in which these simulated calls were encoded in the T-cells was not the same for each stimulus type. At 90dB SPL the T-cell followed the Bat 10 signal perfectly and, when the sound intensity was reduced to 70dB SPL, the ability of the cell to follow the temporal pattern was almost perfect. At 50 dB SPL the ability to follow started to break down. However, at all three sound intensities the T-cell did not follow the Bat 30 or conspecific song perfectly, with the ability to follow the pattern being poor below 90 dB SPL. However, it was clear that other neural units with smaller responses than the T-ceil were following the temporal pattern of the calling song perfectly at 70 and 90 dB SPL. The T-cell has been shown to adapt to the conspecific calling song (Kalmring et al., 1979), but it has also been reported that there is a one-to-one correspondence between syllables and responses (Suga and Katsuki, 1961). In the bushcricket Tettigonia viridissima, where the song is a continuous repetition of a pair of syllables, the T-cell (TNI) follows the individual syllables without adaptation (Schul, 1997). However, experiments with song analogues of different spectral composition have shown that the T-cell is responding to the ultrasonic component of the songs only. Stimulation with pure tones showed that the T-cell responded to a wide range of frequencies (4 to 60 kHz) and the spiking responses were both phasic and adapting, to ultrasonic frequencies. Regular spiking responses were not evoked by the audible range of frequencies. A second, smaller T-shaped interneuron (TN2) has been discovered (Schul, 1997) in T. viridissima. It has a small axonal diameter and its responses to stimulation are not easy to detect using extracellular recordings from the

SOUND SIGNALLING IN ORTHOPTERA

197

connectives. The responses of TN2 are strongly phasic and adapting, and the cell is primarily tuned to ultrasonic frequencies, with the best response being at 30 kHz, where the threshold is below 40 dB. The established view, or assumption, that the T-cell is primarily the agent for the detection and localization of conspecifics has now been challenged, It appears very likely, from the recent evidence (Schul, 1997: Faure and Hoy, 2000b,c), that the role of the T-cell in tettigoniids is not the detection of the conspecific song. Instead, the T-cell detects frequencies in the ultrasonic region and mediates predator avoidance and escape mechanisms. Where the conspecific song contains ultrasonic components these are detected by the T-cell (Schul, 1997), but it remains to be seen how significant this is in the identification and location of conspecifics.

5.2.2

The omega neuron

The omega neuron is shaped like the Greek letter f2 (Fig. 13) and is a first order interneuron that receives auditory input from the receptors in the tibial tympanic organ and connects the auditory neuropiles in each half of the prothoracic ganglion. Omega neurons are found in both major groups of the Ensifera, the Grylloidea (Casaday and Hoy, 1977: Wohlers and Huber, 1982) and the Tettigonioidea (R6mer et al., 1988), and there is evidence that they are homologous (Schul, 1997). The neurons are paired in tettigoniids but there are two pairs in the Gryllidae (Wohlers and Huber, 1982; Stiedl et al., 1997). Each of the paired omega neurons has branches in both halves of the prothoracic ganglion, with the processes on the same side as the cell body being mostly post-synaptic, receiving excitatory inputs fi'om the receptors in the ipsilateral ear. The processes of the neuron in the contralateral side of the ganglion are mostly presynaptic and exert inhibitory control of auditory neurons receiving input from the other ear and the other omega neuron of the pair (Watson and Hardt, 1996). Thus the pair of omega neurons enhances the contrast between the two ears and so improves the ability of the insect to localize the source of sound. 5.2.2.1 The omega neuron 1 ( O N I ) . The omega neuron ONI responds to sound over a wide frequency range but has been found to be most sensitive at the carrier frequency of the conspecific calling song, for example in Gwllus [)ilHacllldlllS (see Popov el al., 1978) and Gryllus campeslris (see Wohlers and ttuber, 1982). A typical response to a conspecific calling song model, recorded in Tettigonia vMdissima, is shown in Fig 13b (Schul, 1997). On the ipsilateral side the response is tonic and has a threshold of about 40 dB. The EPSPs (excitatory post-synaptic potentials) and the spikes match the syllable structure of the song and do not show long-term adaptation over the duration of the song models (up to 20 s), although the spike response to the initial syllables of

198

D.J. ROBINSON AND M. J. HALL

1O0 pm

~as

~

IlOmV

76dB

FIG. 13

(upper) An omega cell in the prothoracic ganglion of the bushcricket

Tettigonia rMdiasima (Schul, 1997, fig. 3A). (lower) The response of the omega neuron 1 (ON) to stimulation from models of the conspecific calling song. Dendritic recordings from ON and PSTHs (peri-stimulus time histogram) to stimulation delivered from the soma-ipsilateral (left) and soma-contralateral side (right). The stimulus intensities are indicated. The top traces correspond with the first stimulus sweep, and the bottom traces to the third one, demonstrating the pronounced responses to the first syllables of a song. PSTH: 15 repetitions, bin width 2 ms (Schul, 1997, fig. 4).

SOUND SIGNALLING IN ORTHOPTERA

199

the stimulus is somewhat greater than for subsequent syllables. The responses to stimulation from the contralateral side were weaker than those on the ipsilateral side, but there was variability in the level of response between different individuals (Schul, 1997). In the cricket Teleogr),llus ocea~licus the omega neuron ON 1 is most sensitive to the frequency of the conspecific song (4.5 kHz), but the latency of the response is longer than for other frequencies (Pollack, 1994: Faulkes and Pollack, 2000). Latency decreases as sound intensity increases but the response at 4.5 kHz remains longer than at other frequencies. The difference in latency is 5 10ms. This latency is observable not only as a delay in the onset of spikes but also of the EPSP. The functional reason for the longer latency at the species-specific frequency is not yet clear, although it is linked to inhibitory inputs onto ascending auditory interneurons (see section 5.2.3). The possible mechanisms that might account for the increased latency have been investigated by Faulkes and Pollack (2001). The possibility that auditory receptors might have different conduction velocities, based upon their frequencies of response was examined by Pollack and Faulkes (1998), and they tk)und no evidence to support the idea. However, unpublished results from lmaizumi and Pollack, quoted by Faulkes and Pollack (2001) indicate that there is a subpopulation of low-frequency auditory receptor neurons that has lower conduction velocities than other auditory receptor neurons. Whether this fact contributes to the latency of response of ON I at the frequency of the conspecific song is not known, as the difference in conduction velocity is only 2 3 ms. A possible mechanism l~r the production of the latency is postsynaptic inhibition that precedes or is simultaneous with the excitation produced by the sound stimulus. When intracellular recordings were made (Faulkes and Pollack, 2001) using microelectrodes filled with potassium acetate (to avoid the use of chloride ions and thus hyperpolarization of the cell), no early IPSPs (inhibitory post-synaptic potentials) were detected. There was no reduction in the rate of firing or of the height of the spikes prior to the response evoked by the stimulating sound. EPSPs that are evoked by ultrasound have a faster rise time than those evoked by the species-specific frequency of 4.5 kHz and also a shorter latency. This demonstrates that at least part of the explanation for the longer latency at 4.5 kHz must lie in events that are presynaptic to ON I. Faulkes and Pollack (2001) argue that the likely explanation is that there is a polysynaptic afferent pathway to ON I that is specific to 4.5 kHz, in addition to a monosynaptic one. They cite as evidence the fact that a very brief 5 k H z stimulation generates two EPSP peaks 12ms apart, but a short ultrasonic stimulus elicits a single peak. The first peak would be evoked via the monosynaptic pathway and the later one via the polysynaptic pathway. An explanation of the functional significance of this arrangement of the inputs to ON1 for species recognition and localization will be of great interest. There is some evidence that the latency is linked to the relationship between ON I and the ascending neurons I and 2 (ANI and AN2) (see section 5.2.3).

200

D.J. ROBINSON AND M. J. HALL

A slowly developing inhibition of ON1 has been observed in Tettigonia viridissima. It has been shown that when two conspecific songs of differing intensity were presented together from the same direction, the louder song (60dB SPL) was represented as spikes in ON1, while the representation of the weaker song (45dB SPL) was completely suppressed (R6mer and Krusch, 2000). The weaker song, when presented on its own, was effective at generating a response in ON1. The explanation for this observation came from further intracellular work, which showed that the synaptic events had two components: excitatory EPSPs encoding the train of syllables and a slowly developing hyperpolarization of the cell. This inhibitory hyperpolarization could take as long as 10s to develop fully and as long as 15 s to disappear and for the resting membrane potential to become re-established. The effect of the hyperpolarization is to reduce the gain of the cell such that only the more intense signals produce a spiking response. One consequence of the gain control is a sharpening of the directional response, such that two sound sources at an angular separation of as little as 7.5 ~' either side of the midline axis will be represented significantly better in the ipsilateral ON I. Thus the auditory field around the bushcricket is divided sharply into two hemispheres. A further point that arises from this study is the very long time course of the development of inhibition. This inhibition is effective in T. viridissima because the males sing for very long periods. In a species with short songs, this gain control mechanism would be ineffective. Thus there could be a link between the evolution of long duration signals, chorusing (see section 8.2) and this inhibitory mechanism. It will be interesting to see what comes out of future studies of long-term inhibition in ON1 and the pattern of distribution of the phenomenon in tettigoniids and gryllids. 5.2.2.2 The omega neuron 2 ( O N 2 ) . The morphology of the omega neuron ON2 in crickets has been described in Gryllus campestris (see Wohlers and Huber, 1982, 1985) and its acoustic responses have been reported (Lewis, 1992), but the only detailed study of its physiology is the recent one of Stiedl et al. (1997). The morphology of ON2 is very similar to that described for ON 1 (Fig. 13). However, unlike ON1, there are one or more cross-over branches that connect the anterior ring tracts in each hemiganglion. A small-diameter ascending axon is sometimes present. It was found in six out of 41 preparations of Acheta domesticus and all six were from females less than eight days old (Stiedl et al., 1997). This is similar to the situation in ON1 in the cricket Teleogo'llus ocea, icus, where ascending axons are found in young females (Atkins and Pollack, 1986). The absence of evidence of ascending axons in males and older females suggests that degeneration occurs during development and that there is an interesting story about the function of these axons still to be researched. ON2 in A. domesticus is a spiking neuron with a frequency response that is about 20 dB more sensitive at 16 kHz than it is at 5 kHz (Stiedl et al., 1997),

SOUND SIGNALLING IN ORTHOPTERA

201

similar to T. oceanicus (see Lewis, 1992). It receives both excitatory and inhibitory inputs from both ears and does not have a directional hearing function. The role played by ON2 in the auditory system of the cricket is speculative at present. The response to high frequency might suggest a role in predator avoidance, although there is no directional response. In combination with other auditory interneurons it could produce patterns of inhibition that enhance syllable pattern processing, although the lack of GABAergic or serotonergic immunoreactivity found in ON2 in G. campestris suggests otherwise (Hardt et al., 1994; Watson and Hardt, 1996). The courtship song in crickets is produced by the male when very close to the female, so it will be received at a high intensity. There remains a possibility that despite the tuning curve being most sensitive at 16kHz, the response to the courtship song is mediated in some way via ON2. 5.2.2.3 Omega neurons h~ acridids. Experiments in which fluorescent tracers were injected into the afferent neuropile of the metathoracic ganglion of the locust showed the presence of omega-shaped neurons in the prothoracic and mesothoracic ganglia (Lakes-Harlan et al., 1998), These were presumably labelled trans-synaptically, although a descending axon, observed in some preparations, may have carried the label. Labelled cell bodies were found on both sides of the ganglion, indicating that input synapses occur on both sides of the aborizations, as observed in ensiferans (Watson and Hardt, 1996). The questions of whether these omega neurons are homologous with those found in ensiferans, and whether they have an auditory function, remain unanswered. 5.2.3

Ascending auditory interneurons

Ascending auditory interneurons have been described in a number of species. In crickets, the neurons AN I and AN2 have been identified (Wohlers and Huber, 1982; Hennig, 1988) and their roles in phonotaxis investigated (Harrison et al., 1988: Schildberger and H6rner, 1988). In tettigoniids a number of authors have described auditory interneurons, and some of them at least can be identified as being homologous with those in crickets (Stumpner, 1997). Two pairs of ascending neurons have been described (Schul, 1997) in Tettigonia viridissima, and in both morphology and physiology are similar to AN I and AN2 in crickets. AN I is tuned to low frequencies and receives inhibitory input during stimulation by ultrasound, while AN2 predominantly receives information about ultrasonic signals and does not display frequencydependent inhibition. These characteristics are also found in crickets. AN2 in T. viHdissima responds tonically to the species song and encodes the syllable pattern, unlike AN2 in crickets, which does not normally resolve the syllable pattern (Hennig, 1988; Schul, 1997). The most complete characterization of an ascending neuron in tettigoniids has been carried out in the phaneropterine bushcricket, Ancistrura n&rovitlata

202

D.J. ROBINSON AND M. J. HALL

by Stumpner (1997). AN1 has a cell body (soma) in the anterior dorsal region of the prothoracic ganglion. An ascending axon runs through the suboesophageal ganglion, without branching, until it reaches the protocerebrum of the brain (Fig. 14). The dendritic branches arise from the primary neurite, in the hemiganglion that is contralateral to the cell body, although some of the branches cross back into the ipsilateral side. Acoustic communication in A. nigrovittata has been described by Dobler et al. (1994a) who showed that the hearing range extended from 2 k H z to well into the ultrasonic range, with the best frequencies being in the range 20 to 25 kHz, where the sensitivity was 30 dB SPL. The male song has peak energy at 15 kHz. Like other phaneropterines studied, A. n~rovittata females produce a song in reply to the male, and they are most sensitive at 12 and 16 kHz. The peak frequency of the female song is 28 kHz. A N I is most sensitive at 16 to 20 kHz and the sensitivity falls off rapidly between 20 and 30 kHz (Stumpner, 1997). The tuning curves are similar for males and females, although males are slightly more sensitive at 20 and 24kHz. At 16 to 20kHz, A N I is more sensitive than the female behavioural thresholds for responses over the same frequency range that were measured by Dobler et al. (1994a). At frequencies in the range 4 to 10kHz and in the range 24 to 50kHz, there are prominent IPSPs, but between 10kHz and 2 0 k H z there are no subthreshold IPSPs and the first response to a stimulus is an excitatory one (Stumpner, 1997). In general, AN1 is more sensitive to stimulus fi-om the soma-contralateral side and the IPSPs are still present when the input from the soma-ipsilateral side is removed by lesion of the tympanic nerve. At intensities up to 55 to 65 dB the soma-contralateral stimuli, evoke more spikes than the soma-ipsilateral stimuli, and the maximum difference is equivalent to a left- right difference of

FIG. 14 A ventral view of the AN1 neuron in the prothoracic ganglion of Ancistrura n(grovittata (Stumpner, 1997, fig. 1). Reprinted with permission from the Company of Biologists.

SOUND SIGNALLING IN ORTHOPTERA

203

30dB. However, at higher intensities the difference is less, and at some frequencies the ipsilateral stimuli may evoke a greater response. The asymmetry in the response implies that there is inhibitory input that is soma-ipsilateral. This was demonstrated by cutting the soma-contralateral tympanic nerve, which eliminated all spiking responses, but displayed IPSPs coupled to the stimuli (Stumpner, 1997). In T. viridissima, white noise stimuli of increasing intensity generated an 1PSP that reduced and delayed the spiking response of ANI (Schul, 1997). This inhibitory input is tuned to ultrasonic frequencies, with stimulation of 7 and 12kHz producing purely excitatory responses. Above 12kHz the IPSPs appear and have a greater effect in delaying and reducing spiking as frequency increases until, at 30kHz, only the inhibitory response is present. The source of the inhibitory inputs acting upon AN I have yet to be determined in A. nigrovittata, but experiments on the cricket Teleogryllus oeeanicus suggest that ON1 is the main, and perhaps the only, source of contralateral inhibition to AN I (Faulkes and Pollack, 2000). When AN1 is stimulated by sound from the soma-contralateral side it receives inhibitory inputs from the ON1 on the right, which receives sound from the ear on the opposite side. The longer latency of the response of ON I at the species song frequency of 4.5kHz means that the first spiking response of AN1 is not influenced by inhibition from ON I. This might be interpreted as being a mechanism that enhanced interaural differences and hence directional hearing, but by adjusting the timing of the stimuli to each ear it was possible to remove artificially the latency of ON1 and show that this increased rather than decreased the latency of AN I (Faulkes and Pollack, 2000). The function of the latency in the response of ON1 still awaits an explanation, although Faulkes and Pollack (2000) suggest that a possible reason might be the fact that ON I is tuned to frequencies that evoke two different behaviours, positive and negative phonotaxis. The delay in the response of ON 1 to the cricket song frequency might be a consequence of other unrelated requirements of the nervous system. 5.2.4

Auditory interneurons in the mole cricket

In a comparison of two mole cricket species, one winged (Seapteriseus borellii) and one flightless (S. abbreviatus), differences in the responses to highfrequency sound were found in identified neurons. The auditory interneurons are similar in anatomy to those found in other ensiferans. Two classes of omega neurons (Wohlers and Huber, 1982; R6mer et ell., 1988) are found in the prothoracic ganglion, although they are not distinguishable morphologically. Anatomically, they resemble the ON1 neurons of the true crickets. The two types differ in their frequency responses. While both have a peak sensitivity at around 3kHz, one type has a second peak in the 20 30kHz range (Fig, 15).

204

D.J. ROBINSON AND M. J. HALL

100 [ A a,23 80 } , -sin 7 0 ,

~

; T,!~'~'-~ ,,\

;~

. i2ii

[

40 30 ~

. . . . . . . . . . . Frequency {kHz}

FIG. 15 The response characteristics of two types of omega neuron in NC(ll?teri,s'cus borellii, showing the difference between a low-frequency tuned neuron and a high frequency tuned one (Mason et al., 1998, fig. 3A). Reprinted with permission from the Company of Biologists.

One auditory unit was recorded (Mason et al., 1998) that had a single peak of sensitivity in the ultrasonic range. This was a high frequency T-neuron, which had a low threshold between 20 and 30kHz but with a wide range, showing strong responses at 70kHz (Fig. 16). Recordings from the cervical connectives showed short-latency responses (15 20 ms) to both audio and ultrasonic acoustic signals, when signal averaging techniques were used. The responses to audio stimuli (3 kHz) were of greater amplitude and duration than those to ultrasonic stimuli (25 kHz), suggesting

I00[ A / m 80 =

>TAT Lt

60 b- 50 4{)

Frequenc> {kHz) FIG.

16 The mean response characteristics

of high-frequency T-neurons in

Scapteriscus borellii (Mason et al., 1998, fig. 6A). Reprinted with permission from the

Company of Biologists.

SOUND SIGNALLING IN ORTHOPTERA

205

that multiple units carried the audio frequency information, but that only one or at most a few units carried the ascending ultrasonic responses. Longer latency responses were detected to ultrasound, but these were likely to be in the descending fibres and related to the ultrasound startle response (Hoy el al., 1989). In S. abbreviatus, which is flightless, the omega neurons had similar tuning and response properties to those observed in the low-frequency omega neurons in S. horellii. One example of a high-frequency T-cell was recorded and it was homologous with the high-frequency T-cell in S. abbreviatus, but with thresholds of greater than 80dB SPL over the whole range of frequencies tested. The presence of responses to high-frequency in the night-flying S. borellii but not in the flightless S. abbreviatus, suggests that the auditory system in S. horellii plays a part in the avoidance of predators (see section 7.3). The two pairs of omega neurons found in the mole crickets are probably homologous with those of crickets, where two pairs are also found (Wohlers and Huber, 1982). Two pairs of omega neurons are also found in haglids (Mason and Schildberger, 1993), in contrast to the tettigoniids where only one pair has been found in the species studied so far (R6mer el al., 1988; Schul, 1997).

5.2.5

Audilo W interneurons in ,grasshoppers

The auditory pathways in the nervous system of grasshoppers have been studied extensively. There have been a number of reviews of auditory processing that include sections on the auditory interneurons, notably those of Stumpner and yon Helversen (2001), while the auditory processing of information from chordotonal organs is reviewed by Field and Matheson (1998). Excitatory and inhibitory inputs to two ascending interneurons were identified by Marquart (1985). Subsequently, Boyan (1991, 1992} showed that there was a common synaptic input from one of the cells recorded by Marquart onto other ascending auditory interneurons. One of these post-synaptic cells was the G neuron, also known as neuron 714, and this neuron is perhaps one of the best-characterized neurons in insects. A recent study of this morphologically prominent neuron by Boyan (1999) has demonstrated that the level of excitability of the neuron depends upon the gating of modulating inputs, including auditory and vibratory sources. Eight auditory interneurons are known to be presynaptic to neuron 714 and all are excitatory. However, it is not yet known how many of the receptors in the ears themselves connect synaptically with neuron 714. Those that do ascend to the metathoracic ganglion to synapse with neuron 714 seem to be predominantly high frequency (15 kHz) ones (Halex et al., 1988). The connections between identified cells in the metathoracic ganglion and neuron 714 in the mesothoracic ganglion are

206

D.J. ROBINSON AND M. J. HALL

summarized by Boyan (1999) and shown in Fig. 17. Neuron 714 is one of several interneurons that provide multimodal outputs in the form of combined information from more than one sensory source, and it is part of the neuronal chain that triggers jumping in locusts (Gynther and Peerson, 1989). The extensive range of both inputs and outputs to this neuron highlight its importance as a component of the locust nervous system and it is clearly a neuron which will repay deeper study in the future.

5.3 SYMMETRYANDASYMMETRY The sound-producing apparatus in Orthoptera shows asymmetry both in anatomy and in motion (for examples see Ewing, 1989). However, orthopteran ears would be expected to be symmetrical as directional information would be derived from the differences between the signals from the left and right ears. However, Rheinlaender and R6mer (1980) found asymmetry in the ears of the bushcricket Tettigonia viridissima when they investigated the responses of the T-cells, particularly at frequencies below 12kHz and above 20kHz. Of the

mesothoracicganglion

I

J,

.... t "£1:ta;h;;aci:;;gl~;n """ "~~

"" I'1" 1"] I

aft ~ " "

\-@ --@

~, @----....--~@1

,@ ,@,

,(

FIG. 17 A summary of the synaptic connections to neuron 714 in the locust mesothoracic ganglion, from cells in the metathoracic ganglion. The numbers in the circles refer either to identified interneurons or to the likely number of interneurons that make up the pathway, based on the measurement of synaptic delays. Arrows indicate an excitatory connection, filled circles an inhibitory one. Auditory afferents make a monosynaptic connection with neuron 714. Reprinted with permission from Boyan, G. S. (1999). Presynaptic contributions to response shape in an auditory shape in an auditory neuron of the grasshopper. J. Comp. Physiol. A 184, 279-294, Springer-Verlag GmbH & Co. KG.

SOUND SIGNALLING IN ORTHOPTERA

207

pairs of T-cells whose thresholds they measured, 53% showed some degree of asymmetry, explained in their view by a loss of function in single primary afferents. Boyan (1979) found an even larger proportion of individuals showing asymmetry of spiking responses in neurons in his study of Teleog~Tllus commodus. These results are puzzling when considering the importance of directionality, particularly in phonotaxis, and a recent study (Faure and Hoy, 2000a) has revisited the problem of the degree of asymmetry in the Tcell response, using Neoconocephalus e,siger. Of 32 males tested with a loudspeaker at 0 <' or 90" to the long axis of the animal, only three showed an asymmetry in their thresholds that was significantly greater than variation due solely to measurement error, all at 90'L Of 16 females tested, only three showed significant asymmetry, two at 9 0 and one at 0'. Overall, this study found that 87 92% of individual pairs of T-cells were bilaterally symmetrical, and that the symmetry covered the whole frequency range of the ear. In contrast, Rheinlaender and R6mer (1980) suggested that within the key fi-equency range for detecting the species song, symmetry was greatest. The reason for the difference between the two studies probably lies in the treatment of measurement error, although the studies were of different species. Rheinlaender and R6mer (1980) were conservative in their analysis, excluding animals where there was a difference of greater than 2dB between two measurements of threshold under the same conditions. Faure and Hoy (2000a) found that the measurement error was 2 3 dB in their study. Thus excluding animals at the 2dB level may have reduced the apparent size of the measurement error and thus enhanced asymmetry. An additional piece of evidence supporting the idea of symmetry comes from a study of the omega neuron (ON 1) of Tettigonia viridissima, which showed that of ten pairs of neurons studied, eight had thresholds that did not deviate by more than 2 dB from each other at any frequency. It appears, from the more recent study of Faure and Hoy (2000a), that symmetry may indeed be a feature of the auditory system, but we must await confirmation from other species that it is a widespread feature of orthopteran hearing.

6

Information content of signals Communication can be defined as the transfer of information from one organism to another, detectable to an outsider only through the appearance of an observable response in the receiving individual. (Alexander, 1967).

Johnstone (1997) has recently reviewed the evolution of animal signals. The most important factor in signal evolution has probably been the reciprocal nature of selection: the properties of the receiver exert strong selection on signal design, and signal design in turn exerts strong selection on receiver behaviour. Thus, on the one hand, selection favours signallers who elicit

208

D.J. ROBINSON AND M. J. HALL

favourable responses, while on the other it favours receivers who can accurately interpret the nature and intentions of senders from their signals. To be effective in eliciting favourable responses, a signaller must use displays that are well suited to detection by the sensory and neural mechanisms of intended receivers (called 'sensory exploitation' by Ryan, 1990), but at the same time unsuitable for detection by unwanted eavesdroppers such as predators. All members of the 'audience', not just intended receivers, can therefore influence the evolution of signalling behaviour. As far as the signaller is concerned, the signal may be a way of manipulating others, for example a means of stimulating a female to mate, but as far as the receiver is concerned, it is solely a source of information. Selection will favour receivers who can best relate their behaviour to the information contained in the signal and at the same time ignore signals that do not provide relevant information. As a result of the selection pressures on signallers, displays will be designed to produce a response efficiently; as a result of the selection pressures on receivers, displays will be designed to prevent them being 'dishonest' (i.e. unreliable). Many display properties can thus be explained in two ways. The high cost of some extravagant signals, for example, can be explained because they make the signal more detectable or stimulating to receivers. On the other hand, if signals are costly to produce, then only signallers of high quality or signallers who stand to gain a lot by getting a favourable response should produce them. According to this 'handicap principle', therefore, honesty can be maintained by costly signals (Zahavi, 1975). From the evidence so far, Johnstone concludes that the detailed design of a signal is more dependent on the environment in which communication takes place, and the audience to which it is directed. In other words, signals must be both efficient and reliable, but the need for efficiency appears to place the greatest constraint on their design. Producing a costly signal may also increase the efficiency of communication; even in an apparently cooperative signalling system where signallers gain by producing a response and receivers gain by responding, there will be conflicts because signallers have to balance efficient communication against the possibility of attracting eavesdroppers, while receivers have to balance the efficiency of responding against the possibility of responding inappropriately, for example to an extraspecific signal (Johnstone, 2000). An orthopteran trying to obtain a mate using acoustic signals must provide information about its species, its location and its suitability as a mate. The physical environment through which its signal has to pass affects the transmission of information to the recipient. The signals of other conspecifics or other species may also obscure the sender's signal. In addition, the sender has to avoid the attentions of unwanted eavesdroppers, for example rival males or predators. The problems of eavesdroppers and interference from other signallers are considered in Sections 7 (predators and parasitoids) and 8 (competition with conspecifics); the information content of the signal and the effects of the physical environment are considered here.

SOUND SIGNALLING IN ORTHOPTERA 6.1

209

C O M P O N E N T S OF SOUND SIGNALS

A sound signal has five main parameters: amplitude (i.e. loudness, measured as sound pressure level); frequency or wavelength spectrum (frequency is equal to the reciprocal of the wavelength); frequency modulation (e.g. a sound may rise in pitch and then fall again); amplitude modulation (e.g. a sound may start loudly, then fade away); and temporal patterning (length of pulse, length of interpulse interval, number of pulses, bout length of a series of pulses, etc). The various parameters of signals reveal different things about the sender, some of which are not under the sender's control (Endler, 2000). For example, dominant wavelength and wavelength spectrum are determined by the biomechanics of the sender, and biomechanical properties are related to size. The energetic cost of the signal is also directly related to its components: for example, costs increase the louder the call, the faster the frequency or amplitude modulation, the longer the pulse length, the shorter the interpulse interval and the longer the bout length (see section 3.8), The signal received by the recipient may differ from the sender's original because of changes that occur during transmission through the environment. Some of the resulting call features can also provide information about the sender. For example, the greater the distance between signaller and recipient, the less coherent the received call is likely to be (i.e. the more echoes it is likely to have that are out of phase with the original: see section 6.2).

6.2

EFFECTS OF THE E N V I R O N M E N T ON SOUND SIGNALS

Sound waves propagate from their source and, in open air, spread spherically and attenuate by the inverse square law (Bennet-Clark, 1998). They travel faster at higher temperatures. The effects of the environment on signals have recently been reviewed by Endler (2000). Earlier reviews include those by Forrest (1994), R6mer (1992) and Michelsen (1985), while Pye (1979) has considered the particular problems and advantages associated with the use of ultrasound in signalling. The environment has two main effects: it modifies signals and masks or distorts them by adding noise (any unwanted sound that interferes with the detection of a signal and the transmission of the information it contains, including environmental noise such as wind and turbulence and biological noise such as sounds from other animals: see Forrest, 1994). Environmental noise is usually low frequency, with most energy below 2 kHz (Forrest, 1994). The signal-to-noise ratio (in its simplest sense, this is the difference in dB, between the level of the background noise and the level of the signal: see Forrest, 1994) will therefore be higher for signals over 2 kHz, so highfrequency signals are inherently easier to detect. Insects, being small, are less efficient sound radiators of low-tYequency sound. There is thus a trade-off

210

D.J. ROBINSON AND M. J. HALL

between producing low-frequency signals that travel farther and highfrequency signals that are produced more efficiently. The problems of added noise can be eliminated to some extent by the tuning of the receiver's auditory system. Some orthopteran signals are confined to narrow frequency bands, and if the auditory system is tuned to the specific frequency of the signal, then the actual signal-to-noise ratio for the receiver will be higher than that measured in the field using a broad-band microphone (Forrest, 1994). If the environment contains objects, such as plants, which reflect the signal and these reflections reach the recipient after various delays, then the combination of original and reflected signals can be so different from the original, both in frequency spectrum and in temporal pattern, that it is unrecognizable (Michelsen, 1985; Endler, 2000). For example, Simmons (1988) showed that the temporal pattern of the calls of the field cricket GJ;vllus bimaculatus became significantly degraded over distances of only 2 m. The intervals between pulses and between chirps decreased because echoes filled in the parts of the call that should be silent. The only parameter transmitted reliably in such a reverberating signal may be the start time of each pulse (Endler, 2000). Reverberation is more of a problem at higher frequencies (shorter wavelengths) because objects reflect wavelengths that are the same size or less, and there are more small objects in the environment than there are large ones. Insects are constrained by their size to produce relatively high frequencies, and choosing a good signalling position free of obstructions may be their best solution to reverberation (Endler, 2000). Some of the adverse effects of reverberation, noise and other interference can be countered by processing in the nervous system of the receiver, for example by averaging repetitive signals (Endler, 2000). Any noise uncorrelated with the signal cancels out if the repeated components of a signal are averaged. Because there is more interference at higher frequencies, and insects' signals tend to be high frequency, insect calls tend to be repetitive (see, for example, BennetClark, 1998). Amplification or filtering within the nervous system can also increase the signal-to-noise ratio, and signal redundancy can reduce noise (Endler, 2000). For example, pulse length and song bout length both provide information about energy expenditure, but the noise associated with each of them is not correlated and can be removed during processing by comparison of the two parameters. A constant sound broadcast from a speaker shows irregular amplitude fluctuations after travelling through turbulent air or past moving objects, and these fluctuations are likely to be greater the higher the frequency of the sound (see R6mer, 1992). R6mer has shown, however, that such amplitude fluctuations decrease with increasing bandwidth. Many orthopterans use broad-band signals simply as a consequence of the sound production mechanism they use, but the broad frequency spectrum may also enable them to communicate information reliably using amplitude-modulated temporal patterns of calling.

SOUND SIGNALLING IN ORTHOPTERA

211

As a result of all the factors that cause degradation as a sound signal passes through the environment, the only cue that is likely to survive is the gross temporal pattern (R6mer, 1992). It is probably because temporal patterning is the most reliable feature of a call, that species recognition (see section 6.3) and mate choice (see section 6.6) often rely on parameters such as interpulse interval or length of song elements. Sound signals also suffer from attenuation as well as degradation as they travel through the environment (see Forrest, 1994: Kalmring et al., 1997). Excess attenuation (i.e. over and above that due to spherical spreading) can be caused by atmospheric absorption, absorption by vegetation, scattering by diffraction or reflection, refraction by wind or temperature gradients, or boundary conditions. Higher frequency sounds attenuate faster than do lower frequency sounds. This degradation can be avoided to some extent by sending signals containing wavelengths that attenuate less than others or for which the background noise is minimized (as, for example, Sciarasaga quadrata does: see R6mer and Bailey, 1998), aiming the signal in a particular direction to minimize spherical loss (as in mole crickets: see Bennet-Clark, 1998), or choosing optimal times or microhabitats for transmission (Endler, 2000). For example, bladder grasshoppers, Bullacris membracioMes, maximize their calling range by calling at night, as described in section 3.1 (van Staaden and R6mer, 1997). Many orthopterans call from on or near the ground, a boundary which has a large effect on attenuation (Forrest, 1994). For example 'softer' ground (i.e. more porous surfaces) attenuates a signal much more than 'harder' ground (i.e. less porous, more reflective surfaces such as water), and a given surface attenuates high frequency sounds more than low frequency ones. Elevating the signaller and/or the receiver decreases the attenuating effect of the ground. For example, R6mer (1992) measured the intensity of Tettigonia viridissima calls at different heights in the field. A signalling male and the recording microphone were placed at the same height, 10-20m apart. He found that, at an elevation of 1.5m, recorded intensities were 10dB greater than those measured at an elevation of 0.75 m. At 2 m above the ground, attenuation was even less and approximated the expected loss due to spherical spreading. R6mer suggests that the boundary effect of the ground may be one of the reasons why many orthopterans prefer if possible to call from elevated perches (although another reason may be to avoid the scattering and absorption by vegetation described above). Attenuation is not necessarily disadvantageous, however. Rapid attenuation benefits the signaller if the distance between the sender and the intended receiver is shorter than between the sender and an unwanted recipient such as a predator (see section 7). It can also provide the recipient with information about the distance to the signaller, since the ratio of the intensities of two or more wavelengths of the signal will be proportional to the distance from the sound source (Endler, 2000).

212

6.3

D.J. ROBINSON AND M. J. HALL SPECIES RECOGNITION

In many orthopterans, male calls are species specific and species recognition often depends on the temporal pattern of the conspecific song (e.g. Dobler et al., 1994a). Auditory neurons in the central nervous system of gryllids, tettigoniids and acridids have been shown to code the temporal parameters of the song (section 5). Rate of singing is dependent on ambient temperature: for example, the warmer it is the faster a cricket chirps (section 3.7). If species recognition is to depend on the temporal pattern of calling, then the ability of a female to recognize a song must also be related to the temperature; in other words the pattern that a female prefers at a particular temperature must be the same as a male produces at that same temperature. Such temperaturedependent behavioural coupling has been demonstrated in several species, including Gwllus bimaculatus and G..Brmus (Hoy, 1992b). In bushcrickets, the songs of closely related species are often very similar but with some species-specific differences in temporal pattern that allow discrimination (see, for example, Schul, 1998; Stumpner and Meyer, 2001). Stumpner and Meyer found that four duetting species of Barbitistes bushcrickets are similar in the carrier frequency of their song and the shape of a single syllable (the sound produced by one wing closure), but that they all differ in the temporal patterning of their song. Species differ in the number of syllables in a chirp, the syllable period (length of time from the start of one syllable to the start of the next), the trigger period (length of time from the start of the last syllable of a chirp to the start of the "trigger syllable', i.e. the single syllable that follows a relatively long time after each chirp and which elicits a female reply), in such a way that each species has a different combination of parameters. Females could therefore discriminate between conspecific and heterospecific males on the basis of the temporal pattern of their call. For males, however, the only cue that a female is conspecific comes from the fact that she replies to his call within a certain time window: female calls in the four species are all very similar in frequency, structure and latency to reply. But the system does seem to work: hybrids between sympatric species of Barbitistes have never been found in the wild. The response delay of the female to the trigger syllable and the time window in which the male recognizes the female response have been measured in several other duetting phaneropterine species (Heller and von Helversen, 1986). Within each species, the delay and the window show very close matching, but, between species, there is considerable variation. In other words, the coupled female response delay/male window times are species specific and incorporate into the calling system an additional species recognition feature that may compensate for the limited amount of temporal patterning that can be contained in the relatively short male call typical of these species. Crickets tend to rely on the temporal patterning and sometimes the frequency of the advertisement call in species recognition. This has been

SOUND SIGNALLING IN ORTHOPTERA

213

reviewed by Doherty and Hoy (1985); see also, for example, Doherty and Callos (1991), Doherty and Storz (1992), Stumpner and yon Helversen (1992) and Hennig and Weber (1997). Hennig and Weber looked at two sympatric species, Teleogryllus commodus and T. oceanicus, which both have calls containing chirps and trills (where a chirp is a series of syllables separated by relatively long intervals, a trill is a series of syllables separated by a relatively short interval, and a syllable is one wing closure). In both species, calls consist of phrases, each containing a single chirp followed by a number of trills, with both chirps and trills of constant amplitude. The two species differ, however, in that they have different syllable periods (a syllable period being the time from the start of one syllable to the start of the next) in the chirp and trill parts of their calls, and a different number of syllables in their trills. Hennig and Weber found that T. c'ommodus females responded only to calls containing the species-specific syllable period during both the chirp and during the trill, while T. oceanicus females would respond as long as the syllable period in the chirp and the frequency of the call were appropriate to their species. For T. commodus, repetition is also important: the chirp and trill parts of the call must each contain a minimum of three to five consecutive syllable periods to elicit a female response. T. oceanicus, on the other hand, can recognize a conspecific on the basis of only one or two "correct' syllable periods. For T. commodus, species recognition is largely based on central nervous processing, while, for T. oceanicus, both peripheral frequency filtering and central temporal filtering are important. Courtship song is also important in species recognition in crickets. In T. oceanicus, for example (Balakrishnan and Pollack, 1996), the courtship song, like the advertisement song, has a carrier frequency of 4 5 kHz and consists of chirps and trills with a chirp syllable period of about 65 ms. It is different from the advertisement song in several ways: the courtship chirp is amplitude-modulated, slightly longer (about nine compared with about five syllables) and followed by a single, very long trill with a slightly shorter trill syllable period (about 30ms compared with about 37ms). Females are reluctant to mount males that do not produce courtship song and, as in the advertisement call, respond only when the frequency of the song and the temporal pattern of the chirp are appropriate to their species, thus relying on the only two features that are consistent between advertisement and courtship song. Species do vary, however, in whether recognition of the conspecific song is absolute or relative: in some cases, only the conspecific song is attractive and heterospecific songs are rejected even if there is no other choice: in other cases, both conspecific and heterospecific signals fulfil the requirements for a response, but the conspecific song is usually preferred in a choice situation (Schul, 1998; Schul et al., 1998). However, in some species, the preference for conspecific song is relatively weak. In Tett&onia cantans, for example, a female will perform phonotaxis towards a T. virMissima male if he is closer and

214

D.J. ROBINSON AND M. J. HALL

therefore perceived as being louder; hybrids are rarely found in sympatric populations, however, and reproductive isolation may be maintained by close-range mechanisms and a partial temporal separation in the breeding season for the two species (Schul el al., 1998). Gwynne and Morris (1986) have even suggested that some species achieve conspecific matings not by recognizing and responding to the conspecific song but instead by recognizing and not responding to closely related sympatric species that present a risk of hybridization. Conocephalus n~ropleurum, for example, shows little if any discrimination between conspecific calls, the calls of congeneric allopatric species, and even random noise, but will not respond at all to the call of the sympatric C. brevipennis, a species which shares its habitat and has a very similar lifehistory (Morris and FuIlard, 1983). Vibration signals may also play a part in species recognition (see review by Kalmring and Kfihne, 1983). Stridulating insects produce both airborne and substrate conducted sound and, in tettigoniids and acridids, all known auditory and vibratory ventral cord neurons ascending to the brain receive inputs from both receptor systems, i.e. the tympanal organs and the vibration receptors. The coding of the conspecific song in the responses of most of the ventral cord neurons is considerably improved when the stimulus consists of both the airborne sound signals and either a maintained vibration or vibration matched to the temporal structure of the song. Some species are unable to discriminate between conspecifics and closely related sympatric species on the basis of their songs. For example, females of the ground crickets Allonemobius.fitsciatus and A. socius show no preference between the calls of either species, even though they differ in carrier frequency and chirp period (Doherty and Howard, 1996). Hybridization between the two species in areas of sympatry is rare and appears to be prevented by a postinsemination barrier to fertilization. Doherty and Howard suggest that it is to the female's advantage not to discriminate between conspecific and heterospecific males in these two species because there is little danger of producing a hybrid, females mate several times so will probably find at least one mate of the right species, energy expended in approaching the male and the risk of predation while doing so are both likely to be low, and the female benefits from courtship feeding by the male. Since song has such a vital role in species recognition and reproductive isolation in many orthopterans, it is likely to have played an important part in speciation. In some cases, calling songs are the only identified characteristics separating closely related species, for example in Oecanthus (Toms, 1985), Chorthippus (Ragge, 1987), and Laupala (Otte, 1994), suggesting that acoustic signals can be among the first features to change during the speciation process. Such change could be due to, among other things, sexual selection resulting from female mate choice. For example, Shaw and Herlihy (2000) have shown that female preference for faster than average pulse rates is exerting a selective pressure on male song in the Hawaiian cricket Laupala cerasina. Pulse rate has

SOUND SIGNALLING IN ORTHOPTERA

215

been the most labile song feature in the evolutionary history of the Laupala genus (Otte, 1994). 6.4

R E C O G N I T I O N OF SEX

In species where both males and females call, individuals have to be able to identify the sex of a caller. It is not adaptive to waste time and energy responding or performing phonotaxis to a member of the same sex. In most duetting bushcrickets this is not a problem because the female song is very different from that of the male in temporal pattern and sometimes in frequency (e.g. Ancistrura nigrovittata: Dobler et al., 1994b), but in many gomphocerine grasshoppers, males and females appear to have similar songs. Von Helversen and von Helversen (1997) have shown, however, in Chorthippus biguttulus, that in many respects the temporal patterns in the calls show no overlap between the sexes, allowing reliable discrimination of potential mates. For example male calls consist of continuous syllables longer than 30ms, without gaps, while female calls consist of "gappy' syllables with pulses shorter than 15 ms. Males and females also have differently tuned auditory systems and use differences in the frequency spectra between male and female calls in sex recognition. No differences with respect to frequency sensitivity, threshold, or temporal properties have been found between males and females in either the tympanal receptor elements or the ascending metathoracic neurons, so the sex recognition system in this species must lie at the level of the brain. The properties of some ascending auditory neurons indicate they are likely to play a role in this process: for example, the neuron AN3 responds more to gappy syllables than to continuous syllables (Stumpner et al., 1991). 6.5

MATE LOCATION

The mechanisms of directional hearing described in section 4.4 should, in theory, allow the sex that performs phonotaxis to locate the position of its potential mate so that it can move towards it. There may still be problems, however. In duetting phaneropterine bushcrickets, for example, the only cue that the male has to help him locate the female is an extremely brief click, delivered at long intervals. Von Helversen and Wendler (2000) found that male Poecilimon q[linis can locate a female on the basis of her acoustic response alone, even though each response lasts only 0.3 ms (Heller and yon Helversen, 1986), but they follow a very roundabout route to reach her. However, the poor quality of the female song as an acoustic beacon is partly compensated for by the use of visual cues. When performing phonotaxis, a male typically walks for a few seconds, stops, calls, and then continues walking. When he receives a female reply, he turns to bring the sound source within his frontal auditory field. Males perform reliable phonotaxis only to female responses occurring between 40 and 170 ms after the end of their own call, demonstrating the

216

D.J. ROBINSON AND M. J. HALL

existence of a time window in this species. Von Helversen and Wendler used a walking compensator to compare phonotaxis with and without visual cues. As the response of the female is so short, the response of the male in the absence of visual cues is an open-loop one, with the male turning towards the source and usually overcompensating, then over-correcting to the next call. Phonotaxis proceeds with the path followed by the male being essentially an oscillation about the axis of the female, resulting in a course made up of circular arcs. When the male walks within a cylinder whose sides have regular stripes, his course is somewhat more direct, but when a single black stripe is provided as a visual cue, his course is much more direct. So, if the male has landmarks in the environment, the visual cues transform an open-loop response into a closedloop response. Von Helversen and Wendler showed that this was because males used the female's first response to establish her direction and then used the visual landmarks to maintain a straight course. They suggest that a stabilizing influence on course maintenance by visual inputs could also be important in other bushcrickets and in gryllids. In duetting bushcrickets, the basic ensiferan neuronal processing system by which song is recognized and located may also impose constraints on the ability of the male to locate his mate. Poecilimon ornatus males have been shown to steer an intermediate course between two females if they both respond to the same call within the time window, even if the responses differ greatly in intensity and are separated by up to 60ms (von Helversen et al., 2001). Thus, in contrast to other ensiferan species, P. ornatus males do not show a precedence effect (see section 6.6). Von Helversen et al. suggest that the poor orientation ability of P. ornatus males in situations where more than one female responds may represent an evolutionary trap for duetting phaneropterines. They argue that the evolution of duetting allowed females to shift the risk of phonotaxis to males and also allowed males to reduce their own risk of attack from acoustically orienting predators or parasitoids by dramatically reducing their duty cycles. Because the female response consists of only a very brief unpatterned click, however, response time became the essential feature for species recognition. Given the serial processing of recognition and localization in ensiferans, this in turn hindered the ability of the male to perform selective phonotaxis. The ability of a female to locate her mate can be influenced by ear size: female Kawanaphila nartee with the largest spiracles (and therefore the greatest hearing sensitivity) have the greatest mating success (Gwynne and Bailey, 1999). In species where males sing and females perform phonotaxis, there are also different selection pressures on the hearing system in males and females. Bailey and Kamien (2001) showed that the hearing system of Requena verticalis is sexually dimorphic, with females having larger spiracles than males and therefore more sensitive hearing. Females also show less variation than males in hearing threshold and bulla volume, and bulla size is related to the dimensions of their thorax. Larger males, with larger wings, have

SOUND SIGNALLING IN ORTHOPTERA

217

proportionately more stridulatory muscles in their thorax, and bulla size is not related to thorax size. There thus seems to be a trade-off in males between hearing sensitivity and sound production. For the male, singing is more important than hearing because he does not perform phonotaxis. For the female, hearing is important because she needs to recognize and locate her mate; Bailey and Kamien suggest that there is stabilizing selection on ear size in females as a result. 6.6

MATE CHOICE

There is increasing evidence in a wide range of orthopterans that mate choice contributes to individual differences in mating success and that male song can provide cues on which female choice is based. Gwynne (2001) discusses the reasons why females may mate with certain males over others, and gives a number of bushcricket examples of female choice based on song. Zuk and Simmons (1997) and Brown (1999) review the evidence for female mate choice in gryllids and tree crickets (Oecanthinae) respectively. Male mate choice has been shown to occur in an acridid (Melanoplus sanguinipes), a gryllid (Acheta domesticus) and a number of tettigoniids, but is based on chemical or tactile cues; male mate choice based on female song has not yet been demonstrated (for a review of male mate choice in insects see Bonduriansky, 2001). Mate choice based on the male song can be "passive' or 'active' (Parker, 1983). In the former, females do not show any preferences between particular males but are simply attracted passively to the male call. As a result, the male whose signal reaches the most females, or is the most stimulating, or the easiest to locate, will attract the most mates. This could be the male who sings most often, for example, or that has the most intense or longest call, or the call that travels the longest distance, or the call frequency most closely tuned to the female auditory system, or the temporal pattern that is most stimulating. Choice that is purely passive could therefore be described as a form of male male competition to reach females, rather than choice per se (Brown, 1999). Females of several species, when given the choice of two otherwise identical male calls, preferentially perform phonotaxis towards the leading call, for example Eigurotettix planum (Minckley and Greenfield, 1995), and Neoconocel)halus aTfiza (Snedden and Greenfield, 1998). N. spiza females prefer the leading call even when the following call is longer, or louder by up to 4dB, which rules out passive attraction to the male that simply sings more (Snedden and Greenfield, 1998). Minckley and Greenfield (1995) suggest, however, that the preference in k. planum may be passive choice due to 'sensory bias' (Ryan and Keddy-Hector, 1992). (For a review of receiver biases and their effect on signal evolution, see Basolo, 2000.) The mechanisms involved in the preference are not clear. The transition from silence to sound at the start of a call could be a critical feature of the call's attractiveness and/or it could aid localization of its source, and the onset of the following call could just be masked by the

218

D.J. ROBINSON AND M. J. HALL

leading call. Snedden and Greenfield (1998) showed, however, that this could not be the only explanation for the female preference in N. spica. Alternatively, the preference might result from a 'precedence effect' (Wyttenbach and Hoy, 1993) in which the first signal suppresses the animal's peripheral or central response to the second. R6mer and his colleagues have shown that stimulation at the ipsilateral ear may cause inhibition at the contralateral ear so that later stimuli do not reach threshold (R6mer, 1993; R6mer et al.. 1997). Wyttenbach and Hoy (1993), on the other hand, looked at directional ultrasound avoidance behaviour in Teleogo,llus oceanicus, which also shows a precedence effect. He found no evidence of ipsilateral--contralateral differences in the responses of a bilateral pair of ascending, second-order auditory interneurons known to initiate ultrasound avoidance, and concluded that interactions in the brain must be responsible for the precedence effect in this case. Just because choice is passive, it does not necessarily mean that it is not adaptive. In the desert clicker, L~gurotettix coquilletti, males defend creosote bushes, which vary in nutritional value, against other males. Females simply approach the most intense signal and this tends to take them to bushes of high nutritional value occupied by groups of males (Bailey et al., 1993a). The females may benefit both from the good diet provided by the bush and from the easy availability of several potential mates within a small area, which decreases the costs of moving between them. In active choice, females show a preference between individual males, rejecting some in favour of others. Choice is adaptive in that females obtain immediate benefits, either for themselves or their offspring, or genetic benefits (Andersson, 1994). Immediate benefits include food (e.g. courtship feeding) or a low-cost copulation (e.g. with reduced predation risk). Genetic benefits can result either because the male has 'good genes', which give his offspring a better chance of surviving, or because the female has 'sexy sons', i.e. her male offspring inherit the male's attractive trait and in turn have high mating success. For orthopterans, male songs provide a useful basis for active female choice because, being both variable and costly to produce (see section 3.8), they are likely to be good ('honest') indicators of the quality of potential mates. Females can also assess each call, or several calls simultaneously, from a distance, so the cost of rejecting a male should be low. Females have been shown to be less choosy when the costs of choice are higher, for example when predation risk is greater (Hedrick and Dill, 1993). Although female choice based on song has been demonstrated in several species of orthopterans, it is difficult to show definitively that choice is active rather than passive, or whether the benefits are immediate or genetic. For example, in laboratory experiments, females of the bushcricket Conocephalus nigropleurum preferentially approach the tape-recorded sounds of a group of ten males, or the sound of just two males singing together, rather than a single calling male, but this could be because females discriminate in favour of groups of males per se, or because they are simply passively attracted to the most

SOUND SIGNALLING IN ORTHOPTERA

219

intense signal (Morris et al., 1978; Morris and Fullard, 1983). Females of some species may simply approach the nearest/most easily located male. For example, Kawanaphila nartee females may prefer males with higher-frequency songs because, as a result of the more rapid attenuation of higher frequencies, they perceive them as being nearer than males with lower-frequency songs (Gwynne and Bailey, 1988). Females of many species prefer the loudest call if all else is equal (see examples in Brown, 1999). This could, however, be the result of passive attraction to the loudest signal, or active choice for the nearest male or the male with the greatest power output. Approaching the nearest male benefits the female because it reduces the energy costs and predation risk associated with performing phonotaxis. The male with the greatest power output is likely to be a high-quality mate because producing a louder song requires more energy (Forrest, 1991; Brown, 1999). Forrest and Raspet (1994) provide a method for distinguishing among the possible decision rules that females might use to choose between males calling at different distances and with different power outputs, but it has not yet been applied to individual species. Female choice has been shown to depend on various parameters of the male advertisement call in different species. Some recent examples are given in Table 2. In many cases, the parameters of male song preferred by females have been shown to correlate with male size, for example in Gryllus bimaculatus (Simmons, 1988), Scu&teria curvicau&~ (Tuckerman et al., 1993), Oecanthus ,iwicol"nis (Brown et al., 1996), Acheta domesticus (Gray, 1997), Phaneroptera nana (Tauber et al., 2001) and Requena verticalis (Schatral, 1990). The benefits from mating with a large male may be direct or indirect. Larger males may provide more or better-quality courtship feeding, and the extra nutrients may increase female fecundity or offspring survival. Proteins in spermatophores have been shown to enhance the fitness of females in some orthopteran species (see. For example, Gwynne, 1988), but in others, no effect of spermatophore consumption on the female's reproductive success has been found (see, for example, Weddell and Arak, 1989). If size is inherited, females mating with a large male will have large offspring; the daughters are likely to be more fecund and the sons to be more attractive as mates. Body size has been shown to be heritable in several species: for example, G. bimaculatus (Simmons, 1987), G. pennsvlvanicus (Simons and Roff, 1994), Acheta domesticus (Gray, 1997), and Poecilimon veluchianus (Reinhold, 1994). For example, in the black-horned tree cricket, Oecanthus n(gricornis, females prefer rnales with lower frequency songs (Brown et al., 1996). Brown et al. have shown that song frequency is a reliable indicator of male size and females mating with larger males have higher fecundity. Fecundity and reproductive lifespan are at least partially resource limited and larger males probably provide better-quality courtship feeding, which in this species is in the form of a glandular secretion (Brown, 1997). Females of the bushcricket Amblyco~37~ha parvipennis prefer males whose call phrases are louder, longer, and leading: sound level and ability to lead are correlated with male weight (although

220

D.J. ROBINSON AND M. J. HALL

TABLE 2 Some parameters of the male advertisement call for which female choice has been demonstrated. Results from laboratory experiments unless otherwise stated

Parameter of call Lower frequency

Example of species showing female mate choice based on parameter

Requena verticalis Oecanthus nigricornis Higher frequency Kawanaphila nartee Amblycowpha parvipennis Loudness (intensity) Requena vertica]is Higher chirp rate Grvllus [ineaticeps More syllables per chirp Acheta domesticus or phrase (longer Go'l/us lineaticeps chirp/phrase Scudderia curvicauda durations) Gryllus texensis b More pulses per trill and shorter interpulse intervals Longer chirps Phaneroptera nana Symmetry in maximum Myrmeleotettix maculalus frequency and chirp duration between chirps produced by different legs Leading call (start of Neoconocephalus ,7)iza call not overlapped Amblycowpha parvipennir by another call) Ephippiger ephipp(ger Ligurotettix planum Li~urotetti.v coquilh, tti

Reference (Schatral, 1990)~ (Brown et al., 1996) (Gwynne and Bailey, 1988) (Galliart and Shaw, 1996) (Bailey el al., 1990) (Wagner and Hoback, 1999) (Gray, 1997) (Wagner and Hoback, 1999) (Tuckerman et al., 1993) (Wagner et el/.. 1995) Tauber et al., 2001) Moiler, 2001)~

Snedden and Greenfield, 1998) Galliart and Shaw, 1996) Greenfield el al., 1997) (Minckley and Greenfield, 1995) (Greenfield et al., 1997)

~'Field observations. t'G. wxensis was called G. h~teger by the author, but this population from Texas has now been recognized as a separate species (Cade and Otte, 2000). ~Laboratory experiments supported by evidence from the field.

phrase length is not) and females may benefit from mating with larger males because they produce larger spermatophores (Galliart and Shaw, 1991, 1996). Such preferences for large males may not always be absolute: Bateman el al. (2001) found that female Gwllus bimaculatus are indiscriminate for their first mating (when the costs of not mating could be very high) and only choose larger males in later matings. In several species, however, there seems to be little or no relationship between size and song parameters or mating success. For example, body size is not related to song structure in a Texas population of Gryllus integer (now called G. texensis) (Souroukis et al., 1992) or in Leptophyes punctatissima

SOUND SIGNALLING IN ORTHOPTERA

221

(Kilduff, 2000). Mating success is not related to male size in sagebrush crickets, Cvphoderris strepitans, and seems linked instead mainly to the number of nights in which a male is able to sing (Snedden, 1996), implying that females are attracted passively to the male song. The relationship between song parameters, size and mating success may not be simple. Field observations by Schatral (1990) indicated that mating success in Requena verticalis depended equally on low song frequency and male size (which are correlated with each other). Laboratory experiments by Bailey et al. (1990), however, have revealed that females show no significant preference for the larger of two calling males, or for the tape-recorded song of the larger male, or for the song with the lower frequency (although the sample sizes in all these experiments were small). But females do show a significant preference for the loudest call with the greatest power in the high-frequency part of the broadfrequency song of this species, even when the difference in intensity is as little as 2dB (R6mer et al., 1998). This would tend to result in the female approaching the nearest male because higher fi'equencies attenuate faster with distance. Song parameters preferred by females have also been shown to correlate, in some species, with other characteristics of the male that may benefit the females that mate with them, including quality of diet, immune function, degree of symmetry and age. For example, female variable field crickets, Grvllus lineaticeps, prefer males who call at higher chirp rates and with longer chirp durations (more syllables per chirp) (Wagner and Hoback, 1999). Higher calling chirp rates are more energetically expensive (Hoback and Wagner, 1997) and males on a better diet chirp at a faster rate; the cost of calling is not related to chirp duration, however, and males on a better diet do not differ in this parameter (Wagner and Hoback, 1999). Females therefore select males on the basis of one nutrition dependent (chirp rate) and one nutrition independent (chirp duration) characteristic. Although they do not rule out the possibility that high chirp rates simply make males easier to locate, Wagner and Hoback suggest that females may benefit by mating with highnutrition males if these males provide better-quality spermatophores; on the other hand, there is no clear benefit to mating with males producing longer chirps, which may just be easier to locate. The number of syllables per chirp in male Acheta domesticus is a reliable indicator of immune function (ability to resist pathogens) as well as size (Ryder and Siva-Jothy, 2000). Ryder and Siva-Jothy suggest that males vary in quality and only high-quality males may be able to invest energy in both traits. The parameters of immune function that they measured are heritable, as is male size. The female may benefit directly and/or indirectly from mating with such a male because they are likely to be able to produce a second spermatophore more quickly (reducing the costs of searching for an additional mate if the female needs to mate more than once to acquire sufficient sperm or nutrients), she is less likely to acquire pathogens from him during copulation, and her offspring will inherit the favourable characteristics of her mate,

222

D.J. ROBINSON AND M. J. HALL

Cricket songs are characterized by almost pure tones of constant frequency. Individual male GJ3,llus campestris" show varying degrees of frequency modulation, however, depending on the degree of directional asymmetry in their harps, which in turn is directly related to the degree of fluctuating asymmetry in a measure of body size (Simmons and Ritchie, 1996). Larger, more symmetrical Grvl/us campestris males have higher mating success in the field (Simmons, 1995). In laboratory experiments, Simmons and Ritchie showed that females prefer pure tones of low carrier frequency, characteristic of large symmetrical harps. Honesty is maintained by the increased costs of symmetry for males with large harps: only males capable of developing large harps while maintaining symmetry will benefit from the increased attractiveness of low-frequency song. The male song may therefore encode information about male quality. Moiler (2001) has also demonstrated that females can discriminate against asymmetrical males directly on the basis of their calls. The number of pegs per row in the two hind legs of the grasshopper Myrmeleotettix maculatus shows fluctuating asymmetry, and the greater the peg asymmetry the greater the asymmetry in maximum frequency of chirps and duration of chirps produced by the two legs. In laboratory experiments, females preferred the calls of males with a larger and more symmetrical number of pegs. Evidence from the field also suggests that more symmetrical males have greater mating success. The reason why females choose more symmetrical males is not known, but there is considerable evidence that fitness is related to symmetry and females may benefit directly or indirectly from mating with a more symmetrical male (Moller and Swaddle, 1997). Females of the duetting bushcricket Ephippiger ephippiger prefer the songs of younger males (Ritchie et al., 1995). Males accumulate damage to their stridulatory apparatus as they get older, and this damage changes the structure of their song. The changes could simply impair pattern recognition by the female, or make the song less stimulating, but Ritchie et al. argue that these possibilities are not supported by the available evidence. They suggest that females could benefit from mating with younger males if they have more nutritious spermatophores, better sperm quality or higher sperm numbers. It is not actually known if any of these differences exist between younger and older males, but E. ephippiger does produce a large spermatophore, which could provide important direct benefits. There could be indirect benefits too: young males have been shown to have higher-quality offspring in Drosophila melanogaster (Price and Hansen, 1998). The preference for younger males is not general, however: in other species, females have been found to prefer older males (Zuk and Simmons, 1997), virgin males (Snedden, 1996; Sakaluk and Ivy, 1999), or novel males (Bateman, 1998), although in most cases song does not appear to be the cue on which choice is based. There is also evidence for female choice based on courtship song. In some cricket species, females will only mate with males that produce courtship song,

SOUND SIGNALLING IN ORTHOPTERA

223

for example in Teleogryllus oceanicus (Burk, 1983) and Acheta domesticus (Nelson and Nolen, 1997). In T. oceanicus, males fight, and fighting success is correlated with mating success because dominant (winning) males are more likely to court females and to produce courtship song. Females do not discriminate among males who produce courtship song. In A. domesticus, Nelson and Nolen suggested that the ability to produce courtship song may depend on the male's condition. Hack (1998) has subsequently shown that courtship song in this species requires more than twice as much energy per unit time as does advertisement song, and may therefore be a more reliable indicator of mate quality. In other species, choice is based on some parameter of the courtship song; Gryllus lineaticeps females, for example, prefer males with higher chirp rates in their courtship song (Wagner and Reiser, 2000). Wagner and Reiser showed that courtship chirp rates, unlike those in advertisement song, are not affected by the quality of the male's diet. Female choice on the basis of courtship song is also independent of choice on the basis of advertisement song. There is no correlation, either, between a male's advertisement chirp rate and his courtship chirp rate. The two songs may therefore provide information to females about different aspects of male quality. Females of some orthopteran species, or under some circumstances, appear to show little or no choice based on song (e.g. Leptophyes punctatissima: Kilduff, 2000). This can happen if the operational sex ratio (OSR: the ratio of sexually active males to sexually receptive females) is less than one. Males of some species produce a very large spermatophore (up to 40% of body weight in Ephippi~er for example: Ritchie et al., 1998) and, having mated, need a long time to produce another one (Gwynne, 1997), so that the interval between matings is greater for males than it is for females. The rate at which a male can produce spermatophores is dependent on the quality of his diet (e.g. Simmons, 1993, Andrade and Mason, 2000; Gwynne. 2001) and in some species, the OSR can become reversed in populations with poor nutrition: under such circumstances, females compete for matings with males, and their selectivity is low (Ritchie et al., 1998). Even with an OSR of 1 or more, other factors may influence how choosy a female is likely to be. Poecilimon ornams is a duetting bushcricket in which the male produces a spermatophore equivalent to about 12% of his body weight and the OSR is likely to be close to unity (Heller et al., 1997). Heller et al. have shown that female P. ornams are indiscriminate and will respond to any type of acoustical signal as long as it is at least 1 3ms in duration, has a similar frequency spectrum to normal male song, and has a sound pressure level of 70dB or more. In this species, females do not perform phonotaxis and so their risk of predation is low. Their response to a male call is very brief and energetically very cheap. Heller et al. suggest that, in species like P. ornatus, the costs for the female of responding to a male's call are so tow that it pays her to respond to all males. She could still show choice after the male has approached

224

D.J. ROBINSON AND M. J. HALL

her, based on proximal cues such as olfaction or visual and tactile assessment of body size. Alternatively, she could benefit from mating with as many males as possible because of the nutrition provided by the male's spermatophore. In this case, cryptic female choice may still operate if, for example, she can control which sperm fertilize her eggs (Eberhard, 1996).

7 Exploitation of heterospecific sounds The acoustic advertisement signals that orthopterans produce travel quickly over long distances, are easy to locate and can be detected at any time of day. These features make them ideal for communicating with potential mates and conspecific male rivals, but they are also likely to make the signaller easy to detect by its enemies. Non-calling females approaching a male also produce noise and vibration that make them more susceptible to predation. Orthopterans are vulnerable to a wide range of natural enemies, including parasitoid flies, lizards, geckos, herons and other birds, bats, domestic cats, and primates such as tarsiers and howler monkeys (Bailey, 1991; Zuk and Kolluru, 1998; Kok and Louw, 2000; Gwynne, 2001). Some of these hunt visually or locate victims by the vibrations that they make when moving. However, many are thought to orient acoustically to the sexual signals of their prey, although this has been demonstrated in relatively few species. For example, the lizard Psammodromus algirus is a predator of the bushcricket Steuropleurus staff and is attracted to its song. It will approach motionless singing individuals, but not motionless silent individuals (Bateman, 2001). Most diurnal birds hunt visually, but some adopt different strategies to hunt orthopterans at night. For example, herons have been observed walking on land at night, tracking calling crickets to their burrows; experiments have confirmed that the birds use the call to locate their prey (Bell, 1979). Parasitoid tachinid flies of the tribe Ormiini have a worldwide distribution. They parasitize ensiferan Orthoptera: field crickets (Gryllidae), mole crickets (Gryllotalpidae) and bushcrickets (Tettigoniidae) are known to be hosts (see Allen, 1995a). The fly deposits its larvae on the host. They burrow into it and feed on it from within. The larvae emerge after seven days, usually killing the host in the process, and then pupate outside the host (Adamo et al., 1995). In the seven days during which the larvae are developing, the host's ability to reproduce is reduced: for example, parasitized male Poecilimon mariannae produce fewer and smaller spermatophores and have reduced reproductive success as a result (Lehmann and Lehmann, 2000), The flies use the advertisement call of the male to locate their host. This was first reported by Cade (1975) for Ormia ochracea (then called Euphasiopteryx ochracea). He carried out experiments with dead crickets attached to a speaker, using Texas Go, llus integer (now called G. texensis). Flies were more likely to attack a cricket attached to a speaker broadcasting the crickets' song than one broadcasting

SOUND SIGNALLING IN ORTHOPTERA

225

a control sound. Lakes-Harlan and Heller (1992) have also shown that Therohia leonidei only parasitizes males and is attracted to them by their song surgically muted males were never attacked. Foliage-gleaning bats in Panama feed heavily on bushcrickets and use the male call to locate their prey. In a cage experiment, bushcricket species calling frequently were located more quickly by the bats than species calling rarely (Belwood and Morris, 1987). Calling crickets and bushcrickets, which produce high-frequency songs, are especially vulnerable to gleaning bats because the frequency of their calls often coincides with the bats' best hearing frequency (Bailey and Haythornthwaite, 1998). However, predators that locate their prey acoustically do not necessarily rely on the sexual signals of their prey. Heller and Arlettaz (1994) have shown that the European scops owl, Otus scops, feeds intensively on bushcrickets but does not take more males than females, even though the males call. They suggest that this may be because the owl has limited sensitivity for sounds above 10kHz, so it is likely that the only song it can hear is that of Tett~onia viridissima, which has a peak spectrum at 10 kHz. The songs of the smaller bushcrickets are much higher, e.g. Platvcleis albopunctata at 29 kHz, PhoIMoptera griseoaptera at 20 kHz, and Barbitistes serricauda at 26 kHz. For T. viridissima, they found no difference in the proportion of males and females in the diet; for smaller bushcrickets such as P. albopunctata, the owls actually took more females than males. They conclude that owls are able to locate T. virMissima males by their call, but otherwise rely mainly on the noise their bushcricket prey make when moving. Females, being more mobile than males because they approach the male to mate and move around to lay eggs, are more vulnerable as a result. Similarly, laboratory experiments have shown that the pallid bat Antrozous p. pallidus, although it uses echolocation for general orientation, locates its prey using the sounds they generate while moving (Fuzessery et al., 1993). The bats never captured stationary calling crickets (Acheta domesticus) but did attack silent walking crickets. The cricket song has a spectral peak between 4.4 and 4.9 kHz, and harmonics as high as 19 kHz, and is audible to the bat. Fuzessery et al. suggest that, since crickets call from protected locations where they are relatively inaccessible, the bats may learn that a walking cricket is much more vulnerable. It is not just predators that exploit the sounds of their prey. Airborne sounds and vibrations transmitted through the air or through the substrate can betray the presence of a predator, and orthopterans, like other insects, use their acoustic senses for predator detection as well as reproduction. This is evident from the structure of orthopteran acoustic systems (see review by Kalmring and KClhne. 1983). There are no neuronal filters at the level of receptor organs or in the ventral cord that are absolutely tuned to conspecific stridulatory signals, and central auditory neurons are much more broad banded with respect to frequency and time parameters than would be necessary just for intraspecific communication. For example, in Locusta mi~ratoria, the auditory

226

D.J. ROBINSON AND M. J. HALL

ventral cord neurons react to sound frequencies up to 100 kHz, even though the song of this species only contains frequencies up to 40kHz, and this allows locusts to recognize many other sounds in the environment as well as conspecific song. The habituation properties of some auditory central cord neurons make them most responsive to new or suddenly changed signals and so suggest that these neurons have a "warning' function. The larvae of grasshoppers have a well-developed auditory system, even though they have no need to find a mate. Some orthopteran species actually have two sets of auditory units: one is tuned to frequencies appropriate for listening to the calls of conspecifics, while the other is used for detecting predators (R6mer and Bailey, 1998). The most sensitive vibration receptors found in insects are the subgenual organs of orthopterans (see review by Kalmring and Kfihne, 1983): they react to vibration stimuli over a wide range, from less than 30 Hz to more than 5000 Hz, with maximal sensitivity between 200 and 1000 Hz. In addition, companiform sensilla react to low-frequency vibration up to about 100 Hz. Vibratory central neurons, in gryllids at least, show habituation and respond primarily to new or changed stimuli, so may also have a warning function. 7.1

DEFENCES A G A I N S T A C O U S T I C A L L Y O R I E N T I N G P R E D A T O R S IN G E N E R A L

Endler (1991) described the stages of a successful predation event: encounter, detection, identification, approach, subjugation and consumption. A number of authors (including Bailey, 199l; Endler, 1991; Endler, 1992: Zuk and Kolluru, 1998; Haynes and Yeargan, 1999) have reviewed the various defences that prey can adopt to counter the predation threat. In the context of sound signals, encounters can be avoided by "one-upmanship' a greater detection distance of the predator by the prey than exists the other way round - or by adopting different activity times from the predator. Detection can be avoided by calling intermittently or not at all, or by confusion (making the detection of a single individual by the predator more difficult or making it difficult for the predator to fix on a single individual long enough to identify it as edible, e.g. by synchronized calling between males). It can also be avoided if the prey stays outside the sensory limits of the predator. This can be achieved by using a "private channel', i.e. a frequency outside the range of the predator's hearing or at least frequencies that it is less sensitive to, or by switching from calling to substrate tremulation, or by using signals that attenuate rapidly with distance so that predators are less likely to overhear them, e.g. by using low-intensity courtship song instead of the advertisement call when the mate is nearby. Identification can be avoided by confusion, by aposematism (conspicuousness associated with unpalatability or ability to inflict harm) or mimicry (resemblance to unpalatable species). For aposematic prey, aggregation helps to enhance the signal or its reinforcement in the predator's brain. Approach (attack) can be avoided by startle responses such as stopping singing when a predator is detected, or aggregation (to saturate the predator). Subjugation can

SOUND SIGNALLING IN ORTHOPTERA

227

be avoided by producing disturbance sounds, i.e. sudden noises that make the predator hesitate and allow the prey to escape.

7.1.1 Aggregating Males of many orthopteran species aggregate spatially and sing in choruses (Greenfield and Shaw, 1983; see section 8.2). Aggregations can be passive, in which individuals cluster together as a result of attraction to resources such as oviposition sites, food, singing perches, etc, or active, in which males are attracted to each other. Males show positive phonotaxis to conspecific male song in many species. Males of some species aggregate temporally and sing in sprees (Walker. 1983a). There may be several benefits to spatial aggregation and synchronized singing (Greenfield and Shaw, 1983; Greenfield, 1994; Allen, 1995b. See also section 8.2). Central individuals in a cluster are less vulnerable to predators (the selfish herd effect: Hamilton, 1971). Spatial aggregations can saturate the appetites of predators and may facilitate predator detection. There may be a dilution effect, whereby the risk of attack to any one individual is reduced the greater the number of conspecifics there are close by. Synchronized calling, which occurs in several orthopteran species, e.g. Pholidoptera griseoaptera (Jones, 1966), may make it more difficult for the predator to locate a single singer within the chorus because sound is emitted from several locations at once. Few of these possibilities have been tested in orthopterans. It is known that bats preying on frogs are attracted more frequently to asynchronous than to synchronous calls (Tuttle and Ryan, 1982). However, Cade (1981) found no evidence for a dilution effect in a Texas population of G13'llus integer (now called G. texensis): males calling in choruses attracted as many parasitoid flies per individual as did isolated males. Walker (1983a) has suggested that sprees may be the result of predators exerting selection pressure on females performing phonotaxis, causing them to cluster their movements in time so as to gain some of the same benefits as they would by aggregating spatially. As a result, males should condense their singing into the same time period in which females perform phonotaxis. 7.1.2

Frequency modulation

Two South American species of cricket, Eneoptera guyanensis and Lerneca .[uscipen,is are unusual among orthopterans in that they show significant changes in frequency within their song elements at constant ambient temperature (Desutter-Grandcolas, 1998). Desutter-Grandcolas suggests that these frequency modulations could make it more difficult for acoustically orienting predators and parasitoids to locate the singing male. The changes in frequency would confuse any predator that identifies prey from the frequency spectrum of its song. This possibility is supported by the unusual singing behaviour of L.

228

D.J. ROBINSON AND M. J. HALL

juscipennis, which may be another adaptation to make it more difficult for a predator to locate them: unlike most other crickets, which stay motionless while singing, the male jumps and turns around completely while singing, and also jumps in the leaf litter between two bursts of song.

7.1.3

Deji, nsive sounds"

Disturbance (alarm or defence) sounds produced 'by arthropods held in the hand, or disturbed in various manners such as pinching, probing or restraining are known in almost every order of insect' (Alexander, 1967). Disturbance sounds (reviewed in Masters, 1979; Bailey, 1991) are very similar between species. They are usually noise-like, and are often emitted as an unstructured hiss; if they do consist of a series of pulses, these are usually presented at an irregular or erratic rate (as, for example, in the Mexican bushcricket species Pterophylla beltrani and P. camell(/blia described by Shaw and Galliart, 1987). Disturbance sounds have a very wide frequency spectrum, including ultrasound, which means that as many predators as possible will be able to hear them. They may aid insects by startling the attacking predator, by acting as an aposematic signal or by mimicking an aposematic signal. There is some experimental evidence that disturbance sounds do deter predators. For example, Sandow and Bailey (1978) showed that active secondary defence (production of disturbance sound, with raised forelegs and open mandibles), had a deterrent effect on potential predators. The disturbance sound alone may have a deterrent effect if the predator learns to associate the sound with the defensive display. Belwood (1990) has described disturbance sounds in bushcricket species in Central and South America. Some, such as Cnemidophyllum eximium, flail their legs rapidly in an up and down motion when handled, which produces a loud raspy sound. Male Steirodon careovirgulatum and C. eximium and others also produce a noisy broad-band disturbance sound in the form of a loud, startling stridulation. In Xestoptera cornea, Scopiorinus Ji'agilis and Parascopioricus exarmatus, females also produce disturbance sounds. Most species of New Zealand weta show defensive stridulation (Field, 1993; Field and Glasgow, 2001). The most common type (observed in all species of Hemideina and Deinacrida) is produced once the insect has assumed a defensive posture. In response to visual, tactile or auditory stimulation, the legs are brought down rapidly in an arc, and a harsh scratching sound of broadband frequency is produced as they rub against the abdomen. All 16 New Zealand species of Hemiandrus stridulate by rocking forward with all legs on the ground and the hindlegs pressed against the abdomen, so that the abdomen rubs against the hindlegs to produce the sound. The large-tusked weta, Motuweta isoh~ta, produces a sharp, loud, barking sound by rapidly opening its mandibles while its tusks are pressed against each other; at the same time it

SOUND SIGNALLING IN ORTHOPTERA

229

stridulates in the same way as the Hemiandrus species. The Deinacrida rugosa group of wetas contract the abdomen and extend it telescopically as their legs kick in an arc towards the source of the disturbance, producing a hissing sound. In two undescribed species of Australian Gryllacrididae belonging to the genera Ametrus and Haeh'ogo,llacris, defensive stridulation is produced as part of an elaborate visual display, by femoro-tergal stridulation (Field and Bailey, 1997). Two rows of spines on abdominal tergites II and Ill of both species are rubbed by an elongate area of tubercules on the inner femoral surface of the hind legs; this motion involves a complex counter-rotation of the leg between leg and abdomen. Both sexes of some American Jerusalem crickets show abdomino-femoral defence stridulation in which the abdomen is moved past the hind femora or the hind femora past the stationary abdomen (Weissman, 2001).

7.2

DEFIANCES AGAINST PARASITOIDS

Ormiine flies hunt at night so cannot orient to their hosts visually. For example, when song tape-recorded from a Texas Gryllus integer (now called G. texensis) population was broadcast, it attracted O. ochracea from sunset to dawn but not during daylight hours (Cade et al., 1996). Flies were attracted in the greatest numbers in the few hours following sunset, and there was least attraction in the hours just before sunrise. In order to locate their hosts, female tachinid flies must perform the sanqe task as a female cricket responding to the call of a potential mate. The flies have tympanate ears which, as a result of convergent evolution, resemble a cricket ear much more than they do a typical fly's ear (Lakes-Harlan and Heller, 1992; Robert et al., 1992). Some respond to ultrasound (LakesHarlan and Heller, 1992). The ear of the female Ormia ochracea is sharply attuned to 4-5 kHz (Robert et al., 1992), which is the song frequency of her main host, the field cricket Gi3,11us rubens (Walker, 1993). Hearing thresholds of the female fly are very low: 20dB SPL at 5 kHz, which is consistent with long-range detection of their hosts (Robert et al., 1992). In the absence of visual or olfactory cues in laboratory experiments the females could locate a sound source from 4m, and land at a mean distance of 8.2cm away from it (Mfiller and Robert, 2001). Even if the stimulus is interrupted, the fly can continue on approximately the correct trajectory and land close to the target. Their performance declines the earlier the stimulus signal is interrupted, but this ability to persist in orientation after the signal has stopped suggests that a strategy of ceasing singing, such as the ASR described for Neoconocephahts ens~,,er, or producing short calls rather than long ones, may not be very effective. There is no physiological evidence yet for an ASR (acoustic startle response) in Orthoptera that is evoked by the sound produced by the fly, which has a fl-equency of 200 to 400 Hz.

230

D.J. ROBINSON AND M. J. HALL

Although female O. ochracea are most sensitive to sound in the range 4 6 kHz, they are also sensitive to ultrasound in the 20 60 kHz range; males also have tympanal ears but they are different both in form and function from the female's ear, being relatively insensitive to the 5-kHz frequencies that characterize cricket calls but just as sensitive as females in the ultrasound range (Robert et al., 1992; Hoy and Robert, 1996). This may be because, being active at night, the parasitoid flies are themselves vulnerable to echolocating bats, and the ability to hear bat echolocation calls may have evolved in response to this predation pressure. Walker (1993) tested the response of O. ochraeea to G. rubens calls with different pulse rates (where a pulse consists of a syllable within a chirp). He found that it changes with temperature, in parallel with temperature-induced changes in the pulse rate of natural songs. He varied the frequency of the call, the pulse rate, and the duty cycle independently of each other and found maximum attraction occurring at a frequency of 4.4 kHz, a pulse rate of 45 per second, and a duty cycle of 20 80%, which compares very closely with the natural song of G. rubens at 21'C of 4.6kHz pulses, produced at a rate of 45 per second, with a duty cycle of 50%. Producing calls divided into chirps instead of a constant trill of pulses made little difference to fly attraction rates, showing that producing pulses in short bursts (i.e. chirps) makes no difference to the chance of being parasitized. Despite the tight coupling between G. rubens song and O. ochracea attraction, Ormiine flies are not necessarily host specific. O. oehraeea is attracted not only to the song of Grvllus rubens but to various other cricket species (although in much smaller numbers), including G. jultoni, G. integer, G. firmus, Orocharis luteolira and Scapteriseus borellii (Walker, 1993). Therobia leonidei, the only ormiine fly in Europe, has a broad range of tettigoniid hosts in three subfamilies: Ephippigerinae, Tettigoniinae and Phaneropterinae, including several Poecilimon species (Lakes-Harlan and Heller, 1992; Lehmann et al., 2001). The Australian species Homotrixa alleni has a number of bushcricket hosts, with very different call frequencies and structures, and its ear is not sharply tuned to the frequency of the host call (Allen, 1995a). How then does the fly recognize its hosts when their calls are so different? A study of the variability of the call structure of the austrosagine bushcricket Seiarasaga quadrata suggests that H. alleni may use short chirp length rather than frequency as an auditory cue when locating its host (Allen, 2000). Allen speculates that some temporal patterns may be easier to locate and/or the ear of H. alleni may have a sensory bias, enabling it more easily to detect shorter chirps. Certainly, males with shorter chirps were lost from the S. quadrata population as the three-month calling season progressed (see section 3.4.2). Parasitism rates vary between different parasitoids and different hosts, but can be very high up to 87% of S. quadrata males can be parasitized by H. alh'ni by the end of the season (Allen, 1995a). Because the flies are attracted to

SOUND SIGNALLING IN ORTHOPTERA

231

the male call, females tend to be less affected although they can pick up larvae deposited near males when approaching them to mate or walking close by them (Cade el a]., 1996). As a result of the difference in parasitism rates between the sexes, the sex ratio tends to become increasingly female biased as the breeding season progresses (Allen, 1995a). In those orthopteran species susceptible to parasitism, a number of aspects of calling behaviour seem to have been affected by the heavy selection pressure imposed by the flies. Rotenberry et al. (1996) looked at the calling characteristics of field crickets, Teleogryllus oceanicus, throughout the Pacific, and found that they differ between populations parasitized and unparasitized by O. ochracea. Their analysis of calls from five geographically separate populations with differing degrees of parasitism strongly suggests that selection pressures by the fly have shaped song characteristics rather than being due to geographical variation. Some species avoid calling in those periods of the night when the flies are most active, although this has to be balanced with pressure from other predatory species, such as diurnal birds, which may make calling during daylight or other times during the night equally dangerous. For example, in a Texas population of GIo~llus integer (now called G. texensis), males call from sunset to dawn but more males call in the hours leading up to sunrise and at sunrise, when the flies are less active, than earlier on; females also mate more frequently when more males are calling, i.e. when flies are less active, and this may reduce the probability that they too will become infected (Cade et al., 1996). Sciarasaga quaeh'ata is unusual in that it calls in daylight, beginning 2 3 hours before sunset and continuing until at least an hour alter midnight (Allen, 1995b). A population of Teleogryllus oceanicus introduced to Hawaii, where it is parasitized by Ormia ochracea, shows a more abrupt onset and cessation of calling at sunset and sunrise than do unparasitized populations elsewhere (Zuk et al., 1993). In Hawaii, O. ochracea is mainly active at dawn and dusk and Zuk et al. suggest that males have been selected not to call during the active periods of the fly. Some species reduce the time they spend calling. For example, a population of Teleogryllus oceanicus in Hawaii, which is parasitized by O. ochracea, called less than an unparasitized population in Moorea (Kolluru, 1999). Seasonal variation in amount of calling related to parasitoid fly abundance has been observed in G. rubens (Burk, 1982). Individual male Grvllus texensis from Texas and Oklahoma (the author referred to this gryllid as a Texas/ Oklahoma population of G, i, teger but it has recently been recognized as a separate species: see Cade and Otte, 2000) and G. rubens from Arkansas and Oklahoma call for about 3 hours per night, while G. integer from California and New Mexico call for about 7 hours per night (Cade, 1991). Males of G. te.vensis are parasitized by O. ochracea: it is not known if the flies parasitize the other populations, although G. rubens is likely to be a host since it is parasitized by the fly elsewhere. The duration of nightly calling is heritable (Cade, 1984b)

232

D.J. ROBINSON AND M. J. HALL

and Cade suggests that the reduced calling time may be the result of the selection pressure from the fly. The shapes of the distributions of mean individual calling duration against proportion of males are skewed for G. texensis and G. rubens, implying that directional selection of some sort has been operating on these populations, but not for G. inte,~er. Lehmann and Heller (1998) studied Poecilimon veluchianus and P. mariannae in Greece. The two species are very closely related, parapatrically distributed and similar in activity patterns, body size and population density. The former produces monosyllabic songs (I syllable with a duration of about 80ms, with chirps repeated at 0.4 Hz and an interchirp interval of 2.5 s) and the latter polysyllabic songs (5-11 repeated syllables per chirp with total duration of about 600 ms, with chirps produced at a mean rate of 0.4 Hz with an interchirp interval of 1.9 s). SPL and frequency of song are the same in the two species. The polysyllabic species was parasitized about three times as much as the monosyllabic species, and Lehmann and Heller suggest this is because its higher duty cycle (24% compared with 3%) makes it easier to locate. The ability of a male to reduce his calling time may be limited, however, by the mating system of the species. Heller and yon Helversen (1993) analysed the calling activity of males for several species of Poecilimon, representing two different communication systems. In species with mute females that approach the males, such as P. veluchianus, P. mariannae and P. propinquus, calling was restricted to darkness, and syllable numbers per hour were high. This is probably because the male has no feedback as to whether and when a female is approaching and so must signal his position continuously. In species where females respond acoustically to male song and thus can induce the male to approach them, such as P. ~['finis, P. artedentatus and P. nobilis, males called during both day and night or during the day only, and syllable numbers per hour were low. In this situation, once a female has responded to him, the male can increase his calling rate and approach the female as rapidly as possible. Therobia leonidei parasitizes Poecilimon, but only those species that sing at night and produce high syllable numbers. Similarly, in Sciarasaga quadrata there seems to be no relationship between individual male calling duration and the likelihood of being parasitized by H. alleni (Allen, 1998). Males show a consistent pattern of calling activity over their lifetime but there is no decline in longer-calling males in the population as the season progresses, showing that these males are no more likely to be attacked by H. alleni than shorter-calling males. Neither is the distribution of call durations within the population skewed, indicating that there is no history of directional selection acting on call duration. Allen suggests that this could be because the species has a long intermating interval of about four days. A longer-calling male is more likely to attract a female and, once he has mated, he stops calling and is no longer vulnerable to fly attack. So the increased risk of calling for longer may cancel out the decreased risk gained by mating more often. He also has some evidence that encounters with females

SOUND SIGNALLING IN ORTHOPTERA

233

are rare, so the advantages of calling for long periods to increase the chance of attracting a mate probably outweigh the disadvantages of attracting a parasitoid fly. O. ochracea has been shown to orient more readily to high-intensity songs, and some hosts may reduce the risk of parasitism by reducing the loudness of their call. Males from a Texas population of Grvllus integ,er (now called G. texensis) call at lower intensities at dawn, when the flies are active, that they do earlier (Cade el al., 1996) Parasitoids may have a more subtle effect on calling, however, than simply reducing calling time, or the loudness of the call. The population of Teh, ogwIlus oceanicus in Hawaii showed a reduction in chirp duration compared with populations unparasitized by O. ochracea (Zuk et al., 1993: Zuk and Simmons, 1997). The song characteristics of males that harboured parasitoid larvae also differed significantly from those of unparasitized males (Zuk el al., 1998). The song in T. oceanicus consists of a 'long chirp' of five to seven pulses, followed by a series of 'short chirps' of one to three pulses. Males with longer long chirps containing shorter interpulse intervals were more likely to be parasitized. Zuk el al. suggest that this is because O. ochracea finds these males more attractive. The intense selection pressure from parasitoids may even have resulted in a shift in the carrier frequency of the call of Sciarasaga quadrata (R6mer and Bailey, 1998). The call, at 5kHz, is unusually low for tettigoniids. The responses of the tympanic nerve and a first-order interneuron (omega neuron) in the afferent auditory pathway also showed that the hearing system is most sensitive to frequencies of 15 20 kHz, an effective mismatch to the conspecific call resulting in a reduced sensitivity of approximately 20dB at the carrier frequency of the call. S. quadrata can occlude its spiracular opening, however, which increases the sensitivity of the ear to lower frequencies, resulting in a match between the best frequency of the ear and the carrier frequency of the call. This ability is unusual among tettigoniids the spiracle is permanently open in most species (Bailey, 1993). By calling at 5 kHz, S. quaeh'ata escapes the interfering calls of several other tettigoniid species. Low frequencies may also help sound transmission. They found that the calls of S. quadrata showed ahnost no attenuation in excess of that produced by spherical spreading alone. This compares with a species calling at 20 kHz, where excess attenuation may be as much as 40dB over 30m (R6mer and Lewald, 1992). R6mer and Bailey suggest, however, that the selection pressure from H. alle, i may be the most important reason for the shift in call frequency. The ear of H. alleni is most sensitive to frequencies between 10 and 20kHz and so is partially mismatched to the call of S. quadrala. Calling at 5 k Hz, instead of at 10 20 k Hz to match the tuning of the S. quadrata ear, results in the parasitoid being less sensitive to its host's call by about 10 15dB (Fig. 18). Reducing the call frequency also reduces call intensity by about 20 dB, which makes the call even less likely to attract a fly. Matching between call frequency and sensitivity of

234

D.J. ROBINSON AND M. J. HALL

the auditory system of the conspecific receiver is then achieved by partially closing the spiracle. The closure of the spiracle is under the individual's control, which enables S. quadrata to hear at a broad range of frequencies and therefore allows it to detect, for example, approaching predators.

7.3

D E F E N C E S A G A I N S T BATS

There are about 700 species of insectivorous bats, most of which use ultrasonic echolocation calls when hunting. Some bats attack flying insects (aerialhawking), while others take stationary insects from the vegetation (foliagegleaning). Bats feed extensively on Orthoptera: 42% of the insects eaten by Micronycteris hirsuta in Panama, for example, are bushcrickets (Belwood, 1990). As well as the calls made by bats, ultrasound is generated by, for example, carnivorous tettigoniids, the calls of small terrestrial mammals, and disturbances to vegetation. Ultrasound therefore represents a serious risk of predation. Many species of Orthoptera, like other nocturnal insects, have tympanal organs that are sensitive to ultrasound, e.g. in the Tettigoniidae (Libersat and Hoy, 1991), the Acrididae (Robert, 1989) and the Gryllidae (Farris and Hoy, 2000). The first evidence that orthopterans are responsive to ultrasound

80 -10 ca

m ~ 60 r~ "O 4o

-20 ~

20 l

5

10

40

-30

Frequency (kHz) FIG. 18 Hearing sensitivity of Sciarasaga quash'am when the spiracle is open (dashed line) and partially blocked (solid line), compared with the sensitivity of the ear of the parasitoid fly Homotrixa alleni, the most common predator of S. quadrata (dotted line). The spectrum of the calling song of S. quadrata is also shown (shaded area). The ability of the fly to detect its host is much reduced for calls at a frequency of 5 kHz compared with calls at 10 20kHz (R6mer and Bailey, 1998, fig. 9). Reprinted with permission from the Company of Biologists.

SOUND SIGNALLING IN ORTHOPTERA

235

was found in mole crickets by Suga (1968). He recorded from the leg nerve and thoracic connections in Gwllotalpa hexadacO, la and Scapteriscus didactylis and found that neurons in both species respond to sounds up to 70kHz. McKay (1969, 1970) found auditory neurons in Ruspolia (Homorocowphus) bushcrickets that respond to sounds above 30kHz. Ultrasonic sensitivity in bushcrickets, to frequencies as high as 100kHz, has also been reported for several species in Europe, e.g. Tettigonia vMdissima (Rheinlaender and R6mer, 1980; Hutchings and Lewis, 1983) and T. cantans (Kalmring and Ktihne, 1980). The ability to detect ultrasound enables Orthoptera to take action to avoid being captured if a bat, or other ultrasound-generating predator, gets too close. Orthoptera show a number of ultrasound-induced startle responses, including stopping singing, stopping flying, and flying away from the source of the sound (see review by Hoy, 1992a). Several species are known to stop singing in response to bats. Sales and Pye (1974) report that stridulating Conocephalus conocephalus and C. maculatus in Nigeria stop singing in the presence of flying bats. The creosote bush cricket, lnsara covilleae, often calls from the tops of bushes where it is vulnerable to foliage-gleaning bats (Spangler, 1984), The call is above 20 kHz and within the bats' hearing range. Spangler recorded insect singing and bat echolocation calls in the field and found that calling ceases during the high-intensity call of a bat approaching closely, but not during the low-intensity call of a distant bat. Calling resumed after the bat had passed, but only intermittently. He also observed that two other tettigoniids, lnsara elegans and Plagiosfira albonotam, also cease calling when echolocating bats approach. Farris and Hoy (2000) looked at ultrasound sensitivity in Eunemobius carolimls, a member of the Nemobiinae (ground crickets). This species shows an ASR while flying, in which the wings are closed and flight is stopped until the ultrasound stimulus ends. Extracellular recordings from the cervical connectives revealed auditory units that are sensitive to frequencies greater than 15 kHz with best sensitivity at 35 kHz (79 dB SPL threshold). Stimuli in this frequency range, which is similar to that emitted by echolocating bats, also elicit the ASR. A recent study (Faure and Hoy, 2000d) has shown that the tettigoniid Neoconoc~j~halus ensiger exhibits both a non-flight and an in-flight ASR. Of 20 N. ensiger tested, 80% responded consistently to pulsed sounds by stopping singing or pausing, The threshold for the response was low (60 70 dB) above 30kHz, being more or less constant up to 100kHz. Below 30kHz thresholds rose sharply and in the audible range below 20 kHz the thresholds were above 100kHz and most animals did not respond at all. This type of ASR in which the animal stops singing appears to be a highly stereotyped one, occurring with a short latency of 20 50ms in this species. The response occurs only when an ultrasonic stimulus arrives during the intersyllabic period of silence, so presumably the insect's own sound masks external signals. In flight, N. ens~er shows a response to ultrasound that is typical of that shown by other

236

D.J. ROBINSON AND M. J. HALL

flying insects to bat echolocation pulses. Within 50 ms of detecting a pulse of ultrasound N. ens~er closes all four wings (Libersat and Hoy, 1991). However, the response is not directional. A comparison of the tuning curves for the ASR cessation of singing and the ASR cessation of flying shows that they are very similar and, in shape, resemble the tuning curve of the T-cell, although the threshold for the T-cell, as might be expected, is lower (mean AdB c.35 dB) (Faure and Hoy, 2000d). A link between the T-cell and in-flight ASR was suggested by Libersat and Hoy (1991) and the physiology of the T-cell makes it well suited to a role as a bat-detector (Faure and Hoy, 2000b) (see also section 5.2.1). Unlike N. ensi,~er, Tetti,~onia viridissima does show a directional response to simulated bat calls. It responds in one of three ways, depending upon the intensity of ultrasound detected (Schulze and Schul, 2001). When simulated bat calls reached an intensity of 55 60 dB, an insect in tethered flight showed a steering response away from the sound source. At increasing intensities the insects stopped beating their hindwings but remained in the normal flight posture. Wing beats resumed 0.3 to 1 s after the bat calls stopped. At higher intensities the forewings folded into the resting position and the insect would have dived if untethered. The median threshold for the steering behaviour was 53.5dB SPL, for the wing beat interruption 64dB SPL and for diving 76dB SPL. These sound intensities have been used to estimate the distance of a calling bat (call taken as 110dB SPL 25 cm in front of the bat) from the insect at the point at which the behaviour is initiated (Schulze and Schul, 2001). Steering behaviour would occur when the bat was 18 m away. Interruption of the wing beat would occur at 10m separation and the diving response at 5 m. The estimated distance at which the bat would first detect the echo from the bushcricket is 5 m. T. viridissima and N. ensiger are closely related and the difference in in-flight responses to bat calls suggests that the evolution of batavoiding strategies is shaped by different selection pressures and is not a trait that is highly conserved within the tettigoniids. Locusts (Robert, 1989) and crickets (Lewis, 1992) also show a directional response to sound sources containing ultrasound. For example, laboratory experiments on Teleogo, llus oceanicus in flight have shown that it will turn towards the song of a conspecific (a series of 5 kHz pulses) and away from ultrasound (sound pulses containing 40kHz) (Moiseff et al., 1978; Nolan and Hoy, 1986a). In this species, electrical stimulation of the high-frequency ascending neuron lnt 1 (ANA), sufficient to produce a response rate above 180 to 220 spikes per second, results in negative phonotaxis, suggesting that Int 1 is both necessary and sufficient to elicit this behaviour (Nolan and Hoy, 1986b), |nt 1 also responds to all three of the sexual songs produced by f . oceanicus (advertisement, aggressive and courtship), and so is involved both in avoidance and courtship (Harrison et al., 1988). Nolan and Hoy (1986b) suggest that the switch between the two behaviours may be based on a difference in spike rate in the neuron. Predator avoidance only occurs if activity in hat 1 is more than

SOUND SIGNALLING IN ORTHOPTERA

237

180 spikes per second, while activity in response to courtship song is in the region of 35 spikes per second. Int 1 could therefore provide input to separate neural networks in the brain concerned with negative phonotaxis and song recognition. For the ASR of ceasing to fly (and thereby dropping out of the echolocation zone of the bat) to work, the insect needs to detect the bat before the bat can hear the echo. Schul et al. (2000) measured the hearing range of the long-winged bushcricket, Phaneroptera fidcata, for echolocation calls of the mouse-eared bat Myotis nO,otis in the field. Male P. falcata often call from exposed perches on the vegetation and, while singing, frequently change position by flying, so that they are exposed both to aerial-hawking and foliage-gleaning bats. The echolocation calls of M. myotis are similar in intensity, duration and frequency range to other aerial-hawking bats large enough to prey on tettigoniids, having a fundamental frequency of about 80 to 27kHz, with strongest amplitudes around 30kHz and durations of 5 10ms. Schul et al. recorded the hearing responses of the bushcrickets to the echolocation calls neurophysiologically and simultaneously recorded the echolocation calls of M. myotis via two microphone arrays, which allowed them to reconstruct the flight path of the bat and determine the maximum distance at which the bushcricket detected the bat. They found that the bushcrickets have hearing ranges of 13 30m. They calculate that the bushcricket has more than I s to recognize the predator and respond evasively before the bat hears the returning echo. The insects' sensitive hearing therefore gives them an advantage over the bat - the 'one-upmanship' described by Endler ( 1991 ). Mason et al. (1998) used sound traps to show that the phonotactic behaviour of both sexes of the mole cricket Scapteriscus borellii is actually affected in the field by ultrasound that represents the possibility of bat predation. Each of two sound traps broadcast synthetic calling song with a carrier frequency of 2.7 kHz, but one of the traps broadcast at a higher sound intensity (increased by 6dB). The two traps were 10m apart and on the basis of the difference in power output from the two traps 17% of individuals would be expected to be attracted to the quieter trap (Forrest and Raspet, 1994). Then 40 kHz sound at a number of different intensities was added to the calling song broadcast by the louder trap and the proportion of insects choosing the quieter trap was measured. Forty kilohertz is a typical peak frequency in the sonar pulses emitted by species of insectivorous bat in North America (Fenton and Bell, 1981). There was a significant positive relationship between the intensity of the ultrasound added to the broadcast from the louder trap and the number of individuals collected at the quieter one. With no ultrasound present, the proportion collected at the quieter trap was as predicted from earlier work (Forrest and Raspet, 1994), but as the intensity of ultrasound increased, the attractiveness of that trap declined and the proportion collected at the quieter trap increased.

238

D.J. ROBINSON AND M. J. HALL

Foliage-gleaning bats often use echolocation for general orientation, but tend to locate their prey passively by listening for the sounds that they make. Four of the six bushcricket species taken by Micronycteris hirsuta on Barro Colorado |sland in Panama call in the 23 27 kHz range, suggesting that the nature of the prey taken depends on whether the bat can hear their call or not (Belwood, 1990). The selection pressure from bats has resulted in a number of adaptations in calling behaviour. Some species avoid predation by calling when bats are least active (Belwood, 1990). For example, wasp mimicking bushcrickets of the genera Aganacriv and Scaphura in Latin America have two active periods, one during the day and one during the night. Their mimicry protects them from predators during the day but not during the night. At night, they sing in a very narrow time band between 03:00 and 05:00, which coincides with a lull in the feeding behaviour of gleaning bats. The strategy appears to be effective because gleaning bats do not appear to take these species. Similarly, Copiphora brevirostris is active between 23:00 and 05:00. It is the most common bushcricket on Barro Colorado Island and Belwood has shown it to be extremely palatable to bats in laboratory feeding experiments, but it is taken by gleaning bats in only small numbers. Some species simply call less, making it more difficult for bats to locate individuals. For example, species of neotropical bushcrickets sympatric with foliage-gleaning bats call for a much smaller proportion of the time compared with species in areas where these bats are absent (Belwood and Morris, 1987). In most understorey forest bushcrickets, where bat predation is high, calls are short (usually less than 1 s) and produced infrequently (every few seconds to every few minutes). One exception to the rule is Ischnomela pulchripennis, which sings for about 50% of the time. However, this species calls from protected sites such as large bromeliads covered in long sharp spines that could kill or injure a gleaning bat; species with short, intermittent calls call from exposed sites (Belwood, 1990). Some or all calling may be replaced with complex species-specific tremulations that generate vibrations inaudible to bats but which reach females through the substrate; females may respond by tremulations of their own which lead males to cease audible calling (Belwood and Morris, 1987; Belwood, 1990; Morris et al., 1994). Males of most pseudophylline and copiphorine species supplement calling with tremulation. Calling and tremulation are produced independently and in a seemingly random fashion within the same block of time. Males tremulate in the absence of females, so the vibrations seem to be true long-distance communication signals. For example, Myopophyllum speciosum is a pseudophylline neotropical bushcricket which calls for a very small proportion of its time (Morris et al., 1994). Male calls are only 148ms long on average and are produced a considerable time apart, with intercall intervals averaging 8.7s. Pairing is completed with tremulation signals generated at closer range. Morris et al. have shown that other neotropical tettigoniids in rain-forest understorey, including species of

SOUND SIGNALLING IN ORTHOPTERA

239

Choeroparnops, Schedocentrus, Docidocercus, and Copiphora, also employ elaborate vibratory signals and some have reduced or even eliminated their use of airborne sound. Belwood (1990) looked at tremulation in five bushcricket species on Barro Colorado Island. Four of these species (Docidocercus gigliotosi, Acanthodis curvidens, Balboa tibialis, Copiphora hrevirostris and Aganacris insectivora) spent more than twice as much time in tremulating as they did in calling, in fact C. brevirostris tremulated more than 90 times as much as they called. Morris et al. (1994) suggest that predation pressure from bats is the most likely reason for the shift from calling to tremu[ation, this being just one of a suite of several anti-bat adaptations shown by neotropical bushcrickets, including very long antennae which may enable them to anticipate a bat strike; the ability to cling tightly to the vegetation when a capture attempt is made; and spines which may deter bats flom eating them. There are costs associated with tremulation, however. The body movements required arc likely to be energetically expensive. Nor do vibrational signals free orthopterans from predators. Bushcrickets observed in the laboratory stop singing or tremulating if anything touches their cage (Belwood, 1990). Belwood speculates that this may be adaptive, especially for tremulation, because individuals are more likely to be detected by predators with vibration receptors when they tremulate. She frequently observed ctenid spiders feeding on bushcrickets in the field, but especially on Copiphora brevirostris, which has the longest and most complicated tremulation signal of all the bushcrickets that she studied on Barro Colorado Island. Some species have calls with unusually high carrier tYequencies. For example, the carrier frequencies of the neotropical species Myol~ophyllum speciosum, Drepanoxiphus angustelaminatus, Haenschie[la ecuadorica and an undescribed species of Haenschiella range from 65 to105 kHz (Morris et al., 1994). Attenuation and scattering of sound are tYequency dependent (Michelsen and Larsen, 1978). Water vapour (Griffin, 1971: Pye, 1979) and vegetation (R6mer and Lewald. 1992) can also act together to reduce the energy of ultrasonic sounds. Thus, the higher the carrier fi*equency, the shorter the range, and the less likely the calls arc to be overheard by bats. M. .V~ecioszm£s call at 81kHz will attenuate from a typical level of 100dB at 10cm to 50dB within 3m, which is just at the species" hearing threshold (Mason et al., 1991). Morris et al. (1994) assess the evidence for several possible reasons why these species might use such high frequencies, such as reducing interference by other species, and conclude that the most likely explanation is bat predation. Some species alter the components of their call. feleogr)'lhts oceanicus has a two-component call: a brief chirp [\)llowed by a trill. Males prefer to call from refuges in natural hollows or vegetation, and unprotected males produce calls containing fewer trill elements than males in refuges (Bailey and Haythornthwaite. 1998). Bailey and Haythornthwaite suggest that this is the

240

D.J. ROBINSON AND M. J. HALL

result of selection pressure from bats. They carried out laboratory experiments to test the vulnerability of T. oceanicus to predation from two species of gleaning bats, Nyctophilus major and N. geql)/?oyi, both of which use passive listening rather than echolocation to locate their prey (although neither is a natural predator). Males calling from refuges were not predated. Unprotected insects were more likely to be attacked if they had longer calls (with more elements); for calls of the same length, they were more likely to be attacked if their calls contained trills rather than chirps. As Rentz (1975) was the first to point out, calling songs ofneotropical understorey forest bushcrickets differ significantly from those of temperate species. In most Latin American understory forest bushcrickets, calls are shorter and produced more infrequently, with lower song duty cycles, and emphasize higher frequencies, compared with temperate species (Belwood, 1990). These differences may be related to differences in predation pressure from bats. For example, in the palaeotropics, there are fewer bat species that use the calls of their prey to locate them than in the neotropics (Heller and Volleth, 1995), so predator pressure is probably less intense. If calling behaviour is affected by the selection pressure imposed by bats, it might be expected, therefore, that the effects would be less extreme in the palaeotropics than in the neotropics. A study by Heller (1995) provides support for this. He looked at a number of Malaysian leaf-mimicking bushcricket species from the family Pseudophyllidae ( Tympanophyllum arct~f'olium, Promeca sumatrana, Promeca perakana, Chomh'oderella borneenses, and Phyllomimus inversus). All three of the ricotropical anti-bat adaptations described above are less well developed in these species. The number of call syllables produced per day ranged from 9000 to 42 000 (giving call duty cycles of 1.4-28.2%), with one exceptionally high value of 360 000 in Chondroderella borneensis. This compares with 200 to 2000 syllables per day (giving call duty cycles of less than 1%) in neotropical species of Pseudophyllidae. None of the Malaysian species appeared to use tremulation. The peak frequency of the call for all species was below 12 kHz, and in fact the lowest frequency long-distance signal yet discovered in insects is the 600 Hz call produced by T. arcu/blium. The Malaysian species are closely related to the African Pseudophyllidae and Heller suggests that the differences between the Malaysian species and the neotropical species probably represent general differences between the neotropical and palaeotropical Pseudophyllidae. He compared carrier frequency in I1 species of palaeotropical and 37 species of neotropical Pseudophyllidae and found a significant difference in carrier frequency between the two groups (Fig. 19), although it is possible that this is a side-effect of the different camouflage strategies adopted by the two groups rather than the result of differential predator pressure. Call frequency is negatively correlated with the length of the stridulatory file. Palaeotropical Pseudophyllidae have longer stridulatory files than neotropical species but this may be linked to the different body and wing shapes that they have adopted in relation to their camouflage strategies.

SOUND SIGNALLING IN ORTHOPTERA

241

120 -O-

Neotropieal species

100,

1

-,O- Palaeotropical species ]

I

8O

,Y

60 40 o --~

20

..... -°W.... d . ~

oe 0

•. •

•

•

•

" , - ' " ' i l l ' ; .... i,

0 ~

u~

• ............

0

.-.•

0

0 2

4

5

6

7

8

9

10

11

12

Pronotunl length (iYtm)

FIG. 19 Pronotum length and carrier frequency in 11 species of palaeotropical and 37 species of neotropical Pseudophyllidae (Heller, 1995, fig. 4). Reprinted with permission from Cambridge University Press.

7.4

SEXUAL VERSUS NATURAL SELECTION

The signals that orthopterans use to attract mates make them vulnerable to acoustically orienting predators and parasitoids. Sexual selection favouring more conspicuous calls that increase attractiveness to females will therefore be countered by natural selection favouring less conspicuous calls that decrease attractiveness to enemies. Often the features that make a call attractive to females and to enemies are the same. In Go'llus 1#Teaticeps, for example, both female crickets and the parasitoid fly Ormia ochracea prefer male calling songs with higher chirp rates, longer chirp durations and higher chirp amplitudes (Wagner, 1996). Similarly, those male Poecilimon thessalicus preferred by females also have a higher risk of being parasitized by Therobia leonidei (Lehmann et al., 2001). Female P. thessalicus prefer males with the loudest and longest calls, perhaps because they are easier to detect and locate, and T. leonidei may prefer these males for the same reason. A male preferred by a female will, if nothing else, be more likely to have a female nearby, thereby providing two potential hosts instead of one (Wagner, 1996). In a Texas population of Gryllus integer (now called G. texensis), Gray and Cade (1999) have shown that both female crickets and O. ochracea prefer male calling song with average numbers of pulses per trill. Preference for the average song would minimize search costs for both conspecific females and parasitoid flies. Pulses per trill are not related either to male size or age, and have a substantial genetic component. Thus female crickets exert stabilizing sexual selection on male song, which reduces genetic variation and maintains

242

D.J. ROBINSON AND M. J. HALL

population cohesiveness, but this is balanced by parasitoid flies exerting disruptive natural selection, which promotes genetic variation and population divergence. The two opposing forces may together maintain the observed high levels of genetic variation in male song. An alternative explanation for the persistence of genetic variation in male song characteristics in the face of natural selection from acoustically orienting predators has been put forward by Hedrick (2000). She has shown that G. integer males differ in their durations of uninterrupted trilling (calling-bout lengths), that these differences are heritable, and that females prefer males with longer calling bouts. However, males with longer, more conspicuous songs compensate behaviourally for their increased vulnerability to predators by acting more cautiously: they take longer to emerge from a safe shelter within a novel, potentially dangerous environment, and stop calling for a longer time when their calls are interrupted by a predator cue. There is some evidence that these differences are heritable. Genetic variation in calling bout duration could therefore be maintained because males with shorter bout lengths, although less attractive to females, may gain opportunities to mate, while males with longer bout lengths are silent following disturbance by a potential predator.

7.5

EVOLUTION OF PREDATOR AVOll)ANCE MECHANISMS

How did predator avoidance mechanisms dependent on the ability to hear ultrasound evolve? It has been argued (Gwynne, 1995) that hearing and intraspecific sound communication have both evolved independently in the Grylloidea and in the Tettigonioidea and, if this is the case, then the ability to detect ultrasound that is found in both groups must have evolved independently. Some insects preyed upon by bats, such as moths, seem to have evolved ultrasound-sensitive ears as a direct result of predation (Hoy, 1992a; Hoy and Robert, 1996). In the Orthoptera, however, there is some dispute as to whether ultrasound sensitivity evolved for this reason or instead in the context of intraspecific communication. For example, Hoy and Robert (1996) and Hoy (1992a) argue that the same hearing organs serve to detect both conspecific signals and echolocation signals of bats; they speculate therefore that the ability to hear ultrasound was added on to the ability to hear lower frequency conspecific calls. Their argument is based on evidence that the Orthoptera have existed for at least 200 million years, while the bats only appeared in the fossil record 50 million years ago, and that crickets and bushcrickets were able to stridulate (and therefore, presumably, send signals that conspecifics could hear) at least 150 000 years ago. Their ears thus evolved in the context of intraspecific communication. They argue further that, by the time the bats appeared on the scene, orthopterans had been established for millions of years, and it is likely that they used ultrasound components in intraspecific signalling systems and

SOUND SIGNALLING IN ORTHOPTERA

243

had hearing organs tuned to ultrasounds, just like their modern counterparts. Hoy and Robert therefore think that in crickets, bushcrickets and locusts, the ability to hear bats was probably a simple step from the ability to hear conspecifics. Others, such as Lewis (1992) and R6mer (1992), disagree, based on the neuroethology of present-day insects. For example, the auditory neuropile is organized differently in crickets, bushcrickets and acridids. Most crickets and bushcrickets sing in early morning or evening and most are heavily preyed upon by bats, Crickets show negative phonotaxis to ultrasound. R6mer et al. (1988) therefore suggest that the bushcricket system evolved to mediate rapid escape behaviour and was only later incorporated into and modified for conspecific acoustic communication. The acridids on the other hand are more highly preyed on by lizards and birds and are less sensitive to ultrasound birds and lizards can be detected by lower frequency sound, vibration and visual cues. Acridids therefore need to combine signals from more than one modality. There is also some evidence that the tympanal organs of orthopterans were present before their sound-generating mechanisms evolved, i.e. before they started to stridulate to communicate with conspecifics (Smart, 1963). In addition, Mason (1991, 1999) has shown that song production can be uncoupled from song reception. In the haglid C3,phoderris motlstrosa, the best frequency of the auditory system (12kHz) is mismatched to the spectral peak of calls (2kHz) (see section 4.1.4). Mason argues that the primary function for hearing in this species is predator detection (although not detection of a bat because 12kHz is below the typical sonar range of most bats).

8

Cooperation and competition between males

In the majority of orthopteran species, the male calls and the female does not reply. Yet the male generally has ears that are fully functional and tuned to the conspecific call (see section 4). Males benefit not only from being able to listen for predators (see section 7) but from eavesdropping on the calls of conspecific males in the competition to acquire mates. Features of the advertisement call relevant to females, such as those that provide information about male size or vigour, should be relevant to male rivals too, since they are also likely to provide them with information about the calling male's fighting ability or his ability to attract females. Males of many species sing in choruses, and various hypotheses have been put forward as to why males should do this. Many of these explain chorusing in adaptive terms: individual males benefit in some way by singing together and so cooperate to create the chorus. Others see chorusing as the result of male competition or simply as an epiphenomenon.

244 8. ]

D.J. ROBINSON AND M. J. HALL SPACING, AGGREGATING AND FIGHTING

The spatial distribution of individuals is under the control of a variety of selective forces. Aggregations can be passive, in that individuals are attracted to resources such as oviposition sites, food, or singing perches. The distribution of males of the desert grasshopper, Ligurotettix coquilletli, for example, is related to the distribution of the food plant most attractive to females, the creosote plant (Greenfield and Shelly, 1985). Aggregations can also be active, in that individuals are attracted to each other. In many acoustically communicating species, sound plays an important role in distribution. For example, in the bushcricket Tettigonia viridissima, males clump in areas where the vegetation provides higher singing perches than the surroundings, a behaviour that would enhance the distance over which the song could be broadcast (Arak et al., 1990). Advertisement calling can act as an agonistic signal directed towards rivals, as well as a signal to attract females (e.g. Simmons, 1988). In 7". viridissima, for example, males are spaced out as far as possible from their rivals (Arak et al., 1990). In the haglid Cvphoderris monslrosa the song of the male is a very loud trill, almost a pure tone, at 12kHz (Mason, 1996). At 10cm from a singing insect the intensity is 105dB SPL. The spacing of males in the field is such that the songs of nearest neighbours are attenuated to about 60 dB SPL (unpublished data quoted in Mason et al., 1999). The role of song in repulsing competitors was demonstrated experimentally by Bailey and Thiele (1983). They released several male Mygalopsis marki in the same place and found that intact males spread out from the release point, while experimentally deafened males did not. They also showed that dispersal rate and the eventual distance between males is dependent on sound intensity: a population of larger males, which sing more loudly, was more widely spaced than a population of smaller males. R6mer and Bailey (1986) showed, however, that males do not space themselves out on the basis of threshold sensitivity to the calls of other males. They measured the maximum hearing distance of M. marki in the field, using the response of an identified auditory interneuron as a 'biological microphone', and found that it was more than twice the mean distance between males. Males of some orthopteran species show no evidence of territoriality and may move considerable distances between one calling session and the next, for example the haglid Qvphoderris strepitans (Dodson et al., 1983) and some tettigoniids such as Leptophyes punctatissima (M. J. Hall, personal observation). Some, however, do defend territories, and a male that calls within the territory of a neighbour is likely to become involved in a fight with the territory owner (Alexander, 1961; Greenfield and Minckley, 1993; Mason, 1996; Gwynne, 2001). In some cases the distance between territorial males is dependent on information contained in the call (e.g. Schatral et al., 1985; R6mer and Bailey, 1986). Nearest-neighbour distances vary greatly, for example 2m in

SOUND SIGNALLING IN ORTHOPTERA

245

Tetli~gonia cantans (Dadour, 1990) compared with 11.5 m in M. marki (Dadour and Bailey, 1990). Greenfield (1990) has suggested that the necessity of hearing conspecific males calling has been an important factor in the evolution of the advertisement song. In the genus Neo('onocephalus, for example, some species have discontinuous songs, with gaps between groups of syllables, while others have continuous songs. The former tend to live in high densities, while the latter live at lower densities. Greenfield argues that discontinuous songs have evolved largely because of the importance to a singing male of being able to hear his neighbours (which he can only do when not singing himself: Hedwig, 2001), especially in higher density populations where acoustic, spacing or territorial interactions are likely to be more frequent. Similarly, in the grasshopper Omocestus viridulus, there is a correlation between male density and the length of the song (Eiriksson, 1992). Males sing shorter songs at higher density, although they sing more often so that total singing time remains the same. In this case, however, Eiriksson argues that males sing shorter songs when malemale competition is high in order to increase the probability that they will hear the female's reply, since females stridulate in response to the mate but do not necessarily wait for the male song to end before replying. Male song may attract other males as well as repulse them: in some species, playback of recorded advertisement song attracts males as well as females (e.g. Walker and Forrest, 1989). As a result of the simultaneous attraction and repulsion that calling males have for each other, they may end up evenly spaced within an aggregation (e.g. Campbell, 1990). There are various possible reasons why aggregating may be adaptive for a male. For example it might make him less vulnerable to predators (see section 7.1.1), increase his mating success if females are more likely to approach groups than single males (Alexander, 1975L or allow an unattractive male to exploit more attractive ones by intercepting females approaching preferred males (satellite behaviour: see section 8.3). Female Conocephalus nigropleurum are more attracted to groups of males (Morris et al., 1978: see also section 6.6) but studies of several other species found little or no evidence that this occurs (e.g. Otte and Loftus-Hills, 1979; Cade, 1981; Walker, 1983b). Cade (1981), for example, found that solitary male field crickets attracted just as many females as clumped males did, on a per capita basis. The degree of separation between males may influence their relative attractiveness to females. In experiments with the cricket Eunemobius carolinus, the attractiveness of a male calling song for a female increased as the intensity relative to other songs increased (Farris et al., 1997). A lower intensity song lost attractiveness as the separation from a more intense song decreased. The ability of crickets to distinguish between calling songs on the basis of intensity decreased as the intensity of the calling songs increased. These findings are consistent with the conclusion from modelling (Forrest and Raspet, 1994) that dense spacing is more costly tk)r less powerful singers.

246

D.J. ROBINSON AND M. J. HALL

In such circumstances, the optimal strategy for weaker singers should be to search silently for females or adopt the role of a satellite male (see section 8.3). The proportion of non-calling males does increase at higher densities in some species (e.g. Hissman, 1990). The switch from calling to satellite behaviour in E. carolinus would only be appropriate; however, if there was a shortage of singing sites at suitable distances from other males (Forrest and Raspet, 1994). Males of many species of orthopteran, including most crickets, produce a distinct aggressive call during male-male aggressive encounters (Greenfield and Minckley, 1993; Zuk and Simmons, 1997; Field, 2001a). Stridulation by one male cricket during a fight usually causes the other to stridulate, although dominant males call more than subordinate males; the winning male also calls after the loser retreats, while the loser rarely does so (Alexander, 1961). In tarbush grasshoppers, Ligurotettix planum, 80% of territorial disputes are settled without a fight by the exchange of aggressive 'shuck' calls (containing up to 60 shucks, each 25 35 ms long); the winner in most of these cases is the male with the highest rate of producing shucks (Greenfield and Minckley, 1993). Greenfield et al. argue that the shuck rate reflects an individual's strength or motivation and therefore the likelihood that he will win should a contest escalate to a fight. In other species, such as certain bushcrickets, aspects of the advertisement song seem to function in male-male competition and may allow males to assess each other's size and strength before engaging in actual physical fighting. For example, male Kawanaphila nartee increase their calling rate when interacting with other males (Simmons and Bailey, 1993). Many Orchelimum and Conocephalus species have 'ticking duels' in which the tick part of the 'tickbuzz' song is extended when males interact (reviewed by Gwynne, 2001). In the common meadow bushcricket, O. vu]gare, an intruding male is usually silent as he approaches his calling neighbour, only occasionally producing the tick part of his advertisement song. But as the two males get closer, tick sequences become more and more common from both males and eventually they may fight. In the haglid Cyphoderris monstrosa, the advertisement song actually seems to play a greater role in male-male competition than it does in female attraction (Mason, 1996). Males may use each other's duty cycle to assess fighting ability: winning males are more sustained singers. Mating success can be heavily dependent on male success in aggressive encounters. In the gregarious cricket Amphiacusta maya, for example, the same type of chirp is used both in courtship and in male-male aggression; males that win fights have the greatest mating success, and experimentally silenced males are less able to achieve high dominance rank and are less successful at mating, not because females are less receptive but because intruding males interrupt silent courtships (Boake, 1983). In A. maya, courtship song appears to be mainly directed at other males: silent males are just as able to elicit mounting as males producing courtship song.

SOUND SIGNALLING IN ORTHOPTERA

8.2

247

CHORUSES

Chorusing is found in many insects that communicate acoustically. Greenfield and Shaw (1983) defined chorusing as +the singing of two or more conspecific males for relatively long periods of time during which there are statistically significant temporal interactions involving the acoustics signals of neighbors'. Males in orthopteran choruses are often aggregated spatially, for example near resources important to females, but they can also form temporal calling assemblies, i.e. sprees (Walker, 1983a). In many species, males respond acoustically (phonorespond) to the sounds of neighbouring males. In others, males probably do not phonorespond, but may end up singing simultaneously because they respond in the same way to environmental conditions such as the onset of dusk (e.g. mole crickets and short-tailed crickets: Walker, 1983a). Greenfield and Shaw (1983) review and define various types of chorusing. In unison singing, males sing during the same extended time period without showing any consistent temporal relationships between any song units. Unison singing is shown by many species with songs consisting of either relatively short or relatively long non-repeating units (NRUs), or of repeating units (RUs) at irregular intervals. Repeating units are very similar sound components of relatively uniform length at a given temperature. They can be aperiodic, in which the sound is produced at irregular intervals, or periodic, in which the sound is produced regularly so that the sound plus the following interval is a consistent length. There may be a hierarchy of sound components song bout, chirp, syllable, etc. which may vary in repeatedness within the same species. In unison singing, individual males may phonorespond by initiating song in response to conspecifics and by singing for longer and more continuously than when singing in acoustic isolation. Examples include Gryllus species (Alexander, 1962), Stenobothrlts lineatus (Haskell, 1957) and Neoconocephalus exiliscanorus (Fulton, 1934). Unison bout singing is unison singing in which collective staging is divided into bouts of irregular length separated by periods of silence of irregular length. Examples include several Neoconoce7~halus species (Greenfield, 1990), Syrbula admirabilis and S./'uscovittata (Otte and LoftusHills, 1979), and Aerochoreutes carlinianus (Otte, 1977). In synchronous chorusing, the phase angles between the RUs of two or more neighbouring males are consistently very small. For example, in the snowy tree cricket, Oecanthus jidtoni, males chirp at the rate of about 2.5 per second at 25~C, but solitary males sing with less uniform chirp periods and lengths than chorusing males (Walker, 1969). Other species that show synchronous chorusing include Platvch, is intermedia (Samways, 1976) and Neocotmcephalus caudellianus (Greenfield, 1990). Leaders and followers (males that habitually initiate their RU first and second respectively) can be distinguished in some species, such as N. nehrascensis (Meixner and Shaw, 1986).

248

D, J. ROBINSON AND M. J. HALL

In alternating chorusing, the phase angles between the RUs of two or more neighhouring males are consistently maintained at about 180 ° and the males do not overlap RUs (e.g. Gryllotalpa ma/or: Hill, 2000). In some alternating species, chorusing males call more rhythmically than isolated males, i.e. intervals between RUs are more variable in isolated males. Some species call more slowly when alternating compared with solo singing (e.g. Acanthogryllusjortipes: Cade and Otte, 1982), while others call more quickly (e.g. Pholidoptera griseoaptera: Jones, 1966). Leaders and followers (where leaders are defined as males that typically initiate calling after a period of silence) can again be distinguished in some species, such as Pterophylla camell(folia (Shaw, 1968). There are also choruses that cannot be described either as synchronous or as alternating but which show statistically significant relationships among RUs. For example, adjacent Amblycor)pha parvipennis males within an aggregation sing with overlapping calls such that each male overlaps his neighbour's phrase about 70% of the time. Where phrases overlap, phonatomes (the sound produced by a single wing-stroke) are synchronized (Shaw et al., 1990). Some species vary in the type of chorusing that they show: phase angles in Orchelimum nigripes, O. vulgare and O. gladiator, for example, are variable and at times males synchronize, while at others they alternate (Feaver, 1977). Chorusing behaviour has rarely been explored in duetting species, but the existence of a time window has interesting implications for chorusing. The response latency of female Phaneroptera nana is about 60ms (Tauber, 2001). Males alternate with their neighbours and their calls rarely overlap. The period of the male calls is about 700ms (Tauber and Cohen, 1997). In experiments with synthetic calls, females were found to respond to both males of an alternating pair, providing that the calls were at least 200ms apart. The timing of the male calls is forced into a pattern that does not jam any female reply and preserves the time window. So in this species the male has two distinct responses to auditory stimulation by the conspecific song (Tauber, 2001). If it is a male song then the response is a delay in calling; if a female song, then phonotaxis is initiated. In this species the male and female songs differ in both temporal pattern and frequency spectrum, allowing the possibility of discriminating between male and female song, unlike L. punclatissima (Robinson et al., 1986) where male and female have similar frequency spectra. It will be interesting to see if male L. punctatiss#na show this discrimination. Is participating in a chorus adaptive for males? Some possible benefits of chorusing were discussed in section 8.1. Greenfield and Shaw (1983) give several other possibilities, and Greenfield (1994) reviews the various hypotheses and studies on the evolution of chorusing, looking in particular at the possible cooperative and competitive functions of signal interactions.

SOUND SIGNALLING IN ORTHOPTERA

249

Greenfield (1994) gives several reasons why chorusing might be adaptive in terms of male-male competitive interactions. Since males cannot hear while they are calling themselves, alternation may help them to assess each other on the basis of their calls, and so play a role in spacing and aggression. Synchronous calling could also result from a kind of acoustic satellite behaviour (see section 8.3) in which a subordinate male produces short calls that are completely overlapped by his dominant neighbour, by which means he avoids detection and possible eviction. Neoconocephalus nebrascensis choruses may provide an example of such behaviour (Meixner and Shaw, 1986). Similarly, in duetting species, a synchronizing male may be able to ~take over" a female responding to his neighbour (e.g. Elephantodeta nobilis: Bailey and Field, 2000). Sprees have been explained in terms of males calling when environmental conditions are most favourable (section 6.2), females most numerous (e.g. Walker, 1983a), or enemies least active (see section 7), which causes them to call together by default. However, males of species that form sprees have been observed to start singing, or increase their rate of singing, in response to playback of conspecific song (e.g. Shaw et al., 1990), so sprees may be the result of competitive or cooperative interactions between males. Males may need to match or outdo other males in order to be attractive to females; the high energetic costs then limit males to singing in periods that coincide with female activity. If females are available over a long period, males may not have enough energy to sing for the whole time. They may break up their calling into bouts to cover the availability period as effectively as possible, leading to unison bout chorusing. Greenfield and Shaw (1983) and Greenfield (1994) also list several possible reasons why males might benefit from cooperating to form choruses, including the reduced vulnerability to predators and increased attractiveness to females of groups compared with single males discussed in section 8.1. In species where the call is produced at regular intervals, the period may be the most important bit of information for species recognition: if neighbouring males synchronize their calls, then any female in the area will hear the species-specific rhythm. This explanation has been suggested for the synchronous chorusing in Oecanthus ji¢ltoni (Walker, 1969). Where information is contained in the song itself, then alternation would preserve the information. If females select males based on relative as opposed to absolute criteria, then they may choose to mate only with members of choruses so that they can assess the attributes of potential mates. Males might also synchronize in order to maximize peak signal amplitude (the 'beacon" effect). If females are more strongly influenced by peak signal amplitude than amplitude averaged over time, then tightly clustered groups of males that synchronize would be more attractive than groups that do not. In duetting species, synchronous calling would avoid masking of the female reply by other males. If female conspecifics suffer from the same confusion effects produced by many males calling simultaneously as has been suggested for predators, then males would

250

D.J. ROBINSON AND M. J. HALL

benefit by alternating their signals. In Ligurotettix planum, it has been shown that females show reduced phonotaxis to overlapping calls (Minckley and Greenfield, 1995) and this may contribute to the maintenance of call alternation in this species (Minckley et al., 1995). Males may alternate to preserve signal components, such as pulse rate, which are critical features in female phonotaxis. This is analogous to the preservation of the species-specific rhythm by synchronous chorusing, but assumes that males are limited in their ability to adjust pulse timing and therefore unable to synchronize pulses: call alternation is therefore selected instead of call synchrony. Finally, if females are preferentially attracted to groups with higher duty cycles, males would benefit by alternating their calls, thus maximizing the number of calls in a particular time period. There is little or no evidence to support any of the functions that have been suggested for alternating and synchronous choruses, however. Greenfield and his colleagues (Greenfield and Roizen, 1993; Greenfield, 1994; Greenfield et al., 1997) argue instead that alternating and synchronous choruses are epiphenomena arising out of the tendency for females to prefer the leading of two male calls (see section 6.6). They suggest that, because of the selection pressure on males imposed by this female preference, individual males compete to jam each other's signals, using mechanisms that modify the timing of their call relative to those of others so that they increase their chance of producing a leading call. For example, female Neoconocephalus spi=a prefer male calls that precede others by 10-80ms. A male N. spiza resets his calling rhythm when he hears a neighbouring male calling, which leads to him producing fewer following calls. But a synchronous chorus results when males, all using this resetting mechanism, call at similar rates (Greenfield and Roizen, 1993). Greenfield et al. (1997) have developed a general model of rhythmic signalling in which calling is controlled by an oscillator in the central nervous system that may be inhibited and reset by a neighbour's call. This inhibitory-resetting mechanism affects one call period only, with the natural rhythm continuing thereafter~ Playback experiments have shown that rhythms and signal interactions in several species fit the assumptions of their model, for example in Pholidoptera griseoaptera (Jones, 1966), Pterophylla camell([olia (Shaw, 1968), Oecanthus jzdtoni (Walker, 1969), and Ligurotettix planum (Minckley et al., 1995), Greenfield et al.'s simulations show that, given female choice for leading calls, the basic resetting mechanism is evolutionarily stable (as long as it includes a 'relativity adjustment') for the velocity of signal transmission and selective attention towards a subset of signalling neighbours. Greenfield et al. present data to show that relativity adjustments are incorporated into the resetting mechanisms of several species, including O. jultoni (Walker, 1969), Neoconocephalus spiza (Greenfield and Roizen, 1993), and L. planmn (Minckley et al., 1995). Selective attention toward loud, nearby neighbours has been demonstrated in L. coquilletti and L. planum by Snedden et al. (1998): see also R6mer (1993).

SOUND SIGNALLING IN ORTHOPTERA

251

Such an inhibitory-resetting mechanism can generate either synchronous or alternating choruses by default. If two neighbouring males both signal at similar rates that vary very little, phase delays tend to mutually align their rhythms within a single period and synchrony continues thereafter; alternation occurs when the rebound interval following resetting is relatively short compared with the signal period, and is seen in species with slower signalling rates (Greenfield, 1994). Thus, in the genus Platycleis, the only species that regularly synchronizes is P. intermedia, the species with the highest signalling rate (Samways, 1976). 8.3

SATELLITE MALES AND SILENT SEARCHING

Males that do not produce advertisement calls have been observed in various orthopteran species, for example in a Texas population of Grrllus integer (now called G. texensis) (Rowell and Cade, 1993), Teleogr),llus oceanic'us (Zuk et al., 1993), Orchelimum nigripes (Feaver, 1977), Acanthoplus speiseri (Mbata, 1992), and gomphocerine grasshoppers (Green, 1995). It is possible that some males do not sing because they do not have enough energy or are too badly damaged by parasites, but there is now good evidence, in some species at least, that males are adopting an alternative mating strategy by acting as a satellite male and/or silently searching for females. Satellite males stay quietly near a calling male and try to intercept females performing phonotaxis towards the caller. The rate at which a satellite encounters females depends on the ability of the calling male to attract them, so males using this strategy could maximize their mating success by moving close to a calling male that produces the kind of song that females prefer. There is some evidence that this may happen in, for example, Acheta domesticus (Kiflawi and Gray, 2000). Various hypotheses have been put forward to explain why males may adopt a silent searching or satellite role. Cade and his colleagues (Cade, 1975, 1979, 1984a) have proposed that satellite behaviour in G. texensis may have arisen as a genetically determined, evolutionarily stable, alternative mating strategy to avoid the risk of attack by acoustically orienting parasitoids (see section 7.2). They collected calling and non-calling males in the field and found that calling males were more likely to be parasitized. More silent males were also observed in this species than in three other gryllines not known to be parasitized (Cade and Wyatt, 1984) and seasonal variation in calling related to parasitoid fly abundance has been observed in G. rubens (Burk, 1982). A study by Zuk and her colleagues (Zuk et al., 1995, Zuk and Simmons, 1997) of the Teleogr),llus oceanicus population in Hawaii has shown, however, that, in this species at least, silent males do not necessarily represent a satellite class of males, Silent males are more likely to be parasitized than calling males and are probably silent because they have too much tissue damage to sing, i.e. their silence is a by-product of parasitism rather than an adaptation to avoid

252

D.J. ROBINSON AND M. J. HALL

parasitism. On the other hand, the association between O. ochracea and T. oceanicus is more recent than for G. texensis, so the Hawaii population may simply not have had time to adapt to the fly. However, silent males are common in non-parasitized populations of T. oceanicus outside of Hawaii (Zuk et al., 1993), so the adoption of a satellite strategy cannot be due entirely to the selection imposed by the fly. Males may be opportunistic, switching between calling and non-calling depending on which strategy maximizes mating success in the prevailing circumstances. Cade (1979) has shown that in G. texensis, for example, noncallers may start to call if a neighbouring male is silenced. Whether males call or not may depend on their competitive status: males unsuccessful in competition with other males could avoid attracting superior competitors by staying silent. How often males call may be correlated with males' competitive ability in G. bimaculams (Simmons, 1986). The strategy that a male adopts also depends on population density. At high densities, the cost of defending a territory is higher (Alexander, 1961) and there is a greater probability of encountering females simply by chance. Males may benefit in this situation from stopping singing and becoming a satellite and/or searching silently for mates. Silent searching at higher densities is found in, for example, GJ:vlhls bh~Taculatus (Simmons, 1986), and G. campestris (Hissman, 1990), and the frequency of satellite behaviour increases with male population density in G. texensis (Cade and Cade, 1992). Cade and Cade (1992) have shown that male G. texensis that call more have greater mating success at low densities, whereas at very high densities there may be no difference in mating success between males who call and males who do not+ Simulations of high-density populations with strongly female-biased operational sex ratios, suggest that mating success for satellite males may actually exceed that of calling males in these circumstances (Rowell and Cade, 1993). In Orchelimum nigripes, satellite males may be 'making the best of a bad job' (Fearer, 1977, 1983). O. nigripes is territorial. Larger, older males are more likely to win fights and can hold the most sought-after central territories where most females are encountered; as a result they have greater mating success. Younger, smaller males are less successful in obtaining mates and, at high densities when they cannot obtain a suitable territory for themselves, often adopt a satellite strategy, staying quietly in the territory of another male and accosting females that respond to the calls of the territorial male. They have very little success, however, because females will not mate with a male who does not begin his courtship ritual by singing, and any satellite male that sings is quickly ousted by the territorial male. There is one major advantage to the satellite strategy: the territorial male produces a spermatophore about 10% of his body mass and, after he has mated, he cannot produce another one for some time, during which period he leaves his territory. The satellite can then take over the territory and start to sing. Fearer demonstrated experimentally that this happens by removing a territorial male.

SOUND SIGNALLING IN ORTHOPTERA

9

253

N e w directions

Work in the different disciplines of bioacoustics has advanced sufficiently that integration of their results is providing a more coherent picture of sound signalling. For example, we now have information on directionality from behavioural and neurophysiological studies that is congruent, enabling behaviour to be explained, at least in part, at a neural level. Studies of the structure and function of ears supports observations made behaviourally, and, although there is much that remains to be explored, we can provide a description of the processes involved in sound communication from the generation of a signal, its transmission through the air, its reception, its analysis and the subsequent response of the recipient. Much of the information comes, inevitably, from laboratory experiments, but the increasing ability to undertake neurophysiological, acoustic and behavioural work in the field promises to enhance the integrative approach to the study of sound signalling. Information about the processing of auditory information in the central nervous system is now extensive and the ability to record from identified neurons has enhanced greatly our understanding of the function of the orthopteran nervous system. The absence of a large section in this review, devoted to the brain, is an indication of the greatest area of ignorance in our understanding of sound signalling. True, we know that stimulation in certain regions of the brain will induce stridulation. We also know that the brain is a centre for decision-making processes in signalling and sound reception, but we know very little indeed about how and where these decisions are made. Investigation of the brain will be perhaps the most important area of research in sound signalling in the future. A lot of information is now accumulating about the natural and sexual selection pressures that affect the evolution of acoustic signals, including environmental effects, predation, mate choice, and competition between signallers. However, there are many questions that remain at best only partially explored (Endler, 2000). For example, how much do orthopterans exploit 'beneficial' environmental effects, perhaps to avoid predation or maximize signal distance? What signal properties and parameters are best for a particular environment and, given that environment and the amount of noise associated with it, what is the best way to send information: when, say, is it better to use repetitiveness and when redundancy'? When we know more about how signal parameters affect decisions made by the recipient, and how the environment degrades those parameters, it should be possible to predict much more about the details not only of signal structure but of sensory system structure and function. This is still at a very early stage. We know a considerable amount about the effects of individual predators or parasitoids on individual orthopteran species but we also need quantitative studies that look at the effects of more than one natural enemy. There are also some unanswered questions. For example, why do neotropical

254

D.J. ROBINSON AND M. J. HALL

bushcrickets under heavy predation from bats seem to have evolved much lower calling rates than bushcrickets attacked to a similar or even greater degree by parasitoid flies? It is possible that those species studied so far that have been attacked by flies have not been able to evolve more cautious calling behaviour because they are constrained by their shorter breeding lifetime (e.g. about four weeks for Poecilimon) compared with the neotropical species attacked by bats (Gwynne, 2001), but much more work needs to be done before we can arrive at a more complete understanding of the effects of natural enemies on the evolution of calling behaviour. Acoustic signals have many parameters that can be used to convey intbrmation, but more work needs to be done to find out which parameters are actually used, for example, by females to assess male quality, or to code for such things as species, gender or age. Which parameters are inherently more honest and which can be faked'? What are the trade-offs between different functions of a signal (e.g. female attraction versus repulsion of rival males), and how does this affect what parameters are used, possibly within the same call? Our knowledge of the parameters of acoustic signals that influence mating success is increasing, but distinguishing between passive and active female choice is still a big problem. And we still know very little about how female choice actually translates into male lifetime reproductive success, how individual females differ in their preferences, or how the characteristics of her mate affects a female's reproductive success. Because females mate more than once and sperm mixing is common, assigning paternity to offspring is very difficult, but DNA fingerprinting techniques should help in future research. More work is also needed, especially in the field, to see how differences between individual males in song structure, diel patterns of calling, and calling durations translate into mating success, how the calling behaviour of individual males is affected by population density and nutritional status, and how competition between signallers affects the transfer of information to females. Finally, there is an enormous potential in the Orthoptera for divergence between populations in signalling systems. There are many ways to modify acoustic signals because there are so many different parameters that can carry information. Selection on sensory processes or preferences can also result in divergence. Thus signal divergence may play a very important role in speciation in the Orthoptera but, apart from observations that different species have different signals, this area has not yet been well studied (Endler, 2000).

Acknowledgements

D JR would like to thank Professors Yamada and lkeda of the Centre for the Studies of Higher Education at the University of Nagoya for their hospitality during his six-month visit. MJH would like to thank Roger Lowry for his invaluable help and support. We would also like to thank Patricia Ash for

SOUND SIGNALLING IN ORTHOPTERA

255

her c o m m e n t s on an earlier draft o f the m a n u s c r i p t , Denise R o w e and C a t h e r i n e Eden for assistance with l i b r a r y research, a n d Brian R i c h a r d s o n for assistance with the d i a g r a m s .

References

Adamo, A., Robert, D. and Hoy, R. R. (19953. Effects of a tachinid parasitoid, Ormia ochracea, on the behavior and reproduction of its male and female field cricket hosts (Gryllus spp). J. bisect Physiol. 41, 26%277. Alexander, R. D. (1961). Aggressiveness, territoriality and sexual behavior in field crickets. Behaviour 17, 130 223. Alexander, R. D. (1962). Evolutionary change in cricket acoustical communication. Evolution 16, 433 467. Alexander, R. D. (1967). Acoustical communication in arthropods. AmT. Rev. Entomol. 12, 495 526. Alexander, R. D. (19753. Natural selection and specialised chorusing behavior in acoustical insects. In 'Insects, Science and Society' (D. Pimentel, ed.), pp. 35 77. Academic Press, New York. Allen, G. R. (1995a). The biology of the phonotactic parasitoid, ttomotrixa sp. (Diptera, Tachinidae), and its impact on the survival of male Sciarasaga quadrata (Orthoptera, Tettigoniidae) in the field. Ecol. Entomol. 20, 103-110. Allen, G. R. (1995b). The calling behavior and spatial distribution of male bush-crickets (Sciarasaga quadrata) and their relationship to parasitism by acoustically orienting tachinid flies. Ecol. Entomo/. 20, 303-310. Allen. G. R. (1998). Diel calling activity and field survival of the bushcricket, Sciarasaga qmuh'ata (Orthoptera: Tettigoniidae): a role for sound-locating parasitic flies'? Ethology 1114, 645 660. Allen, G. R. (2000). Call structure variability and field survival among bushcrickets exposed to phonotactic parasitoids. Ethology 106. 409 423. Andersson, M. (19943. "Sexual Selection'. Princeton University Press, Princeton, New Jersey. Andrade, M. C. B. and Mason, A. C. (2000). Male condition, female choice, and extreme variation in repeated mating in a scaly cricket, Orm, hius aperta (Orthoptera: Gryllidae: Mogoplistinae). J. hTsect Behav. 13, 483~497. Arak, A., Eiriksson, T. and Rades'/iter, T. (19903. The adaptive significance of acoustic spacing in male bushcrickets Tettigonia viridissima: a perturbation expe,iment. Behav. Ecol. Sociobiol. 26, 1-7. Atkins, G. and Pollack, G. S. (1986). Age-dependent occurrence of an ascending axon on the omega neuron of the cricket Teleog~?,llus ocealEcus. J. Comp. Nem'o/. 243, 527 -534. Atkins, G. and Pollack, G. S. (1987). Correlations between structure, topographic arrangement and spectral sensitivity of sound-sensitive interneurons in crickets. J. Comp. Neurol. 266, 398 412. Bailey, W. J. (1970). The mechanics of stridulation in bush crickets (Tettigonioidea, Orthoptera). 1 The tegminal generator. J. Evp. Biol. 52, 495 505. Bailey, W. J. (1990). The ear of the bush cricket. In "Thc Tettigoniidae: Biology, Systematics and Evolution" (W. J. Bailey and D. C. F. Rentz, eds), pp. 217 247. Crawford House Press, Bathurst, Australia. Bailey, W. J. (1991). "Acoustic Behaviour of Insects: an Evolutionary Perspective'. Chapman and Hall, London.

256

D.J. ROBINSON AND M. J. HALL

Bailey, W. J. (1993). The tettigoniid (Orthoptera, Tettigoniidae) ear: multiple functions and structural diversity, lnt. J. Insect Morphol. Embrvol. 22, 185-205. Bailey, W. J. (1995). Costs of calling in tettigoniid Orthoptera: a case study of Requena verticalis (Tettigoniidae: Listroscelidinae). J. Orth. Res. 4, 65-73. Bailey, W. J. (1998). Do large bushcrickets have more sensitive ears? Natural variation in hearing thresholds within populations of the bushcricket Requena verticalis (Listroscelidinae: Tettigoniidae). Physiol. Entomol. 23, 105- 112. Bailey, W. J. and Broughton, W. B. (1970). The mechanics of stridulation in bush crickets (Tettigoniidea, Orthoptera). I1 Conditions for resonance in the tegminal generator. J. Exp. Biol. 52, 507 517. Bailey, W. J. and Field, G. (2000). Acoustic satellite behaviour in the Australian bushcricket Elephantodeta nobilis (Phaneropterinae, Tettigoniidae, Orthoptera). Anita. Behav. 59, 361 369. Bailey, W. J. and Haythornthwaite, S. (1998). Risks of calling by the field cricket Teleogryllus oceanic'us; potential predation by Australian long-eared bats. J. Zool. 244, 505 513. Bailey, W. J. and Kamien, D. (2001). Hearing dimorphism, trait variation and conflicts over space in the thorax of the bushcricket Requena verticalis (Listroscelidinae: Tettigoniidae: Orthoptera. J. Comp. Physiol. A 187, 647 652. Bailey, W. J. and Rentz, D. C. F. (1990). 'The Tettigoniidae: Biology, Systematics and Evolution'. Crawford House Press, Bathurst, Australia, Bailey, W. J. and R6mer, H. (1991). Sexual differences in auditory sensitivity: mismatch of hearing threshold and call frequency in a tettigoniid (Orthoptera, Tettigoniidae: Zaprochilinae). J. Comp. Physiol. A 169, 349-353. Bailey, W. J. and Stephen, R. O. (1978). Directionality and auditory slit function: a theory of hearing in bush crickets. Science 201, 633-634. Bailey, W. J. and Stephen, R. O. (1984). Auditory acuity in the orientation behaviour of the bushcricket Pachysagella australis Walker (Orthoptera Tettigoniidae, Saginae). Anita. Behav. 32, 816 829. Bailey, W. J. and Thiele, D. R. (1983). In 'Orthopteran Mating Systems' (D. T. Gwynne and G. K. Morris, eds), Male spacing behaviour in the Tettigoniidae: an experimental approach, pp. 163 184. Westview Press, Boulder, Colorado. Bailey, W. J., Cunningham, R. J. and Lebel, L. (1990). Song power, spectral distribution and female phonotaxis in the bushcricket Requena verticalis (Tettigoniidae, Orthoptera): active female choice or passive attraction. Anita. Behav. 40, 33--42. Bailey, W. J., Greenfield, M. D. and Shelly, T. E. (1993a). Transmission and perception of acoustic signals in the desert clicker, Ligurotettix coquilletti (Orthoptera, Acrididae). J. hTsect Behav. 6, 141-154. Bailey, W. J., Withers, P. C., Endersby, M. and Gaull, K. (1993b). The energetic costs of calling in the bush-cricket Requena verticalis (Orthoptera, Tettigoniidae, Listroscelidinae). J. Exp. Biol, 178, 21 37. Balakrishnan, R. and Pollack, G. S. (1996), Recognition of courtship song in the field cricket, Teleogl~vllus oceanicus. Anita. Behav. 51, 353 366. Ball, E. and Cowan, A. (1978). UItrastructural study of the development of the auditory tympana in the cricket Teleogo,Ilus commodus (Walker). J. Embo, ol. Exp. Morphol. 46, 75 87. Ball, E. and Hill, K. G. (1978). Functional development of the auditory system of the cricket TeleogtTllus commodus. J. Comp. Physiol. A 127, 131-138. Ball, E. and Young, D. (1974). Structure and development of the auditory system in the prothoracic leg of the cricket Go, l]us commodus (Walker). II Postembryonic development. Z. ZeE/~rsch. 147, 313-324.

SOUND SIGNALLING IN ORTHOPTERA

257

Bangert, M., Kahnring, K., Sickmann, T., Stephen, R. O., Jatho, M. and Lakes-Harlan, R. (1998). Stimulus transmission in the auditory receptor organs of the foreleg of bushcrickets (Tettigoniidae) I. The role of the tympana. Hear. Res. 115, 27 38. Basolo, A. L. (2000). Problems in studying receiver biases and their effects on signal evolution. In 'Animal Signals: Signalling and Signal Design in Animal Communication" (Y. Espmark, T. Amundsen and G. Rosenqvist, eds), pp. 177-193. Tapir Academic Press, Trondheim, Norway. Bateman, P. W. (1998). Mate preference for novel partners in the cricket GwIhls himaculatus. Ecol. Entonlol. 23, 473 475. Bateman, P. W. (2001). Changes in phonotactic behavior of a bushcricket with mating history. J. bisect Behav. 14, 333 343. Bateman, P. W. and Toms, R. B. (1998a). Mating, mate guarding and male-male relative strength assessment in an African king cricket (Orthoptera: Mimnermidae). Traits. Am, Entomol. Soc. 124, 69-75. Bateman, P. W. and Toms, R. B. (1998b). Olfactory intersexual discrimination in an African king cricket (Orthoptera: Mimnermidae). J. hlsect Behav. 11, 159 163. Bateman, P. W., Gilson, L. N. and Ferguson, J. W. H. (2001). Male size and sequential mate preference in the cricket Gwllus bimacuhltus. Anita. Behav. 61, 631 637. Bell. P. D. (1979). Acoustic attraction of herons by crickets. J. N.Y. Entomol. Soc. 87, 126 127. Belwood. J. J. (1990). Anti-predator defences and ecology of neotropical forest katydids, especially the Pseudophyllinae. In 'The Tettigoniidae: Biology, Systematics and Evolution' (W. J. Bailey and D. C. F. Rentz, eds), pp. 8 26. Crawford House Press, Bathurst, Australia. Belwood, J. J. and Morris, G. K. (1987). Bat predation and its influence on calling behavior in Neotropical katydids. Science 238, 64 67. gennet-Clark, H. C. (1970). The mechanism and efficiency of sound production in mole crickets. J. Exp. Biol. 52, 619 652. Bennet-Clark, H. C. (1989). Songs and the physics of sound production. In "Cricket Behavior and Neurobiology' (F. Huber, W. Loher and T. E. Moore, eds), pp. 227 261. Cornell University Press, Ithaca, New York. Bennet-Clark, H. C. (1995). Insect sound production: transduction mechanisms and impedance matching. In 'Biological Fluid Dynamics' (C. P. Ellington and T. J. Pedley, eds), pp. 199 218. The Company of Biologists, Cambridge. Bennet-Clark, H. C. (1998). Size and scale effects as constraints in insect sound communication. Phil. Trans. R. Soc. Lond, Set. B 353, 407 419. Bennet-Clark. H. C. (1999). Resonators in insect sound production: how insects produce loud pure-tone songs, J. Exp. Biol. 2112, 334%3357. Bentley, D. R. (1977). Control of cricket song patterns by descending interneurons. J. Comp. Physiol. A 116, 19-38. Bickmeyer, U., Kalmring, K., Halex, H. and Mucke, A. (1992). The bimodal auditory vibratory system of the thoracic ventral nerve cord in Locusla mi,graloria (Acrididae, Locustinae, Oedipodini). J. E.vp. Zool. 264, 381 394. Boake, C. R. B. (1983). Mating systems and signals in crickets. In 'Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects" (D. T. Gwynne and G. K. Morris, eds), pp. 28M-4. Westview Press, Boulder, Colorado. Bonduriansky, R. (2001). The evolution of male mate choice in insects: a synthesis of ideas and evidence. Biol. Rev. 76, 305 339. Boyan, G. S. (1979). Directional responses to sound in the central nervous system of the cricket Teh,ogryllus commodus (Orthoptera: Gryllidae). 1. Ascending interneurons. J. Comp. Physiol. A 130, 137 150.

258

D.J. ROBINSON AND M. J. HALL

Boyan, G. S. (1991). A neuronal network defined by lineage in the locust CNS. Naturwi~senscha/ien 78, 565-567. Boyan, G. S. (1992). Common synaptic drive to segmentally homologous interneurons in the locust. J. Comp. Neurol. 321, 544~554. Boyan, G. S. (1993). Another look at insect audition: the tympanic receptors as an evolutionary specialization of the chordotonal system. J. hTsect Physiol. 39, 187 200. Boyan, G. S. (1998). Development of the insect nervous system. In ~Comparative Hearing: Insects" (R. R. Hoy, A. N. Popper and R. R. Fay, eds), pp. 97 138. Springer-Verlag, New York. Boyan, G. S. (1999). Presynaptic contributions to response shape in an auditory neuron of the grasshopper. J. Comp. Physiol. A 184, 279 294. Brekkow, J. and Sippel, M. (1985). Mechanisms of the transduction of sound in the tympanal organ of adults and larvae of locusts. J. Comp. Physiol. A 157, 619 629. Broughton, W. B. (1963). Method in bio-acoustic terminology. In 'Acoustic Behaviour of Animals' (R.-G. Bushel, ed.), pp. 3 24. Elsevier, Amsterdam. Broughton, W. B. (1976). Proposal for a new term 'echeme' to replace "chirp' in animal acoustics. Physiol. Entomol. I, 103-106. Brown, W. D. (1997). Courtship feeding in tree crickets increases insemination and female reproductive life span. Anhn. Behav. 54, 1369 1382. Brown, W. D. (1999). Mate choice in tree crickets and their kin. Am1. Rev. Entomol. 44, 371 -396. Brown, W. D., Wideman, J., Andrade, M. C. B., Mason, A. C. and Gwynne, D. T. (1996). Female choice for an indicator of male size in the song of the black-horned tree cricket Oecanthus n~ricornis (Orthoptera: Gryllidae: Oecanthinae). Evohaion 50, 2400 2411. Burk, T. (1982). Evolutionary significance of predation on sexually signalling males. Ha Emomol. 65, 90 104. Burk, T. (1983). Male aggression and female choice in a field cricket (Teleogryllus oceanicus): the importance of courtship song. In 'Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects" (D. T. Gwynne and G. K. Morris, eds), pp. 97 119. Westview Press, Boulder, Colorado. Burk, T. (1988). Acoustic signals, arms races and the costs of honest signalling. Fla Iz)ztomol. 71,400~409. Cade, W. (1975). Acoustically orientating parasitoids: fly phonotaxis to cricket song. Science 190, 1312 1313. Cade, W. H. (1979). The evolution of alternative male reproductive strategies in field crickets. In "Sexual Selection and Reproductive Competition in Insects' (M. S. Blum and N. A. Blum, eds), pp, 343 380. Academic Press, London. Cade, W. H. (1981). Field cricket spacing, and the phonotaxis of crickets and parasitoid flies to clumped and isolated cricket songs. Z. Tierpsychol. 55, 365 375. Cade, W. H. (1984a). Effects of fly parasitoids on nightly calling duration in lield crickets. Can. J. Zool. 62, 226-228. Cade, W. H. (1984b). Genetic variation underlying sexual behavior and reproduction. Am. Zool. 24, 355 366. Cade, W. H. (1991). Inter- and intraspecific variation in nightly calling duration in field crickets. Grv/lus i,t~Lger and G. rubens (Orthoptera, Gryllidae). J. bisect Bekav. 4, 185 194. Cade. W. H. and Cade, E. S. (1992). Male mating success, calling and searching behaviour at high and low densities in the field cricket, G~3'llus bragger. Anita. Behav. 43, 49~ 56. Cade, W. H. and Otte, D. (1982). Alternation calling and spacing patterns in the field cricket Acanthogrvllus./ortipes (Orthoptera, Gryllidae). Can. J. Zoo/. 60, 2916-2920.

SOUND SIGNALLING IN ORTHOPTERA

259

Cade, W. H. and Otte, D. (2000). GtTIlus texensis n. sp.: a widely studied field cricket (Orthoptera; Gryllidae) from the southern United States. Trans. Am. Entomol. Soc. 126, 117 123. Cade, W. H. and Wyatt, D. R. (1984). Factors affecting calling behaviour in field crickets, TeleogJ~vllus and Gryllus (age, weight, density, and parasites). Behavioz, 88, 61 75. Cade, W. H., Ciceran, M. and Murray, A.-M. (1996). Temporal patterns of parasitoid fly (Ormia ochracea) attraction to field cricket song (Go'llus integer). Cam J. Zoo/. 74, 393-395. Campbell, D. J. (1990). Resolution of spatial complexity in a field sample of singing crickets TeleogJ3,11us commodus (Walker) (Gryllidae): a nearest-neighbour analysis. Anita. Behav. 39, 1051 1057. Casaday, G. B. and Hoy, R. R. (1977). Auditory interneurons in the cricket Teleogrvllus oceanicus: physiological and anatomical properties. J. Comp. Physiol. 121, 1 13. Ciceran, M., Murray, A. M. and Rowell, G. (1994). Natural variation in the temporal patterning of calling song structure in the field cricket Grvllus pemlsyh'anicus: effects of temperature, age, mass, time of day, and nearest neighbor. Can. J. Zool. 72, 38 -42. Crnokrak. P. and Roff, D. (1995). Fitness differences associated with calling behavior in the two wing morphs of male sand crickets, Gryllus .ffrmus. A,im. Behav. 50, 1475 148l. Crnokrak, P. and Roff, D. A. (2000). The trade-off to macroptery in the cricket Grvllu,~ ./irmus: a path analysis in males. ,I. Evol. Biol. 13. 396-408, Dadour, I. R. (1990). Dispersion, dispersal, and mating behavior in Tettigonia cantans (Orthoptera: Tettigoniidae). J. Insect Behav. 3. 805 812. Dadour, I. R. and Bailey, W. J. (1990). The acoustic behaviour of male Mygalopsis marki (Copiphorinae). In 'The Tettigoniidae: Biology, Systematics and Evolution" (W. J. Bailey and D. C. F. Rentz. eds), pp. 98-111. Crawford House Press. Bathurst, Australia. Dambach, M. and Gras, A. (1995). Bioacoustics of a miniature cricket, C):cloptiloides canarieHsis (Orthoptera, Gryllidae, Mogoplistinae). J. F~\-p. Biol. 198, 721 728. Desutter-Grandcolas, L. (1998). Broad-frequency modulation in cricket (Orthoptera, Grylloidea) calling songs: two convergent cases and a functional hypothesis. Can. J. Zoo/. 76, 2148 2163. Dobler, S., Heller, K.-G. and von Helversen, O. (1994a). Song pattern recognition and an auditory time window in the female bush-cricket Ancistrura n~rovittata (Orthoptera, Phaneropteridae). J. Comp. Physiol. A 175, 67 74. Dobler, S., Stumpncr, A. and Heller, K.-G. (1994b). Sex-specific spectral tuning l\)r the partner's song in the duetting bush-cricket Ancistrura niy;rovitlata (Orthoptera. Phaneropteridae). J. Comp. Physiol. A 175, 303 310. Dodson, G. N., Morris, G. K. and Gwynne. D. T. (1983). Mating behavior of the primitive orthopteran genus Cypho&,rris (Haglidae). In "Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects" (D. T. Gwynne and G. K. Morris, eds), pp. 305-318. Westview Press, Boulder, Colorado. Doherty, J. A. (1985). Temperature coupling and 'trade-oft" phenomena in the acoustic communication system of the cricket, Grvllus himaculalHx (Orthoptera, Gryllidae). J. Exp. Biol. 114, 17 35. Doherty, J. A. and Callos, J. D. (1991). Acoustic communication in the trilling field cricket, Gryllus rubens (Orthoptera, Gryllidae). J. hlsec't Behav. 4, 67 82. Doherty, J. A. and Howard. D. J. (1996). Lack of preference |k~r conspecific calling songs in female crickets. Anita. Behav. 51, 981 990.

260

D.J. ROBINSON AND M. J. HALL

Doherty, J, A. and Hoy, R. (1985). Communication in insects. III. The auditory behavior of crickets: some views of genetic coupling, song recognition, and predator detection. Q. Rev. Biol. 60, 457472. Doherty, J. A. and Storz, M. M. (1992). Calling song and selective phonotaxis in the field crickets, GryllusJirmus and G. pennsvh,anicus (Orthoptera, Gryllidae)..l. Insect Behav. 5, 555-569. Eberhard, W. G. (1996), 'Female Control: Sexual Selection by Cryptic Female Choice'. Princeton University Press, Princeton, New Jersey, Eberl, D. F. (1999). Feeling the vibes: chordotonal mechanisms in insect hearing. Curr. Opin. Neurobiol. 9, 389 393. Eiriksson, T. (1992). Density dependent song duration in the grasshopper Omocestus viridulus. Behaviour 122, 121-132. Elliott, C. J. H. and Koch, U. T. (1985). The clockwork cricket, Naturwissenschq/?en 72, 150-153,

Elsner, N. (1974). Neuroethology of sound production in gomphocerine grasshoppers I. Song pattern and stridulatory movements. Z. Vergl. Physiol. 88, 67-102. Eisner, N. (1994). The search for the neural centres of cricket and grasshopper song. In 'Neural Basis of Behavioural Adapations" (K. Schildberger and N. Elsner, eds), pp. 167-194. Gustav Fischer, Jena, Germany. Elsner, N. and Wasser, G. (1994). Leg and wing stridulation in various populations of the gomphocerine grasshopper Stenobothrus rubicundus (Germar 1817). I. Sound patterns and singing movements. Zoology 98, 179 190. Endler, J. A. (1991). Interactions between predators and prey. In "Behavioural Ecology: an Evolutionary Approach' (J. R. Krebs and N. B. Davies, eds), pp. 169 201. Blackwell Scientific Publications, Oxford. Endler, J. A. (1992). Signals, signal conditions and the direction of evolution. Am. Nat. 139, S125 S153. Endler, J. A, (2000). Evolutionary implications of the interaction between animal signals and the environment. In 'Animal Signals: Signalling and Signal Design in Animal Communication' (Y. Espmark, T. Amundsen and G. Rosenqvist, eds), pp. 11 46. Tapir Academic Press, Trondheim, Norway. Ewing, A. W. (1989). Arthropod Bioacoustics: Neurobiology and Behaviour. Edinburgh University Press, Edinburgh. Farris, H. E. and Hoy, R. R. (20(10). Ultrasound sensitivity in the cricket, Etmemobius carol#ms (Gryllidae, Nemobiinae). J. Acoust. Soc. Am. 107, t727 1736. Farris, H. E., Forrest, T. G. and Hoy, R. R. (1997). The effects of calling song spacing and intensity on the attraction of flying crickets (Orthoptera: Gryllidae: Nemobiinae). J. Insect Behav. 10, 639 653. Faulkes, Z. and Pollack, G. S. (2000). Effects of inhibitory timing on contrast enhancement in auditory circuits in crickets (Teleogrvllus oceanicus). J. Neuroph)siol. 84, 1247-1255. Faulkes, Z. and Pollack, G. S. (2001). Mechanisms of frequency-specific responses of omega neuron 1 in crickets (Teleog13,11us oceanicus): a polysynaptic pathway for song? J. EaTs. Biol. 204, 1295 1305. Faure, P. A. and Hoy, R. R. (2000a). Auditory symmetry analysis. J. Erp. Biol. 203, 3209-3223. Faure, P. A. and Hoy, R. R. (2000b). Neuroethology of the katydid T-cell I. Tuning and responses to pure tones. J. E_vp. Biol. 203, 3225-3242. Faure, P. A. and Hoy, R. R. (2000c). Neuroethology of the katydid T-cell ll. Responses to acoustic playback of conspecific and predatory signals. J, Exp. Biol. 203, 3243 3254.

SOUND SIGNALLING IN ORTHOPTERA

261

Faure, P. A. and Hoy, R. R. (2000d). The sounds of silence: cessation of singing and song pausing are ultrasound-induced acoustic startle behaviors in the katydid Neoco,ocephalus ensi,~er (Orthoptera; Tettigoniidae). J, Comp. Physiol. A 186, 129 142. Feaver, M. N. (t 977). Aspects of the behavioral ecology of three species of Orc'helimum (Orthoptera: Tettigoniidae), Ph.D. thesis, University of Michigan, Ann Arbor. Fearer, M, N. (1983). Pair formation in the katydid Orchelimum ,igripes (Orthoptera: Tettigoniidae). In +Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects' (D. T, Gwynne and G. K. Morris, eds), pp. 2(15 239. Westview Press, Boulder, Colorado. Fenton, M. B. and Bell, G. P. (1981). Recognition of species of insectivorous bats by their echolocation calls. J. Mammal. 62, 233 243. Field, L. H. (1993). Structure and evolution of stridulatory mechanisms in New Zealand wetas (Orthoptera. Stenopelmatidae). hTt. d. Insect Morphol. Emhrvol. 22, 163 183. Field, L. H. (2001a). Aggression behaviour in New Zealand tree wetas. In "The Biology of Wetas, King Crickets and Their Allies" (L. H. Field, ed.), pp. 333-349. CABI Publishing, Wallingford, Oxfordshire. Field, L. H. (2001b). Stridulatory mechanisms and associated behaviour in New Zealand wetas. In 'The Biology of Wetas, King Crickets and Their Allies' (k. H. Field, ed.), pp. 271 295. CAB] Publishing, Wallingford, Oxfordshire. Field, L. H. and Bailey, W. J. (1997). Sound production in primitive Orthoptera from Western Australia: sounds used in defence and social communication in Ametrus sp. and Hadrogrylhlcris sp. (Gryllacrididae: Orthoptera). J. Nat. Hist. 31, 1127 1141. Field, L. H. and Glasgow, S. (2001). Defence behaviour. In +The Biology of Wetas, King Crickets and Their Allies" (L. H. Field, ed.), pp. 297 316. CABI Publishing, Wallingford, Oxfordshire. Field, L. H. and Jarman, T. H. (2001). Mating behaviour. In "The Biology of Wetas, King Crickets and Their Allies' (L. H. Field, ed.), pp. 317 332. CABI Publishing, Wallingford, Oxfordshire. Field, L. H. and Matheson. T. (1998). Chordotonal organs of insects. Adv. lnsect Physiol. 27, 1 228. Fischer, F. P., Schubert, H., Fenn, S. and Schulz, U. (1996). Diurnal song activity of grassland Orthoptera. Acta Oecol. 17, 345 364. Forrest, T. G. (1983). Calling songs and mate choice in mole crickets. In 'Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects" (D. T. Gwynne and G. K. Morris. eds), pp. 185 204. Westview Press, Boulder, Colorado. Forrest, T. G. (1991). Power output and efficiency of sound production by crickets. Behar. Ecol. 2, 327-338. Forrest, T. G. (1994). From sender to receiver: propagation and environmental effects on acoustic signals. Am. Zool. 34, 644 654. Forrest, T. G, and Raspet, R. (1994). Models of female choice in acoustic communication. Behav. Ecol. 5, 293 303. Fnhr, G., Ronacher, B., Krahe. R., Fcst, S., Shirley, S. G. and Rogaschewski, S. (1999). UV-laser ablation of sensory cells in living insects. Appl. Phys. A: Mater. Sci. Process. 68, 379 385. Fullard, J. H. (1984). Listening for bats: pulse repetition rate as a cue for a defensive behaviour in Cyc,ia re,era (Lepidoptera: Arctiidae). J. Comp. Physiol. ,4 154, 249 252. Fulton, B. B. (1934), Rhythm, synchronism and alternation in the stridulation of Orthoptera. J. Elisha Mitchell Sci. Soc. 50, 263 267.

262

D.J. ROBINSON AND M. J. HALL

Fuzessery, Z. M., Buttenhoff, P., Andrews, B. and Kennedy, J. M. (1993). Passive sound localization of prey by the pallid bat (Antrozous p. pallidus). J. Comp. Physiol. A 171, 767 777. Galliart, P. L. and Shaw, K. C. (1991). Role of weight and acoustic parameters, including nature of chorusing+ in the mating success of males of the katydid, Amblycorypha parvipennis (Orthoptera, Tettigoniidae). Fla Enlomol. 74, 453 464. Galliart, P. L. and Shaw, K. C. (1996). The effect of variation in parameters of the male calling song of the katydid, Amhlycoo'pha parvipennis (Orthoptera: Tettigoniidae), on female phonotaxis and phonoresponse. J. bisect Behav. 9, 841 855. Gilbert, F. and Elsner, N, (2000). Directional hearing of a grasshopper in the field. ,1. E.q~. Biol. 203, 983 993. Gorochov, A. V. (2001). In 'The Biology of Wetas, King Crickets and Their Allies" (L. H. Field, ed.), The higher classification, phylogeny and evolution of the superfamily Stenopelmatoidea. pp. 3-33. CABI Publishing, Wallingford, Oxfordshire. Gray, D. A. (1997). Female house crickets, Acheta domesticus, prefer the chirps of large males. Anita. Behav. 54, 1553 1562. Gray, D. A. and Cade, W. H. (1999). Sex+ death and genetic variation: natural and sexual selection on cricket song. Proc. Roy. Soc. Lond. Set. B 266, 707-709. Green, S. V. (1995). Song characteristics of certain Namibian grasshoppers (Orthoptera: Acrididae: Gomphocerinae). ,~17. Entomol. 3, 1-6. Greenfield, M. D+ (1990). Evolution of acoustic communication in the genus Neoconocephalus: discontinuous songs, synchrony and interspecific interactions, in +The Tettigoniidae: Biology, Systematics and Evolution' (W. J. Bailey and D. C. F. Rentz, eds), pp. 71-97. Crawford House Press, Bathurst, Australia. Greenfield, M. D. (1994). Cooperation and conflict in the evolution of signal interactions. Ann. Rev. Ecol, ~vst. 2& 97-126. Greenfield, M. D. (1997). Acoustic communication in Orthoptera. In +The Bionomics of Grasshoppers, Katydids and Their Kin" (S. K. Gangwere, M. C. Muralirangan and M. Muralirangan, eds), pp. 197-230. CAB International, New York and Wallingford, Oxfordshire. Greenfield, M. D. and Minckley, R. L. (1993). Acoustic dueling in tarbush grasshoppers: settlement of territorial contests via alternation of reliable signals. Ethology 95, 309 326. Greenfield, M. D. and Roizen, 1. (1993). Katydid synchronous chorusing is an evolutionary stable outcome of female choice. Nature 364, 618 620. Greenfield, M. D. and Shaw, K. C. (1983). In "Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects" (D. T. Gwynne and G. K. Morris, eds), Adaptive significance of chorusing with special reference to the Orthoptera. pp. 1 27. Westview Press, Boulder, Colorado. Greenfield, M. D. and Shelly, T. E. (1985). Alternative mating strategies in a desert grasshopper: evidence for density dependence. Anita. Behav. 33, 1192 1210. Greenfield, M. D., Tourtellot, K. K. and Snedden, W. A. (1997). Precedence effects and the evolution of chorusing. Proc. Roy. Soc. Lond. Ser. B 264, 1355 1361. Griffin, D. R. (1971). The importance of atmospheric attenuation for the echolocation of bats (Chiroptera). An#n. Behav. 19, 55-61. Gwynne, D. T. (1984). Sexual selection and sexual differences in mormon crickets (Orthoptera: Tettigoniidae, Anahrus simplex). Evolution 3& 1011 1022. Gwynne, D. T. (1988). Courtship feeding and the fitness of female katydids (Orthoptera: Tettigoniidae). Evolution 42, 545-555. Gwynne, D. T. (1990). The katydid spermatophore: evolution of a parental investment. In 'The Tettigoniidae: Biology, Systematics and Evolution' (W. J. Bailey and D. C. F. Rentz, eds), pp. 27 40. Crawford House Press, Bathurst, Australia.

SOUND SIGNALLING IN ORTHOPTERA

263

Gwynne, D. T. (1995). Phylogeny of the Ensifera (Orthoptera): a hypothesis supporting multiple origins of acoustical signalling, complex spermatophores and maternal care in crickets, katydids and weta. J. Orth. Res. 4, 203 218. Gwynne, D. T. (1997). The evolution of edible "sperm sacs' and other forms of courtship feeding in crickets, katydids and their kin (Orthoptera: Ensifera). In "The Evolution of Mating Systems in Insects and Arachnids" (J. C. Choe and B. J. Crespi, eds), pp. 110 129. Cambridge University Press, Cambridge. Gwynne, D. T. (2001). "Katydids and Bush-crickets: Reproductive Behavior and Evolution of the Tettigoniidae'. Cornell University Press, Ithaca, New York. Gwynne, D. T. and Bailey, W. J. (1988). Mating system, mate choice and ultrasonic calling in a zaprochiline katydid (Orthoptera: Tettigoniidae). Behaviour 105, 202 223. Gwynne, D. T. and Bailey, W. J. (1999). Female female competition in katydids: sexual selection lk)r increased sensitivity to a male signal'? Evolution 53, 546-551. Gwynnc, D. T. and Morris, G. K. (1986), Heterospecific recognition and behavioral isolation in acoustic Orthoptera (Insecta). Evo/. Theory 8, 33 38, Gynther, 1. C. and Peerson, K. G. (1989). An ewfluation of the role of identified interneurons in triggering jumps and kicks in the locust. J. Neurophysiol. 61, 45 57. Hack, M. A. (1998). The energetics of male mating strategies in field crickets (Orthoptera: Gryllinae: Gryllidae). J. Insect Behav. 11, 853 867. Hale. R. J. and Rentz, D. C. F. (2001). The Gryllacrididae: an overview of the world fauna with emphasis on Australian examples. In 'The Biology of Wetas, King Crickets and Their Allies" (L. H. Field, ed,), pp. 95 110. CABI Publishing, Wallingford, Oxfordshire. Halex, H., Kaiser, W. and Kalmring, K. (1988). Projection areas and branching patterns of the tympanal receptor-cells in migratory locusts. Locusta migratoria and Schistocerca gregaria. ('ell Tissue Res. 253, 517 528. Hamilton, W. D. (1971). Geometry for the selfish herd. J. Theor. Biol. 31, 295 311. Hardt, M., Sporhase-Eichmann, U. and Agricola, H. (1994). Serotonin immunoreactivity of identified sensory interneurons of the cricket. Soc. Neurosci. Ahstr. 20, 913. Harrison, L.. Horseman, G. and Lewis, D. B. (1988). The coding of the courtship song by an identified auditory interneurone in the cricket Teleo~ryllus oceanicus (Le Guillou)..I. Comp, Physiol. A 163, 215 225. Hartley, J. C. (1993). Acoustic behaviour and phonotaxis in the duetting ephippigerines. S:cropleur,s md~rei and Steropleurtts staff (Tettigoniidae). Zoo/..I. Lim~. Soc. 107, 155 167. Hartlcy. J. C. and Robinson, D. J. (1976). Acoustic behaviour of both sexes of the speckled bush cricket LtT)tophyes puncmtissima. Physiol. Entomol. 1, 21 25. Hartley, J. C. and Stephen, R. O. (1989). Temporal changes in the quality of the song of a bush cricket..I. E\p, Biol. 147, 189 202. Hartley, J. C., Jatho, M.. Kahnring, K., Stephen, R. O. and Sch6rder, H. (2000). Contrasting sound production in Tettigoniidae. J. Orth. Res. 9, 121 127. Haskell, P. T. (1957). Stridulation and associated behaviour in certain Orthoptera. 1. Analysis of the stridulation of, and behaviour between, males. Brit. J. Anita. Behav. 5, 139 148. Haynes, K. F. and Yeargan, K. V. (1999), Exploitation of intraspecific communication systems: illicit signalers and receivers. Am1. Entomol. Soc. Ant. 92, 960 970. Hedrick, A. V. (200(/). Crickets with extravagant mating songs compensate for predation risk with extra caution. Proc. Roy. Soc. L(md. Set'. B 267, 671 675. Hedrick, A. V. and Dill, L. M. (1993). Mate choice by female crickets is influenced by predation risk. Alfim. Behav. 46. 193 196.

264

D.J. ROBINSON AND M. J. HALL

Hedwig, B. (1994). A cephalothoracic command system controls stridulation in the acridid grasshopper Omoceslus viridulus L. J. Neurophysiol. 72, 2015 2025. Hedwig, B. (1995). Control of grasshopper stridulation by descending brain neurons. Verh. Dt. Zool. Gesellsch. 88, 181-190. Hedwig, B. (1996). A descending brain neuron elicits stridulation in the cricket Gryllus bimaculatus (de Geer), Naturwissenschq/ten 83, 428M29. Hedwig, B. (2001). Singing and hearing: neuronal mechanisms of acoustic communication in orthopterans. Zoology 103, 140-149. Hedwig, B. and Heinrich, R. (1997). Identified descending brain neurons control different stridulatory motor patterns in an acridid grasshopper. J. Comp. Physiol. A 180, 285-294. Heinrich, R. and Elsner, N. (1997). Central nervous control of hindleg coordination in stridulating grasshoppers. J. Comp. Physiol. A 180, 257 269. Heinrich, R., Hedwig, B. and Elsner, N. (1997). Cholinergic activation of stridulatory behaviour in the grasshopper Omocestus vMdulus (L.). J. Exp. Biol. 200, 1327 1337. Heinrich, R., Jatho. M. and Kalmring, K. (1993). Acoustic transmission characteristics of the tympanal tracheas of bush-crickets (Tettigoniidae). 2. Comparative studies of the tracheas of seven species. J, Acoust. Soc. Am. 93, 3481-3489. Heinrich, R., Rozwod, K. and Elsner, N. (1998). Neuropharmacological evidence for inhibitory cephalic control mechanisms of stridulatory behaviour in grasshoppers. J. Comp. Physiol. A 183, 389-399. Heller, K.-G. (1984). Zur Bioacustik und Phylogenie der Gattung Poecilimon (Orthoptera: Tettigoniidae: Phaneropterinae). Zoo/. Jb. Syst. I 11, 69 117. Heller, K.-G. (1995). Acoustic signalling in palaeotropical bush-crickets (Orthoptera, Tettigonioidea, Pseudophyllidae): does predation pressure by eavesdropping enemies differ in the Palaeotropics and Neotropics? J. Zool. 237, 469 485. Heller, K.-G. and Arlettaz, R. (1994). Is there a sex ratio bias in the bushcricket prey of the scops owl due to predation on calling males? J. Orth. Res. 2, 41 42. Heller, K.-G. and Volleth, M. (1995). Community structure and evolution of insectivorous bats in the Palaeotropics and Neotropics. Tropic. Ecol. 11,429-442. Heller, K.-G. and von Helversen, D. (1986). Acoustic communication in phaneropterid bushcrickets: species specific delay of female stridulatory response and matching male sensory time window. Behav. Ecol. Sociohiol. 18, 189 198. Heller, K.-G. and von Helversen, D. (1993). Calling behavior in bush-crickets of the genus Poecilimon with differing communication systems (Orthoptera, Tettigonioidea, Phaneropteridae). J. h~sect Behav. 6, 361-377. Heller, K.-G., von He[versen, O. and Sergejeva, M. (1997). Indiscriminate response behaviour in a female bushcricket: sex role reversal in selectivity of acoustic mate recognition? Naturwissenschaften 84, 252-255. Hennig, R. M. (1988). Ascending auditory interneurons in the cricket Teh'ogo, Ilus commodus (Walker): comparative physiology and direct connections with afferents. J. Comp. Physiol. A 163, 135 143. Hennig, R. M. and Weber, T. (1997). Filtering of temporal parameters of the calling song by cricket females of two closely related species: a behavioral analysis. J. Comp. Ph)siol. A 180, 621 630. Hill, P. S. M. (1999). Lekking in Gwllotalpa ma/or, the prairie mole cricket (Insecta: Gryllotalpidae). Ethology 105, 53l 545. Hill, P. S. M. (2000). Elements of the acoustic repertoire of the prairie mole cricket (Orthoptera: Gryllotalpidae: Gryllotalpa mq/or Suassure). J. Kuns. Entomol. Soc. 73, 95 102. Hill, P. S. M. and Shadley, J. R. (1997). Substrate vibration as a component of a calling song. Naturwissenschal?en 84, 460 463.

SOUND SIGNALLING IN ORTHOPTERA

265

Hissman, K. (1990). Strategies of mate finding in the European field cricket (Gryllus campestris) at different population densities: a field study. Ecol. Entomo/. 15, 281 291. Hoback, W. W. and Wagner, E. W., Jr (1997). The energetic costs of calling in the variable field cricket, Gwllus lineaticeps. Physiol. Entomol. 22, 286 290. Hoffmann, E. and Jatho, M. (1995). The acoustic trachea of tettigoniids as an exponential horn: theoretical calculations and bioacoustical measurements. J. Acoust. Soc. Am. 98, 1845-1851. Hoy, R. R. (1992a). The evolution of hearing in insects as an adaptation to predation from bats. In "The Evolutionary Biology of Hearing' (D. B. Webster, R. R. Fay and A. N. Popper, eds), pp. 115 129. Springer-Verlag, New York. Hoy, R. R. (1992b). Genetic and temperature coupling between sender and receiver in acoustic biocommunication systems. Semin. Neurosci. 4, 377-383. Hoy, R. R. and Robert, D. (1996). Tympanal hearing in insects. AmT. Rev. Enlomol, 41, 433~50. Hoy, R. R,, Nolen, T. G. and Brodfuehrer, P. (1989). The neuroethology of acoustic startle and escape in flying insects. J. Exp. Biol. 146, 287 306. Hnber, F. (1960). Untersuchungen fiber die Funktion des Zentralnervensystems und insbesondere des Gehirnes bei der Fortbewegung und der Lauterzeugung der Grillen. Z. I*ergl. Physiol. 44, 60 132. Huber, F., Moore, T. E. and Loher, W. (1989). 'Cricket Behavior and Neurobiology'. Cornell University Press, Ithaca, NY. Hutchings, M. and Lewis, B. (1983). Insect sound and vibration receptors, in "Bioacoustics: A Comparative Approach' (B. Lewis, ed.), pp. 181 2/)5. Academic Press, London. hnaizumi, K. and Pollack, G. S. (1999). Neural coding of sound frequency by cricket auditory receptors. J. Neurosci. 19, 1508 1516. Jacobs, K., Otte, B. and Lakes-Harlan, R. (1999). Tympanal receptor cells of Schistocerca gregaria: correlation of soma positions and dendrite attachment sites, central projections and physiologies. J. Evp. Zool. 283, 270 285. Jamieson, I. G., Forbes, M. R. and McKnight, E. B. (2000). Mark recapture study of mountain stone weta ttemideina maori (Orthoptera: Anostostomatidae) on rock tor 'islands'. N . Z . J . Ecol. 24, 209-214. Jatho, M., Weidemann, S. and Kretzen, D. (1992). Species specific sound production in three ephippigerine bush-crickets. Behav. Process. 26, 31~42. Jatho, M., Schul, J., Stiedl, O. and Kalmring, K. (1994). Specific differences in sound production and pattern recognition in tettigoniids. Behav. Procesx. 31, 293 300. Johns, P. M. (1997). The Gondwanaland weta: family Anostostomatidae (formerly in Stenopelmatidae, Henicidae or Mimnermidae): nomenclature problems, world checklist, new genera and species. J. Orth. Res. 6, 125-138. Johnstone, R. A. (1997). The evolution of animal signals. In "Behavioural Ecology: an Evolutionary Approach" (J. R. Krebs and N. B. Davies, eds), pp. 155 178. Blackwell Scientilic Publications, Oxford. Johnstone, R. (2000). Conflicts of interest in signal evolution. In 'Animal Signals: Signalling and Signal Design in Animal Communication" (Y. Espmark, T. Amundsen and G. Rosenqvist, eds), pp. 465 485. Tapir Academic Press, Trondheim. Norway. Jones, M. D. R. (1966). The acoustic behaviour of the bush cricket Pholidoptera ,ariseoaptera. 1. Alternation, synchronism, and rivalry between males..I.E.vp. Biol. 45. 15 30.

266

D.J. ROBINSON AND M. J. HALL

Kalmring, K. and K/ihne, R. (1980). The coding of airborne-sound and vibration signals in bimodal ventral-nerve cord neurons of the grasshopper Tettigonia cantans. J. Comp. Ph3siol. A 139, 267 275. Kahnring, K. and Kfihne, R. (1983). The processing of acoustic and vibrational information in insects. In 'Bioacoustics: a Comparative Approach' (B. Lewis, ed.), pp. 261-282. Academic Press, London. Kalmring, K., Rehbein, H.-G. and Kfihne, R. (1979). An auditory giant neuron in the ventral nerve cord of Detritus verrucivorus (Tettigoniidae). J. Comp. Physiol. A 132, 225-234. Kalmring, K., R6ssler, W., Ebendt, R., Ahi, J. and Lakes, R. (1993). The auditory receptor organs in the forelegs of bush-crickets: physiology, receptor cell arrangement, and morphology of the tympanal and intermediate organs of three closely related species. Zool. Jb.: Abteil. Allgem. Zool. Physiol. Tiere 97, 75 94. Kalmring, K., R6ssler, W. and Unrast, C. (1994). Complex tibial organs in the forelegs, midlegs, and hindlegs of the bush-cricket Gampsocleis gratiosa (Tettigoniidae): comparison of the physiology of the organs. J. Exp. Zool. 270, 155 -161. Kalmring, K., R6ssler, W., Hoffmann, E., Jatho, M. and Unrast, C. (1995). Causes of the differences in detection of low-frequencies in the auditory receptor organs of two species of bush-crickets. J. Exp. Zool. 272, 103 115. Kalmring, K., Hoffmann, E., Jatho, M., Sickmann, T. and Grossbach, M. (1996). The auditory-vibratory sensory system of the bushcricket PoO~sarcus denticauda (Phaneropterinae, Tettigoniidae). 2. Physiology of receptor cells. J. Exp. Zool. 276, 3 l 5-329. Kalmring, K., Jatho, M., Rassler, W. and Sickmann, T. (1997). Acousto vibratory communication in bushcrickets (Orthoptera: Tettigoniidae). Entomol. Gen. 21, 265 291. Keuper, A. and Kfihne, R. (1983). The acoustic behavior of the bushcricket Tetfigonia cantans. 2. Transmission of airborne-sound and vibration signals in the biotope. Behav. Process. 8, 125-145. Keuper, A., Otto, C., Latimer, W. and Schatral, A. (1985). Airborne sound and vibration signals of bushcrickets and locusts; their importance for the behaviour in the biotope. In "Acoustic and Vibrational Communication in Insects" (K. Kahnring and N. Elsner, eds), pp. 135 142. Paul Parey, Berlin and Hamburg. Kiflawi, M. and Gray, D. A. (2000). Size-dependent response to conspecific mating calls by male crickets. Proc. R~O~. Soc. Lond. Ser. B 267, 2157 2161. Kilduff I. A. (2000). "Mate Choice and Reproductive Success in the Speckled Bushcricket, Leptoph)es punctatissima', Ph.D. thesis, Open University, Milton Keynes, UK. Kleindienst, H.-U. (1978). Schallbeugung und -reflexion am Grillenk6rper im Frequenzbereich 5 20kHz. Verb. Dr. Zool. Gesellsch. 71, 160. Kok, O. B. and Louw, S. V. (2000). Avian and mammalian predators of Orthoptera in semi-arid regions of South Africa. South/U?. J. Wihtl. Res. 30, 122 128. Kolluru, G. R. (1999). Variation and repeatability of calling behavior in crickets subject to a phonotactic parasitoid fly. J. hTsect Behav. 12. 611 626. Koning, J. W. and Jamieson, I. G, (2001). Variation in size of male weaponry in a harem-defence polygynous insect, the mountain stone weta Hemidehta maori (Orthoptera: Anostostomatidae). N . Z . J . Zool. 28, 109-117. Lakes-Harlan, R. and Heller, K.-G. (1992). Ultrasound-sensitive ears in a parasitoid fly. Naturwissensch~{/ie, 79, 2 2 4 226. Lakes-Harhm, R., Bailey, W. J. and Schikorski, T. (1991). The auditory system of an atympanate bush-cricket Phasmodes ranatr(/ormes (Westwood) (Tettigoniidae, Orthoptera). J. Exp. Biol. 158, 307 324.

SOUND SIGNALLING IN ORTHOPTERA

267

Lakes-Harlan, R., Jacobs, K. and Heinrich, R. (1998). Identification of auditory interneurons in situ and in vitro by tracer injection into an afferent neuropil of Locusta migratoria, Naturwissenscha/?en 85, 240 243, Lang, F. and Eisner, N. (1994). Leg movement and hearing: biophysics and electrophysiology of the tympanal organ in Locusta migratoria. J. Comp. Physiol. A 175, 251 260. Larom, D., Garstang, M., Payne, K., Rasper, R. and Lindeque, M. (1997). The influence of surface atmospheric conditions on the range and area reached by animal vocalizations, J. Exp. Biol. 200, 421 431. Larsen, O. N. and Michelsen, A. (1978). Biophysics of the ensiferan ear. lII The cricket ear as a four input system. J. Comp. Physiol. A 123, 217 227. Latimer, W. (1981a). The acoustic behavior of Platvcleis alhoptmctata (Goeze) (Orthoptera, Tettigoniidae). Behaviour 76, 182-206. Latimer, W. (1981b). Variation in the song of the bush cricket Platycleis alhopunct(aa (Orthoptera, Tettigoniidae). J. Nat. Hist. 15, 245 263. Lehmann, G. U. C. and Heller, K.-G. (1998). Bushcricket song structure and predation by the acoustically orienting parasitoid fly Therohia leonidei (Diptera: Tachinidae: Ormiini). Behav. Ecol. Sociohiol. 43, 239-245. Lehmann, G. U. C. and Lehmann, A. W. (2000). Spermatophore characteristics in bushcrickets vary with parasitism and remating interval. Behav. Ecol. Sociohiol. 47, 393 399. Lehmann, G. U. C., Helter, K.-G. and Lehmann, A. W. (2001). Male bushcrickets favoured by parasitoid flies when acoustically more attractive for conspecific females (Orthoptera: Phanopteridae/Diptera: Tachinidae). Entomol. Gen. 25, 135-140. Lewis, D. B. (1974). The physiology of the tettigoniid ear II. The response characteristics of the ear to differential inputs: lesion and blocking experiments. J. £xt~. Biol. 60, 839 851. Lewis, D. B. (1992). The processing of auditory signals in the CNS of Orthoptera. In 'The Evolutionary Biology of Hearing" (D. B. Webster, R. R. Fay and A. N. Popper, eds), pp. 95-I 14. Springer-Verlag, New York. Libersat, F. and Hoy, R. R. (1991). Ultrasonic startle behavior in bushcrickets (Orthoptera, Tettigoniidae). J. Comp. Physiol. A 169, 507 514. Lin, Y. Z., Kahnring, K., Jatho, M., Sickmann, T. and R6ssler, W. (1993). Auditory receptor organs in the forelegs of Gampsocleis gratiosa (Tettigoniidae): morphology and function of the organs in comparison to the frequency parameters of the conspecific song. J. Exp. Zool. 267, 377-388. Lin, Y. Z., R6ssler, W. and Kalmring, K. (1994). Complex tibial organs in forelegs, midlegs, and hindlegs of the bush-cricket Gampsocleis :zratiosa (Tettigoniidae): comparison of morphology of the organs. J. Moq~hol. 221. t91 198. Lins, F. and Elsner, N. (1995). Descending stridulatory interneurons in the sub()esophageal ganglion of two grasshopper species. II: Influence upon the stridulatory patterns. J. Comp. Physiol. A 176, 823 833. Mclntyre, M. (2001). The ecology of some large weta species in New Zealand. In 'The Biology of Wetas, King Crickets and Their Allies' (L. H. Field. ed.), pp. 225 242. CABI Publishing, Wallingford, Oxfordshire. McKay, J. M. (1969). The auditory system of Homorocor37)hu.s (Tettigonioidea, Orthoptera). J. Exp. Biol. 51. 787 802. McKay, J. M. (1970). Central control of an insect auditory interneurone. J. Exp. Biol. 53. 137 146. Marquart, V. (1985). Local interneurons mediating excitation and inhibition onto ascending neurons in the auditory pathway of grasshoppers. Naturwixsenschq/ten 72, 42- 45.

268

D.J. ROBINSON AND M. J. HALL

Masaki, S., Kataoka, M., Shirato, K. and Nakagahara, M. (1987). Evolutionary differentiation of right and left tegmina in crickets. In "Evolutionary Biology of Orthopteroid Insects" (B. M. Bacetti, ed.), pp. 347-358. Ellis Horwood, Chichester. Mason. A. C. (1991). Hearing in a primitive ensiferan: the auditory system of Qvphoderris monstrosa (Orthoptera: Haglidae)..I. Comp. Physiol. A 168, 351 363. Mason, A. C. (1996). Territoriality and the function of song in the primitive acoustic insect Cvphoderris monstrosa (Orthoptera: Haglidae). Anita. Behav. 51,211 224. Mason, A. C. and Bailey, W. J. (1998). Ultrasound hearing and male-male communication in Australian katydids (Tettigoniidae: Zaprochilinae) with sexually dimorphic ears. Physiol. Entomol. 23, 139 149. Mason, A. C., Morris, G. K. and Wall, P. (I 991 ). High ultrasonic hearing and tympanal slit function in rainforest katydids. Naturwissenscha/~en 78, 365 367. Mason, A. C. and Schildberger, K. (1993). Auditory interneurons in Qvpho~k'rris monstrosa (Orthoptera, Haglidae). J. Comp. PIo,siol. A 171, 749 757. Mason, A. C., Forrest, T. G. and Hoy, R. R. (1998). Hearing in mole crickets (Orthoptera: Gryllotalpidae) at sonic and ultrasonic frequencies. J. Exp. Biol. 201, 1967 1979. Mason, A. C., Morris, G. K. and Hoy, R. R. (1999). Peripheral frequency mis-match in the primitive ensiferan Cyphoderris monstrosa (Orthoptera: Haglidae). J. Comp. Physiol. A 184, 543 551. Masters, W. M. (1979). Insect disturbance stridulation: its defensive role. Behav. Ecol. Sociohiol. 5, 187 200. Mbata, K. J. (1992). Some observations on the reproductive behavior of Acanthoplus ,~peiseri Brancsik (Orthoptera, Tettigoniidae, Hetrodinae). bisect Sci. Appl. 13, 19 26. Meixner, A. J. and Shaw, K. C. (1986). Acoustic and associated behavior of the coneheaded katydid, Neoconocephalus nebrascensis (Orthoptera, Tettigoniidae). Ann. Entomol. Soc. Am. 79, 554~565. Meyer, J. and Elsner, N. (1995). How respiration affects auditory sensitivity in the grasshopper (7~orthippus b~uttuhts (L.). J. Uomp. Physiol. A 176, 563 573. Meyer, J. and Hedwig, B. (1995). The influence of tracheal pressure changes on the responses of the tympanal membrane and auditory receptors in the locust Locusta migratorh, L. J. EYp. Biol. 198, 1327 1339. Michelsen, A. (1971a). The physiology of the locust ear I. Frequency sensitivity of single ceils in the isolated ear. Z. Vergl. Physiol. 71, 49 62. Michelsen, A. (1971b). The physiology of the locust ear 11. Frequency discrimination based upon resonances in the tympanum. Z. Vergl. Physiol. 71, 63 101. Michelsen, A. ( 1971 c). The physiology of the locust ear. III. Acoustical properties of the intact ear. Z. Verg/. Physiol. 71, 102 128. Michelsen, A. (1985). Environmental aspects of sound communication in insects. In "Acoustic and Vibrational Communication in Insects' (K. Kalmring and N. Eisner, eds), pp. 1-9. Paul Parey, Berlin and Hamburg. Michelsen, A. and Larsen, O. N. (1978). Biophysics of the ensiferan ear I. Tympanal vibrations in bush crickets (Tettigoniidae) studied with laser vibrometry. J. Comp. Physiol. A 123, 193 203. Michelsen, A. and Larsen, O. N. (1985). Hearing and sound, In 'Comprehensive Insect Physiology, Biochemistry and Pharmacology' (G. A. Kerkut and L. 1. Gilbert, eds), pp. 496 556. Pergamon Press, Oxford. Michelsen, A. and Rohrseitz, K. (1995). Directional sound processing and interaural sound transmission in a small and a large grasshopper. J. Exp. Biol. 198, 1817 1827.

SOUND SIGNALLING IN ORTHOPTERA

269

Michelsen, A., Heller, K.-G., Stumpner, A. and Rohrseitz, K. (1994a). A new biophysical method to determine the gain of the acoustic trachea in bush-crickets. J. Comp. Physiol. A 175, 145 151. Michelsen, A., Popov, A. V. and Lewis, B. (1994b). Physics of directional hearing in the cricket GwIlus himaculatus. J. Comp. Physiol. A 175, 153 164. Miller, L. A. (1977). Directional hearing in the locust Schistocerca gre,gaHa Forskfil (Acrididae, Orthoptera). J. Comp. Physiol. A 119, 85 98. Minckley, R. L. and Greenfield, M. D. (1995). Psychoacoustics of female phonotaxis and the evolution of male signal interactions in Orthoptera. Ethol. Ecol. Evol. 7, 235 243. Minckley, R. L., Greenfield, M. D. and Tourtellot, M. K. (1995). Chorus structure in tarbush grasshoppers: inhibition, selective phonoresponse and signal competition. Anita. Behav. 50, 579 594. Moiseff, A., Pollack, G. S. and Hoy, R. R. (1978). Steering responses of flying crickets to sound and ultrasound: mate attraction and predator avoidance. Proc. Nail A~ad. Sci. U.S.A. 75, 4052 4056. Moiler, A. P. (2001). Female preference for symmetric calls in a grasshopper. Ethol. Ecol. Evol. 13, 261 272. Moller, A. P. and Swaddle, J. P. (1997). "Asymmetry, Developmental Stability, and Evolution'. Oxford University Press, Oxford. Monteith, G. B. and Field, L. H. (2001). Australian king crickets: distribution, habitats and biology (Orthoptera: Anostostomatidae). In 'The Biology of Wetas, King Crickets and Their Allies" (L. H. Field, ed.), pp. 79 94. CABI Publishing, Wallingford, Oxfordshire. Morgan-Richards, M. and Gibbs, G. W. (2001). A phylogenetic analysis of New Zealand giant and tree weta (Orthoptera: Anostostomatidae: Deitmcrida and Hemideina) using morphological and genetic characters. Im,ertehr. Taxon. 15, 1-12. Morris. G. K. and Fullard, J. H. (1983). Random noise and congeneric discrimination in ComWel~halus (Orthoptera: Tettigoniidae). In "Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects' (D. T. Gwynne and G. K. Morris, eds), pp. 73 96. Westview Press, Boulder, Colorado. Morris, G. K., Kerr, G. E. and Fullard, G. H. (1978). Phonotactic preferences of female meadow katydids (Orthoptera: Tettigoniidae, Conocephah~s ni~,,ropleurum). Can. J. Zoo/. 56, 1479-1487. Morris. G. K., Mason, A. C., Wall, P. and Belwood, J. J. (1994). High ultrasonic and tremulation signals in neotropical katydids (Orthoptera, Tettigoniidae). J. Zoo/. 233, 129 163. Mfiller, P. and Robert, D. (2001). A shot in the dark: the silent quest of a free-flying phonotactic fly..I.E.vp. Biol. 204, 1039 1052. Nelson, C. M. and Nolen, T. G. (1997). Courtship song, male agonistic encounters, and female mate choice in the house cricket, Acheta domesticus (Orthoptera: Gryllidae). ,I. Insect Behav. 10, 557 570. Nocke, H. (1971). Biophysik der Schallerzeugung dutch die Vorderflfigel der Grillen. Z. l,'er~'/. Physiol. 74, 272 314. Nocke, H. (1975). Physical and physiological properties of the tettigoniid ('grasshopper') ear. J. Comp. Physiol. A 100, 25 57. Nolan. T. G. and Hoy, R. R. (1986a), Phonotaxis in flying crickets. 1. Attraction to the calling song and avoidance of batqike ultrasound are discrete behaviours. J. Comp. Physiol. A 159, 423M-39. Nolam T. G. and Hoy, R. R. (1986b). Phonotaxis in flying crickets. 11. Physiological mechanisms of two-tone suppression of the high frequency avoidance steering behavior by the calling song. J. ('Olllp. Phvsiol, A 159, 441-456.

270

D.J. ROBINSON AND M. J. HALL

Ocker, W., Hedwig, B. and Elsner, N. (1996). Pharmacological induction and modulation of stridulation in two species of acridid grasshoppers. J. Exp. Biol. 198, 17011710. Oldfield, B. P. (1982). Tonotopic organization of auditory receptors in Tettigoniidae (Orthoptera: Ensifera). J. Comp. Physiol. A 147, 461 469. Oldfield, B. P. (1984). Physiology of auditory receptors in two species of Tettigoniidae (Orthoptera: Ensifera). Alternative tonotopic organizations of the auditory organ. J. Comp. Physiol. A 155, 689 696. Oldfield, B. P. (1985). The tuning of auditory receptors in bush crickets. Hear. Res. 17, 27 35. Oldfield, B. P. and Hill, K. G. (1983). The physiology of ascending auditory interneurons in the tettigoniid Caedicia simplex (Orthoptera: Ensifera): response properties and a model of integration in the afferent auditory pathway. J. Comp. Physiol. A 152, 495-508. Otte, D. (1977). Communication in Orthoptera. In 'How Animals Communicate (T. A, Sebeok, ed.), p..p. 334-361. Indiana University Press, Bloomington, Indiana. Otto, D. (1978). Anderung von Gesangsparametern bei der Grille (Gryllus campestris) nach Injektion von Pharmaka ins Gehirn. Verh. Dr. Zool. Gesellsch. 71, 245. Otte, D. (1990). The relation between hearing and flying in insects. Entomol. News 101, 29 34. Otte, D. (1994). 'The Crickets of Hawaii: Origin, Systematics and Evolution'. The Orthopterists' Society, Academy of Natural Sciences of Philadelphia, Philadelphia. Otte, D. and Loftus-Hills, J. (1979). Chorusing in Syrbula (Orthoptera: Acrididae). Cooperation, interference competition, or concealment? Entomol, News 90, 159 165. Parker, G. A. (1983). Mate quality and mating decisions. In ~Mate Choice" (P. Bateson, ed.), pp. 141-164. Cambridge University Press, Cambridge. Pollack, G. S. (1994). Synaptic inputs to the omega neuron of the cricket Teleog~3,11us oceanicus: differences in EPSP wave-forms evoked by low and high sound frequencies. J. Comp. Physiol. A 174, 83--89. Pollack, G. S. (1998). Neural processing of acoustic signals. In ~Comparative Hearing: Insects' (R. R. Hoy, A. N. Popper and R. R. Fay, eds), pp. 139 196. SpringerVerlag, New York. Pollack, G. S. and Faulkes, Z. (1998). Representation of behaviorally relevant sound frequencies by auditory receptors in the cricket Teleogo,llus oceanicus. J. Exp. Biol. 201, 155-163. Pollack, G. S. and Imaizumi, K. (1999). Neural analysis of sound frequency in insects. Bioessays 21,295-303. Popov, A. V., Markovich, A. M. and Andjan, A. S. (1978). Auditory interneurons in the prothoracic ganglion of the cricket, Go'llus bimaculatus. I. The large segmental auditory neurons (LSAN). J. Comp. Physiol. A 126, 183 192. Popov, A. V., Michelsen, A. and Lewis, B. (1994). Changes in the mechanics of the cricket ear during the early days of adult life. J. Comp. Physiol. A 175, 165 170. Poulet, J. F. A. and Hedwig, B. (200l). Tympanic membrane oscillations and auditory receptor activity in the stridulating cricket Gl~Hlus bimaculatus. J. Exp. Biol. 204, 1281 -1293. Prestwich, K. N. (1994). The energetics of acoustic signaling in anurans and insects. Am. Zool. 34, 625-643. Prestwich, K. N., Lenihan, K. M. and Martin, D. M. (2000). The control of carrier frequency in cricket calls: a refutation of the subalar-tegminal resonance/auditory feedback model. J. Exp. Biol. 203, 585-596. Price, D. K. and Hansen, T. F. (1998). How does offspring quality change with age in male Drosophila mehmogaster? Behav. Genet. 28, 395M02.

SOUND SIGNALLING IN ORTHOPTERA

271

Prier, K. R. and Boyan, G. S. (2000). Synaptic input from serial chordotonal organs onto segmentally homologous interneurons in the grasshopper Schistocerca <wegaria. J. Insect Physiol. 46, 297 312. Pye, J. D. (1979). Why ultrasound? Endeavour 3, 57 62. Ragge, D. R. (1987). Speciation and biogeography of some southern European Orthoptera, as revealed by their songs. In 'Evolutionary Biology of Orthopteroid Insects" (B. M. Bacetti, ed.), pp. 418~427. Ellis Horwood, Chichester. Ragge, D. R. and Reynolds, W. J. (1998). ~The Songs of the Grasshoppers and Crickets of Western Europe'. Harley Books, Colchester. Reinhold, K. (19943. Inheritance of body and testis size in the bush-cricket PoecilimoJt veluchiamts Ramme (Orthoptera, Tettigoniidae) examined by means of subspecies hybrids. Biol. J. Limt. Soc. 52, 305 316. Reinhold, K. (1999). Energetically costly behaviour and the evolution of resting metabolic rate in insects. Funct. Ecol. 13, 217 224. Rentz, D. C. (1975). Two new katydids of the genus Melanonotus t¥om Costa Rica with comments on their life history strategies (Tettigoniidae: Pseudophyllinae). Entomol. News 86, 129 140. Rheinlaender, J. (1984). Das akustische Orientierungsverhahen von Heuschrecken. Grillen und Fr6schen: eine vergleichende neuro- und verhaltensphysiologische Untersuchung, Habilitation thesis, Ruhr-University, Bochum. Rheinlaender, J. and R6mer, H. (1980). Bilateral coding of sound direction in the CNS of the bushcricket Tettigonia vMdissima L. (Orthoptera, Tettigoniidae)..l. Conq~. Physiol. A 140, 101 111. Rheinlaender, J. and R6mer, H. (19863. Insect hearing in the held 1. The use ofidentihed nerve cells as ~biological microphones'..I. Comp. Physiol. A 158, 647 651. Rheinlaender, J., Kalmring, K. and R6mer, H. (19723. Akustische Neuronen mit TStruktur im Bauchmark von Tettigoniiden. Z. I'er~,l. Physiol. 77, 208 224. Riede, K. (19873. A comparative study of mating behaviour in some neotropical grasshoppers (Acridoidea). Ethology 76, 265 296. Riede, K., Kiimper, G. and H6fler, I. (19903. Tympana, auditory thresholds, and projection areas of tympanal nerves in singing and silent grasshoppers (lnsecta, Acrididae). Zoomorphology 109, 223 230. Ritchie, M. G., Couzin, I. D. and Snedden, W. A. (1995). What's in a song: female bush-crickets discriminate against the song of older males. Proc. Roy. Soc. Lond. Set'. B 262. 21 27. Ritchie, M. G., Sunter, D. and Hockham, E. R. (1998). Behavioral components of sex role reversal in the tettigoniid bushcricket Ephil~pi~er ~7~hil)piger. J. htsect Behav. 1 I. 481 491. Robert, D. (1989). The auditory behaviour of flying locusts. J. Exp. Biol. 147. 279 301. RoberL D. J., Amoroso, J. and Hoy, R. R. (1992). The evolutionary convergence of hearing in a parasitoid fly and its cricket host. Science 258, 1135 1137. Robinson, D. J. (1980). Acoustic communication between the sexes of the bush cricket, Leptophyes punctatissima. Physiol. I:nlomol. 5. 183 189. Robinson, D. J. (1990). Acoustic communication between the sexes in bushcrickets. In "The Tettigoniidae: Biology, Systcmatics and Evolution' (W. J. Bailey and D. C. F. Rentz, cds), pp. 112 129. Crawt\~rd House Press, Bathurst, Australia. Robinson, D. J., Rheinlaender, J. and Hartley, J. C. (1986). Temporal parameters of male female sound comlnunication in Lel~mphyes puncmtis.~ima. Physiol. Entomol, 11. 317 323. R6mer, H. (1976). Die infornlationsverarbeitung tymanaler Rezeptorelemente yon kocttsta mi,vraloria (Acrididae, Orthoptera). J. Comp. Physiol. A 109, 101 122.

272

D, J. ROBINSON AND M. J. HALL

Ramer, H. (1983). Tonotopic organization of the auditory neuropile in the bushcricket Tettigonia viHdJssima. Nature 306, 60-62. Ramer, H. (1985). Anatomical representation of frequency and intensity in the auditory system of Orthoptera. In 'Acoustic and Vibrational Communication in Insects" (K. Kalmring and N. Eisner, eds), pp. 25 32. Paul Parey, Hamburg. Ramer, H. (1992). Ecological constraints for the evolution of hearing and sound communication in insects. In "The Evolutionary Biology of Hearing' (D. B. Webster, R. R. Fay and A. N. Popper, eds), pp. 79 93. Springer-Verlag, New York. R6mer, H. (1993). Environmental and biological constraints for the evolution of longrange signalling and hearing in acoustic insects. Phil. Trans. R. Soc. Lond. Ser. B 340, 179 185. ROmer, H. and Bailey, W. J. (1986). Insect hearing in the field II. Male spacing behaviour and correlated acoustic cues in the bushcricket Mygalopsis marki. J. Comp. Physiol. A 159, 627 638. R6mer, H. and Bailey, W. J. (1998). Strategies for hearing in noise: peripheral control over auditory sensitivity in the bushcricket Sciarasaga quadrata (Austrosaginae: Tettigoniidae). J. Exp. Biol. 201, 1023 1033. R6mer, H. and Krusch, M. (2000). A gain-control mechanism for processing of chorus sounds in the afferent auditory pathway of the bushcricket Telt~gonia viridissima (Orthoptera: Tettigoniidae). J. Comp. Physiol. A 186, 181 191. R6mer, H. and Lewald, J. (1992). High-frequency sound transmission in natural habitats: implications for the evolution of insect acoustic communication. Behav. Ecol. Sociobiol. 29, 437~44. R6mer, H., Marquart, V. and Hardt, M. (1988). Organization of a sensory neuropil in the auditory pathway of two groups of Orthoptera. J. Comp. Neurol. 275, 201215. R6mer, H., Hedwig, B. and Ott, S. (1997). Proximate mechanism of female preference for the leader male in synchronizing bushcrickets (Mecopoda ehmgata). In 'From Membrane to Mind: Proceedings of the 25th G6ttingen Neurobiology Conference' (N. Eisner and H. Wfissle, eds), p. 322. Thieme, Stuttgart. R6mer, H., Spickermann, M. and Bailey, W. (1998). Sensory basis for sound intensity discrimination in the bushcricket Requena verticalis (Tettigoniidae, Orthoptera). J. Comp, Physiol. A 182, 595-607. Ronacher, B. (1989). Stridulation of acridid grasshoppers after hemisection of thoracic ganglia: evidence for hemiganglionic oscillators. J. Crimp. Physiol. A 164, 723 736. Ronacher, B. (1991). Contribution of abdominal commissures in the bilateral coordination of the hindlegs during stridulation in the grasshopper Chorthippus dorsatus. J, Comp. Physiol. A 169, 191 200. Ronacher, B. and Krahe, R. (1997). Long songs of a short-horned grasshopper: how important is song duration for pattern recognition and localisation? In 'From Membrane to Mind: Proceedings of the 25th G6ttingen Neurobiology Conference" (N. Elsner and H. W~ssle, eds), p. 317. Thieme, Stuttgart. Ronacher, B. and Krahe, R. (1998). Song recognition in the grasshopper Chorthiplms higuttulus is not impaired by shortening song signals: implications for neuronal encoding..l. Comp. Physiol. A 183, 729 735. Ronacher, B., Krahe, R. and Hennig, R. M. (2000). Effects of signal duration on the recognition of masked communication signals by the grasshopper Chorthippus biguttulus. J. Comp. Physiol. A 186, 1065-1072. R6ssler, W. (1992a). Functional morphology and development of tibial organs in the legs I, II, and II1 of the bush cricket Ephipp~ger ephippiger (lnsecta, Ensifera). Zoomorphology 112, 181 188.

SOUND SIGNALLING IN ORTHOPTERA

273

R6ssler, W. (1992b). Postembryonic development of the complex tibial organ in the foreleg of the bush cricket Ephippiger ephippiger (Orthoptera, Tettigoniidae). Cell Tissue Res. 269, 505-514. Rotenberry, J. T., Zuk, M., Simmons, L. W. and Hayes, C. (1996). Phonotactic parasitoids and cricket song structure: an evaluation of alternative hypotheses. Evol. Ecol. 10, 233 243. Rowell, G. A. and Cade, W. H. (1993). Simulation of alternative male reproductive behavior: calling and satellite behavior in field crickets. Ecol. Model. 65, 265-280. Ryan, M. J. (1990). Sexual selection, sensory systems, and sensory exploitation. Ox. Surv. Evol. Biol. 7, 156 195. Ryan, M. J. and Keddy-Hector, A. (1992). Directional patterns of female choice and the role of sensory biases. Am. Nat. 139, $4 $35. Ryder, J. J. and Siva-Jothy, M. T. (2000). Male calling song provides a reliable signal of immune function in a cricket. Proc. Roy. Soe. Lond. Set'. B 267, 1171- 1175. Sakaluk, S. K. and Ivy, T. M. (1999). Virgin-male mating advantage in sagebrush crickets: differential male competitiveness or non-independent female mate choice? Behaviour 136, 1335-1346. Sales, G. and Pye, D. (1974). 'Ultrasonic Communication by Animals'. Chapman and Hall, London. Samways, M. J. (1976). Song modification in the Orthoptera. 1. Proclamation songs of Plao, cleis spp. (Tettigoniidae). Physiol. Entomol. 1, 131 149. Sandow, J. D. and Bailey, W. J. (1978). An experimental study of defensive stridulation in Mygalopsisjbrruginea Redtenbacher (Orthoptera: Tettigoniidae). Anita. Behav. 26. 1004 1011. Schatral, A. (1990). Body size, song frequency and mating success of male bush-crickets Requena vertiealis (Orthoptera, Tettigoniidae, Listrocelidinae) in the field. Anita. Behav. 40, 982-984. Schatral, A., Latimer, W. and Kalmring, K. (1985). The role of the song for spatial dispersion and agonistic contacts in male bushcrickets. In 'Acoustic and Vibrational Communication in Insects' (K. Kalmring and N. Elsner, eds), pp. 111 116. Paul Parey, Berlin and Hamburg. Schildberger, K. and H6rner, M. (1988). The function of auditory interneurons in cricket phonotaxis. I. Influence of hyperpolarization of identified neurons on sound localization. J. Comp, Physiol. A 163, 621 631. Schul, J. (1997). Neuronal basis of phonotactic behaviour in Tettigonia viridiss#~m: processing of behaviourally relevant signals by auditory afferents and thoracic interneurons. J. Comp. Physiol. A 180, 573 583. Schul, J. (1998). Song recognition by temporal cues in a group of closely related bushcricket species (genus Tettigonia). J. Comp. Physiol. A 183, 401~410. Schul, J., yon Helversen, D. and Weber, T. (1998). Selective phonotaxis in Tettigonia eantans and T. viridissima in song recognition and discrimination. J. Comp. Physiol. A 182, 687 694. Schul, J., Holderied, M., yon Helversen, D. and yon Helversen, O. (1999). Directional hearing in grasshoppers: neurophysiological testing of a bioacoustic model. J. Exp. Biol. 202, 121-133. Schul, J., Matt, F. and von Helversen, O. (2000). Listening for bats: the hearing range of the bushcricket Pha,eropterafalcata for bat echolocation calls measured in the field. Proc. Rot. Soc. Lond. Set'. B 267, 1711 1715. Schulze, W. and Schul, J. (2001). Ultrasound avoidance behaviour in the bushcricket Tettigo,ia viridissima (Orthoptera: Tettigoniidae). J. Exp. Biol. 204, 733 740. Schumacher, R. (1979). Zur funktionellen Morphologie des auditiven Systems der Laubheuschrecken (Orthoptera: Tettigonioidea). Entomol. Gen. 5, 321 356.

274

D.J. ROBINSON AND M. J. HALL

Schwabe, J. (1906). Beitrage zur Morphologie und Histologie der tympanalen Sinnerapparate der Orthopteren. Zoologica 20, 1-54. Seymour, C., Lewis, B., Larsen, O. N. and Michelsen, A. (1978). Biophysics of the ensiferan ear II. The steady state gain of the hearing trumpet in bush crickets. J. Comp. Physiol. A 123, 205 216. Shaw, K. C. (1968). An analysis of the phonoresponse of males of the true katydid, Pterophylla camell(/blia (Fabricius) (Orthoptera: Tettigoniidae). Behaviour 31, 203 260. Shaw, K. C. and Galliart, P. (1987). Acoustic and mating behavior of a Mexican katydid, Pteroplo, lla beltrani (Orthoptera, Tettigoniidae). Fla Entomo/. 70, 354-368. Shaw, K. C., Galliart, P. L. and Smith, B. (1990). Acoustic behavior of Amblycorypha parvipennis (Orthoptera, Tettigoniidae). Ann. Entomol. So('. Am. 83, 617 625. Shaw, K. L. and Herlihy, D. P. (2000). Acoustic preference functions and song variability in the Hawaiian cricket Laupala cerash, a. Pro('. Roy. So('. Lond. Ser. B 267, 577 584. Shen, J.-X. (1993). A peripheral mechanism for auditory directionality in the bushcricket Gampsoeleis gratiosa acoustic tracheal system. J. Acoust. So(. Am. 94, 1211 1217. Sickmann, T., Kalmring, K. and Muller, A. (1997). The auditory-vibratory system of the bushcricket Polysareus dentieauda (Phaneropterinae, Tettigoniidae). 1. Morphology of the complex tibial organs. Hear, Res. 104, 155 166. Simmons, L. W. (1986). lntermale competition and mating success in the field cricket, Gryllus bimaculatus (De Geer). Anita. Behav. 34, 567 579. Simmons, L. W. (1987). Heritability of a male character chosen by females of the field cricket Gryllus himaculatus. Behav. Ecol. So('iobiol. 21, 129 133. Simmons, L. W. (1988). The calling song of the field cricket Grylhts bhnaculatus (De Geer): constraints on transmission and its role in intermale competition and female choice. Anita. Behav. 36, 380 394. Simmons, L. (1993). Some constraints on reproduction for male bush crickets, Requeml verticalis: diet, size and parasite load. Behav. Ecol. Sociobiol. 32, 135 139. Simmons, L. W. (1995). Correlates of male quality in the field cricket, GJ3,llus campestris L.: age, size, and symmetry determine pairing success in field populations. Behav. Ecol. 6, 376-381. Simmons, L. and Bailey, W. J. (1990). Resource influenced sex roles of Zaprochiline tettigoniids (Orthoptera: Tettigoniidae). Evolution 44, 1853 1868. Simmons, L. W. and Bailey, W. J. (1993). Agonistic communication between males of a zaprochiline katydid (Orthoptera: Tettigoniidae). Behav. Ecol. 4, 364~368. Simmons, L. W. and Ritchie, M. G. (1996). Symmetry in the songs of crickets. Proe. Roy. Soc. Lond. Ser. B 263, 1305 1311. Simmons, L. W., Teale, R. J., Standish, R. J., Bailey, W. J. and Withers, P. C. (1992). Some costs of reproduction for male bushcrickets, Requena verticalis (Orthoptera, Tettigoniidae): allocating resources to mate attraction and nuptial feeding. Behav. Ecol. Sociohiol. 31. 57-62. Simons, A. M. and Roff, D. A. (1994). The effect of environmental variability on the heritabilities of traits of a field cricket. Evolution 48, 1637 1649. Smart, J. (1963). Explosive evolution and the phylogeny of insects. Proc. Lira1. So('. Lond. 1974, 125 126. Snedden, W. A. (1996). Lifetime mating success in male sagebrush crickets: sexual selection constrained by a virgin male mating advantage. Anita. Behav. 51. 11191125. Snedden, W. A. and Greenfield, M. D. (1998). Females prefer leading males: relative call timing and sexual selection in katydid choruses. Anita. Behav. 56, 1091 1098.

SOUND SIGNALLING IN ORTHOPTERA

275

Snedden, W. A. and Sakaluk, S. K. (1992). Acoustic signalling and its relation to male mating success in sagebrush crickets. Anita. Behav. 44, 633 639. Snedden, W. A., Greenfield, M. D. and Jang, Y. W. (1998). Mechanisms of selective attention in grasshopper choruses: who listens to whom? Behav. Ecol. Sociobiol. 43, 59 66. Souroukis, K., Cade, W. H. and Rowell, G. (1992). Factors that possibly influence variation in the calling song of field crickets: temperature, time, and male size, age, and wing morphology. Can. J. Zool. 70, 950 955. Spangler, H. G. (1984). Silence as a defense against predatory bats in two species of calling insects. Southwest Nat. 29, 481M88. Stephen, R. O. and Bailey, W. J. (1982). Bioacoustics of the ear of the bush cricket ttemisaga (Saginae). J. Acoust. Soc. Ant. 72, 13 25. Stephen, R. O. and Bennet-Clark, H. C. (1982). The anatomical and mechanical basis of stimulation and frequency analysis in the locust ear. J. Exp. Biol. 99, 279-314. Stephen, R. O. and Hartley, J. C. (1995a). Control of call carrier frequency in the bush cricket Ruspolia nitidula. Bioacoustics 6, 163 170. Stephen, R. O. and Hartlcy, J. C. (1995b). Sound production in crickets. J. Exp. Biol. 198, 2139 2152. Stiedl. O. and Bickmeyer, U. (1991). Acoustic behavior of Ephippiger ephippiger Fiebig (Orthoptera, Tettigoniidae) within a habitat of southern France. Behav. Process. 23. 125 135. Stiedl, O., Hoffmann, E. and Kalmring, K. (1994). Chirp rate variability in male song of Ephippigerida taeniata (Orthoptera, Ensifera). J. Insect Behav. 7, 171 181, Stiedl, O,, Stumpner, A., Mbungu, D. N., Atkins, G. and Stout, J. F. (1997). Morphology and physiology of local auditory interneurons in the prothoracic ganglion of the cricket Achela domestieus. J. Exp. Zool. 279, 43 53. St61ting, H. and Stumpner, A. (1998). Tonotopic organization of auditory receptors of the bushcricket Pholidoptera griseoaptera (Tettigoniidae, Decticinae). ('ell Tissue Res. 294, 377 386. Stumpner, A. (1997). An auditory interneurone tuned to the male song frequency in the duetting bushcricket Ancistrura nigrovittata (Orthoptera, Phaneropteridae). J. Exp. Biol. 200. 1089 1101. Stumpner, A. and Heller, K.-G. (1992). Morphological and physiological differences of the auditory system in three related bush-crickets (Orthoptera, Phaneropteridae, Poecilimon). Physiol. Entomol. 17, 73 80. Stumpner, A. and yon Helversen, D. (2001). Evolution and function of auditory systems in insects. Naturwissenschal~en 88, 159 170. Stumpner, A. and yon Helversen, O. (1992). Recognition of a two-element song in the grasshopper Chorthippus dorsatus (Orthoptera: Gomphocerinae). J. Comp. Physiol. A 171,405-412. Stumpner, A. and Meyer, S. (2001). Songs and the function of song elements in lk~ur duetting bushcricket species (Ensifera, Phaneropteridae, Barhitistes). J. lnsect Behav. 14, 511 534. Stumpner, A., Ronacher, B. and yon Helversen, O. (1991). Auditory interneurons in the mctathoracic ganglion of the grasshopper Chorthippus biguttulus. II. Processing of temporal patterns of the song of the male. J. Exp. Biol. 158, 411 430. Suga, N. (1966). Ultrasonic production and its reception in some neotropical Tettigoniidae. J. htsect Physiol. 12, 1039-1050. Suga, N. (1968). Neural responses to sound in a Brazilian mole cricket. J. Audit. Res. 8, 129 134. Suga, N. and Katsuki, Y. (1961). Central mechanism of hearing in insects. J. Kvp. Biol. 38, 545 558.

276

D.J. ROBINSON AND M. J. HALL

Tauber, E. (2001). Bidirectional communication system in katydids: the effect on chorus structure. Behav. Ecol. 12, 308- 312. Tauber, E. and Cohen, D. (1997). Acoustic communication and duet singing in the bush cricket Phaneroptera nana. Isr. J. Zool. 43, 117 118. Tauber, E., Cohen, D., Greenfield, M. D. and Pener, M. P. (2001). Duet singing and female choice in the bushcricket Phaneroptera nana. Behaviour 138, 411-430. Toms, R. B. (1985). Speciation in tree crickets (Gryllidae: Oecanthinae). In 'Species and Speciation' (E. S. Vrba, ed.), pp. 109 114. Transvaal Museum, Pretoria. Toms, R. B. (2001). South African king crickets (Anostostomatidae). In ~The Biology of Wetas, King Crickets and Their Allies" (L. H. Field, ed.), pp. 73-78. CABI Publishing, Wallingford, Oxfordshire. Tuckerman, J. F., Gwynne, D. T. and Morris, G. K. (1993). Reliable acoustic cues for female mate preference in a katydid (Scudderia curvicauda, Orthoptera, Tettigoniidae). Behav. Ecol. 4, 106 113. Tuttle, M. D. and Ryan, M. J. (1982). The role of synchronized calling, ambient light, and ambient noise, in anti-bat-predator behavior of a treefrog. Behav. Ecol. Sociobiol. 11, 125 131. Ulagaraj, S. M. (1976). Sound production in mole crickets (Orthoptera: Gryllotalpidae: Scapteriscus). Ann. Entomol. So~. Am. 69, 299-306. van Staaden, M. J. and R6mer, H. (1997). Sexual signalling in bladder grasshoppers: tactical design for maximizing calling range. J. Evp. Biol. 200, 2597 2608. van Staaden, M. J. and R6mer, H. (1998). Evolutionary transition from stretch to hearing organs in ancient grasshoppers. Nature 394, 773-776. von Helversen, D. (1972). Gesang des M/innchens und Lautscheme des Weibchens bei der Feldheuschrecke Chorthippus biguttulus (Orthoptera, Acrididae). Z. l/el2~,,/. Physiol. 81, 381~22. yon Helversen, D. (1997). Acoustic communication and orientation in grasshoppers. In 'Orientation and Communication in Arthropods' (M. Lehrer, ed.), pp. 301-341. Birkhafiser, Basle. yon Helversen, D. (1998). Is the ramped shape of pulses in the song of grasshoppers adaptive for directional hearing'? Naturwissenschc~l?en 85, 186 188. von Helversen, D. and Rheinlaender, J. (1988). lnteraural intensity and time discrimination in an unrestrained grasshopper: a tentative behavioural approach. .L Comp. Physiol. A 162, 333 340. von Helversen, D. and von Helversen, O. (1975). Verhaltensgenetische Untersuchungen am akustischen Kommunikationsystem der Feldheuschrecken (Orthoptera, Acrididae) ]. De Gesang von Artbastarden. J. Comp. Physiol. 104, 273 299. von Helversen, D. and von Helversen, O. (1983). Species recognition and acoustic localization in acridid grasshoppers: a behavioural approach. In 'Neuroethology and Behavioural Physiology' (F. Huber and H. Markl, eds), pp. 95 107. Springer, Berlin. von Helversen, D. and von Helversen, O. (1995). Acoustic pattern recognition and orientation in orthopteran insects: parallel or serial processing'? J. Uomp. Physiol. A 177, 767 774. von Helversen, D. and yon Helversen, O. (1997). Recognition of sex in the acoustic communication of the grasshopper Chorthippus biguttulus (Orthoptera, Acrididae). J. Comp. Physiol. A 180, 373-386. von Helversen, D. and Wendler, G. (2000). Coupling of visual to auditory cues during phonotactic approach in the phaneropterine bushcricket Poecilimon qDfnis. J. Comp. Physiol. A 186, 729-736. von Helversen, D., Schul, J. and Kleindienst, H.-U. (2001). Male recognition mechanism for female responses implies a dilemma for their localisation in a phaneropterine bushcricket. J. Comp. Physiol. A 186, 1153 1158.

SOUND SIGNALLING IN ORTHOPTERA

277

yon Helversen, O. and yon Helversen, D. (1994). Forces driving coevolution of song and song recognition in grasshoppers. In 'Neural Basis of Behavioural Adaptations" (K. Schildberger and N. Elsner, eds), pp. 253 284. Gustav Fischer, Jena. Wadepuhl, M. (1983). Control of grasshopper singing behaviour by the brain: responses to electrical stimulation. Z. Tierpsychol. 63, 173 200. Wagner, W. E., Jr (1996). Convergent song preferences between female field crickets and acoustically orienting parasitoid flies. Behav. Ecol. 7, 279 285. Wagner, W. E., Jr and Hoback, W. W. (1999). Nutritional effects on male calling behaviour in the variable field cricket. Anita. Behav. 57, 89 95. Wagner, W. E., Jr and Reiser, M. G. (2000). The importance of calling song and courtship song in female mate choice in the variable field cricket. Anita. Behav. 59, 1219 1226. Wagner, W. E., Jr, Murray, A. M. and Cade, W. H. (1995). Phenotypic variation in the mating preferences of female field crickets, Gryllus imeger. Anita. Behav. 49, 1269 1281. Walker, T. J. (1969). Acoustic synchrony: two mechanisms in the snowy tree cricket. Science 166, 891 894. Walker, T. J, (1975). Effects of temperature on rates in poikilothermic nervous systems: evidence from the calling songs of meadow katydids (Orthoptera, Tettigoniidae: Orchelh~umO and reanalysis of published data. J. Comp. Physiol. A 101, 57 69. Walker, T. J. (1983a). Diel patterns of calling in nocturnal Orthoptera. In 'Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects' (D. T. Gwynne and G. K. Morris, eds), pp. 45 72. Westview Press. Boulder. Colorado. Walker. T. J. (1983b). Mating modes and female choice in short-tailed crickets (Amtrogryllus arboreus). In 'Orthopteran Mating Systems: Sexual Competition in a Diverse Group of Insects' (D. T. Gwynne and G. K. Morris, eds), pp. 240 267. Westview Press, Boulder, Colorado. Walker, T. J. (1993). Phonotaxis in female Ormia ochracea (Diptera. Tachinidae), a parasitoid of field crickets. J. bisect Behav. 6, 389410. Walker, T. J. and Figg, D. E. (1990). Song and acoustic burrow of the prairie mole cricket, Gryllotalpa mq/or (Orthoptera. Gryllidae)..I. Karts. Entomol. Soc. 63, 237 242. Walker, T. J. and Forrest, T. G. (1989). Mole cricket phonotaxis: effects of intensity of synthetic calling song (Orthoptera. Gryllotalpidae, Scopteriscu,s' acletus). Fla Entomol. 72, 655 659. Watson, A. H. D. and Hardt, M. (1996). Distribution of synapses on two local auditory interneurones, ON I and ON2 in the prothoracic ganglion of the cricket: relationships with GABA-immunoreactive neurones. Cell Tissue Res. 283, 231 246. Weddell, N. and Arak, A. (1989). The wartbiter spermatophore and its effect on female reproductive output. Behav. Ecol. Sociobiol. 24, 117 125. Weissman, D. B. (2001). Communication and reproductive behaviour in North American Jerusalem crickets (Stenopeh~uttus) (Orthoptera: Stenopelmatidae). In "The Biology of Wetas, King Crickets and Their Allies' (L. H. Field, ed.), pp. 351 375. CABI Publishing, Wallingford, Oxfordshire. Wohlers, D. W. and Huber, F. (1982). Processing of sound signals by six types of neurons in the prothoracic ganglion of the cricket Grvlhts campestris L. J. Comp. Physiol. A 146, 161 173. Wohlers, D. W. and Huber, F. (1985). Topographical organization of the auditory pathway within the prothoracic ganglion of the cricket, Gryllus campestris L. ('ell Tisxue Res. 239, 555 565. Wyttenbach, R. A. and Hoy, R. R. (1993). Demonstration of the precedence effect in an insect. J. Acoust. Soc. Am. 94, 777 784.

278

D.J. ROBINSON AND M. J. HALL

Yager, D. D. (1999). Structure, development, and evolution of insect auditory systems. Microsc. Res. Tech. 47, 380400. Young, D. and Ball, E. (1974). Structure and development of the auditory system in the prothoracic leg of the cricket GwIlus commodus (Walker). l. Adult structure. Z. Zel(/brsch. 147, 293 312. Zahavi, A. (1975). Mate selection: a selection for a handicap. J. Theor. Biol. 53, 205 214. Zhantiev, R. D. (197t). Frequency characteristics of tympanal organs in grasshoppers (Orthoptera: Tettigoniidae). Zool. Zh. 50, 507 514. Zhantiev, R. D. and Korsunovskaya, O. S. (1978). Morpho-functional organisation of tympanal organs in Tettigonia cantans (Orthoptera, Tettigoniidae). Zool. Zh. 57, 1012 1016. Zimmermann, U., Rheinlaender, J. and Robinson, D. (1989). Cues for male phonotaxis in the duetting bushcricket Leptophyes punctatissima. J. Comp. Physiol. A 164, 621 628. Zuk, M. and Kolluru, G. R. (1998). Exploitation of sexual signals by predators and parasitoids. Q. Rev. Biol. 73, 415 438. Zuk, M. and Simmons, L. W. (1997), Reproductive strategies of the crickets (Orthoptera: Gryllidae). In 'The Evolution of Mating Systems in Insects and Arachnids' (J, C. Choe and B. J. Crespi, eds), pp. 8%109. Cambridge University Press, Cambridge. Zuk, M., Simmons, L. W. and Cupp, L. (1993). Calling characteristics of parasitized and unparasitized populations of the field cricket Teleogryllus oceanicus. Behav. Ecol. Sociobiol. 33, 339-343. Zuk, M., Simmons, L. W. and Rotenberry, J. T. (1995). Acoustically orienting parasitoids in calling and silent males of the field cricket Teleogryllus oceanicus. Ecol. Entomol. 20, 380 383. Zuk, M., Rotenberry, J. T. and Simmons, L. W. (1998). Calling songs of field crickets (Teleogryllus oceanicus) with and without phonotactic parasitoid infection. Evolution 52, 166 171.

Insect Diuretic and Antidiuretic Hormones Geoffrey M. Coast, a l a n Orchard, b John E. Phillips c and David A. Schooley d aDepartment of Biology, Birkbeck (University of London), Malet Street, London WC1E 7HX, UK bUniversity of Toronto, Department of Zoology, 25 Harbor Street, Toronto, Ontario M5S 3G5, Canada CUniversity of British Columbia, Department of Zoology, Vancouver, British Columbia V6T 1Z4, Canada dDepartment of Biochemistry (330), University of Nevada, NV 89557-0014, USA 1 Introduction 280 2 Physiology of excretion 282 2.1 Introduction 282 2.2 Transport processes in Malpighian tubules that are controlled 285 2.3 Transport processes in the hindgut that are controlled 286 2.4 Fluid reabsorption across the cryptonephric complex 288 3 The functions of diuretic and antidiuretic hormones 289 3.1 Introduction 289 3.2 Postprandial diuresis 289 3.3 Post-eclosion diuresis 290 3.4 Excretion of excess metabolic water 290 3.5 Clearance of toxic wastes 291 3.6 Restricting metabolite loss 291 4 Isolation and structural characterisation of active factors 291 4.1 Introduction 291 4.2 Purification and chemical structure of neuropeptides that act on Malpighian tubules 293 4.3 Purification and chemical structure of neuropeptides that stimulate locust hindgut 312 5 Cellular actions 324 5.1 Introduction 324 5.2 Regulation of Malpighian tubule secretion 324 5.3 Regulation of hindgut activity 339 5.4 Fluid uptake from the cryptonephric complex 342 5.5 Structure/activity studies 343 6 Distribution 349 6.1 Introduction 349 6.2 Serotonin 349 6.3 AVP-like insect DH 352 6.4 CRF-related peptides 353 6.5 Kinins 358 A D V A N C E S IN INSECI PHYSIOLO(}~ VOL, 29 ISBN O-12-()24229-X

( ~4~~-(~I;; ! 2(H~2 ELsericr S ~ i c m c L i d

.Ill ~ie/ll~ O~ r~j~r°ducli(m in apll /i~rm ~c~¢,rr~,d

280

7

8

9 10

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

6.6 CAP2b/periviscerokinins 362 6.7 Calcitonin-like peptides 364 6.8 Ion Transport Peptide 364 6.9 Co-localisation 364 Physiological relevance 367 7.1 Circulating levels in relation to physiological status 367 7.2 Interfering with the natural titres of circulating factors 372 7.3 Degradation and inactivation 375 Integrated activities of diuretic and antidiuretic hormones 379 8.1 Introduction 379 8.2 Maintenance of haemolymph volume and compositon 379 8.3 Synergism between diuretic hormones 381 8.4 Co-ordinating Malpighian tubule and hindgut activities 383 The excretory system as a target for pest control strategies 384 Future directions 386 Acknowledgements 389 References 389 Addendum 389

Abstract The excretory system of insects comprises the Malpighian tubules and hindgut (ileum and rectum), the functions of which are controlled by diuretic and antidiuretic hormones. Strategies that have been used for the isolation, purification and characterisation of factors (putative hormones) that influence Malpighian tubule secretion and hindgut reabsorption in vitro are reviewed, along with information about their receptors, second messengers, effects on ion transport and structure/activity relationships. This is followed by a detailed description of the immunocytochemical localisation of identified diuretic and antidiuretic factors in neurosecretory cells and in neurohaemal organs from where they might be released into the haemolymph. Consideration is then given to the evidence for these factors functioning as circulating neurohormones to regulate Malpighian tubule secretion and hindgut reabsorption in vivo, and to how their activities are integrated for the maintenance of haemolymph volume and composition. Finally, mention is made of the potential use of diuretic and antidiuretic hormone agonists and/or antagonists as novel insecticides that act by disrupting the endocrine control of the excretory system. 1

Introduction

Despite their small size and hence large surface area to volume ratio insects have evolved to be the most abundant and diverse of all terrestrial animals. This has been made possible by a variety of behavioural, morphological and physiological adaptations that reduce water loss to a minimum, particularly in those species that occupy very dry habitats. To maintain their state of hydration, water gained from the diet, metabolism and, in some species, from the atmosphere, must equal water lost by evaporation, respiration and excretion.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

281

Indeed, for growth and reproduction the insect must be in net positive water balance. Since water availability varies greatly with developmental stage, activity and nutritional state, it follows that water loss must be regulated. Although evaporative and respiratory losses may vary with environmental factors and the state of hydration, the major site of regulation is the excretory system, and water loss via this route can change dramatically depending on the physiological status of the insect. Excretory water loss is determined by the rate at which fluid enters the hindgut from the midgut and Malpighian tubules (MTs), and the rate of reabsorption therein. These processes are under endocrine and possibly neural control, which allows the insect to regulate haemolymph volume and composition while permitting nitrogenous waste, toxic substances, and excess ions and/or water to be voided. The endocrine factors are referred to as diuretic hormones (DHs) and antidiuretic hormones (ADHs), although this may not accurately describe their actions (see below). Generally, DHs stimulate primary urine secretion by MTs, whereas ADHs increase fluid reabsorption from the hindgut, but there are exceptions to this. The endocrine control of the insect excretory system was last comprehensively reviewed by Phillips (1983). At that time, insects were known to have DHs and ADHs, but none of the active lectors had been characterised, although they were thought to be peptides produced by neurosecretory cells (NSCs) in the central nervous system (CNS) and released into the circulation from neurohaemal sites such as the storage (neural) lobe of the corpora cardiaca (CC). The first definitive DH was characterised by Kataoka el al. (1989), and a further 10 years elapsed before the complete sequence of an ADH was known (Meredith el ell., 1996). There have been a number of separate reviews of the hormonal control of MT and hindgut function since 1983 (Audsley et al., 1994; Phillips and Audsley, 1995; Coast, 1996; Coast, 1998b; Phillips et al., 1998b; O'Donnell and Spring, 2000), but none of these attempted an integrated account of the regulation of the entire excretory process. Many technical advances since 1983 have contributed to the characterisation of an ever-increasing number of peptides with diuretic and/or antidiuretic activity. These include improved methods for the isolation, purification and characterisation of neuropeptides, most notably with improvements in highperformance liquid chromatography (HPLC), automated peptide sequencing and mass spectrometry (MS). There has also been the development of routine molecular biology protocols for mRNA isolation, amplification and sequencing, and for gene expression. Of great long-term significance is the sequencing of the entire genome of the fruit fly Drosophila melanogaster (Adams et al., 2000), with a similar project nearing completion for the malarial mosquito Anopheles gambiae. These genomic databases can be used to search for genes encoding neuropeptides and their receptors, and much work over the next few years will be devoted towards determining their functions. In light of developments since 1983, this seems an appropriate time to review current understanding of the endocrine control of the insect excretory

282

G. M, COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

system. We begin with a brief overview of the major transport processes in Malpighian tubules (MTs) and hindgut, before describing identified DHs and ADHs, their receptors, second messengers and modes of action. This is followed by a consideration of their immunocytochemical localisation and actions in the intact insect. We conclude by reviewing progress in developing novel insect pest control compounds aimed at disrupting the endocrine control of excretion.

2 2. l

Physiology of excretion INTRODUCTION

Figure 1 summarises the excretory process in insects. Malpighian tubules secrete primary urine that is isosmotic to haemolymph and rich in KC1 and/ or NaCl. In many respects, it resembles a filtrate of haemolymph in that it contains all low molecular weight solutes (_<5000 Da), but minus large proteins. This non-selective process may be augmented by active secretion of urate (O'Donnell et al., 1983), organic anions (Maddrell et al., 1974; Linton and O'Donnell, 2000), various plant alkaloids (Maddrell and Gardiner, 1976b), and the cardiac glycoside ouabain (Meredith et al., 1984). Primary urine generally flows into the gut at the midgut hindgut junction where it mixes with fluid from the midgut. Normally, urine flow is directed posteriorly into the hindgut, but some may also move anteriorly into the midgut and be recycled to the haemolymph (Dow, 1981; Nicolson, 1991 ). Within the hindgut, the selective and controlled reabsorption of water, ions and essential metabolites means that the final urine generally differs markedly in volume and composition from MT fluid. The initial non-selective process of primary urine formation ensures that toxic compounds, even ones not previously encountered, are first cleared from the circulation and then concentrated in the excreta by the reabsorption of essential solutes and water in the hindgut. Malpighian tubules are not innervated and must therefore be controlled by factors (hormones) released into the circulation. Making use of an in vitro assay to measure fluid secretion by isolated MTs (the Ramsay assay; Ramsay, 1954), Maddrell (1963) showed that the extensive diuresis that follows a blood meal in the triatomid bug Rhodnius prolixus is initiated by release of a DH from the fused mesothoracic ganglion mass (MTGM), which stimulates primary urine production by up to 1000-fold. The Ramsay assay has since been used to demonstrate the presence of diuretic factors in extracts of the CNS and/ or retrocerebral complex (CC and corpora allata, CA) of many insects. Many of these factors (neuropeptides and serotonin) have been chemically characterised (reviewed by Coast, 1996, 1998b), which has allowed detailed studies of their mode of action, cellular localisation and, in a few cases, their haemolymph concentration in relation to physiological status. The majority of these studies have looked only at effects on fluid secretion, and few have attempted

INSECTDIURETICANDANTIDIURETICHORMONES midgutcontents

MIDGUT

]

L ~. / MALPIGHIAN TUBULES

organicacids,urate,

plan°tu~ :'°ids' iI-K+j'Na+

/iso.o,,c\ /

283

wer water(isosmoticabsorbate)

~

\isosmotic/

co-9 / ( ~

-~

essentialmetabolites

RECTUM -~C~'ater (hyp°-°r hyper°sm°ticabs°rbate) / ~ K+/Na+

hypo-orhyperosmotie excretacontaining toxicwastes FIG. l An overview of the excretory process in insects. Malpighian tubule secretion is driven by the active transport of KCI and/or NaCI into the lumen, with water tkfllowing by osmosis. Other ions and low molecular weight solutes enter the lumen by pzlssive diffusion, which may be augmented by active transport. Primary urine enters the gut at the midgut-hindgut junction and for the most part is directed posteriorly into the ileum, where it is modified by isosmotic fluid reabsorption driven by an apical membrane electrogenic CI-- pump. The reduced volume of fluid entering the rectum is extensively modified by reabsorption of essential metabolites and fluid, which may be hypo- or hyperosmotic to the luminal contents, Again the main driving force is an apical electrogenic CI pump. Toxic wastes are retained in the hindgut lumen and are voided in the excreta, which can be strongly hypo- or hyperosmotic to haemolymph.

284

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

to measure effects on the composition of the secreted fluid, which can change profoundly after stimulation by DHs (Williams and Beyenbach, 1983; Williams et al., 1983). Additionally, a number of diuretic and myotropic peptides have been shown to increase the frequency and amplitude of writhing movements by locust MTs, which could assist the flow of urine in the lumen and reduce unstirred layers at the basolateral surface (Coast, 1998a). The presence of antidiuretic factors (ADFs) that reduce MT secretion, first proposed by Spring et al. (1988), has recently been confirmed (Quinlan et al., 1997: Eigenheer et al.. 2002) and adds further to the complex control of primary urine formation. Many of the factors (putative hormones) that act on MTs have been characterised, but relatively little is known of their role as physiological regulators of tubule secretion in vivo. Ionic and osmotic regulation in most insects studied to date depends ultimately on selective, active, and controlled reabsorption of solutes and water from the primary urine in anterior (ileum) and posterior (rectum) hindgut segments. This was evident from early physiological observations on whole insects that revealed very large changes in the volume and solute composition of the final excreta in response to severe fluctuations in external conditions, thus suggesting neuronal or endocrine control of hindgut reabsorption. The most studied insect, the desert locust (Schi~'tocerca gregaria), when dehydrated and starved produces faecal pellets only infrequently and these are very dry (hyperosmotic). In contrast, when feeding on succulent plants, they daily eliminate their own body weight of fluid containing very low NaC1 (reviewed by Phillips, 1981). Accordingly, the neuroendocrine control of insect hindgut function received early attention (reviewed by Phillips, 1981, 1983). There are several early reports of putative diuretic and antidiuretic factors in tissue homogenates that inhibit or stimulate respectively the rate of fluid reabsorption (Jr) by rectal sacs of several larger insects in vitro (reviewed by Phillips, 1983; see also Proux et al., 1984, 1985). These early studies generally used uncharacterised in vitro hindgut preparations in transient states and with inadequate salines or oxygenation. None of these studies considered effects on specific solute transport processes. More recent work on hindguts has indicated only antidiuretic factors in insect neuroendocrine tissues, possibly because extracts have only been assayed on unstimulated rectal preparations. These studies consistently suggested that the brain/CC/CA contain stimulants of fluid absorption, and hence presumably ion transport. Stimulants of ileal and rectal Jv were also observed in the fourth to seventh ventral ganglia (VG) of desert locusts (Lechleitner et al., 1989a,b; Lechleitner and Phillips, 1989: Audsley and Phillips, 1990) and terminal abdominal ganglia of cockroaches (Goldbard et al., 1970). More recent identification and isolation of specific neuropeptides acting on specific hindgut solute transport processes have been restricted largely to locusts (reviewed by Phillips and Audsley, 1995; Phillips et al., 1998a,b, 2001). The functioning of these neuropeptides as natural hormones in vivo remains, however, to be firmly established.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

2.2

285

TRANSPORT PROCESSES IN M A L P I G l l I A N TUBULES T H A T ARE CONTROLLED

Malpighian tubule ion transport mechanisms have been reviewed by Nicolson (1993) and Beyenbach (1995). Fluid secretion is driven by the active transport of K + and/or Na + into the lumen accompanied by CI , with water following osmotically. Principal ceils provide the route for cation transport into the lumen, and the close association of mitochondria with their apical microvilli and basolateral infoldings is typical of cells engaged in active transport. Whether K + or Na + is preferentially transported depends on their uptake across the basolateral membrane and their affinity for the 'cation transporter" m the apical membrane. The latter is now known to comprise a parallel arrangement of amiloride-sensitive alkali cation/H + antiports and a Bafilomycin Awsensitive proton pump (vacuolar ATPase; V-ATPase) (Wieczorek el al., 1991). The V-ATPase generates an electrochemical gradient for protons to enter the cell from the lumen in exchange for K + and/or Na + (Bertram el al., 1991; Maddrell and O'Donnell, 1992: Weltens el al., 1992). The alkali cation/H + antiports of R. prolixus have a preference for Na -~ over K + (Maddrell, 1978; Maddrell and O'Donnell, 1993), whereas there appears to be no such preference in the house cricket, A c h e m domesticus (Coast, 2001b). Alkali cations enter principal cells through ion channels in the basolateral membrane, if the electrochemical gradient permits, or by secondary (e.g. cation/C1- cotransport) and primary (Na+/K + ATPase) active transport (see Fig. 2). In general, the electrochemical gradient over the epithelium favours CI diffusion into the lumen through a shunt pathway. The C1-selective shunt described by Pannabecker et al. (1993) in MTs of the yellowfever mosquito, Aedes aegypti, appears to lie outside of the principal cells, and may be paracellular or through a second cell type, the stellate cells, which form thin (3 51,m deep) windows between the haemolymph and tubule lumen. The osmotic permeability of the distal segment of R. prolixus tubules is high enough for classical osmosis to account for the observed rate of secretion even though the osmotic gradient is only a few milli-osmolar (O'Donnell et al., 1982). Importantly, the osmotic permeability is not increased by serotonin, although it stimulates rapid fluid secretion (Maddrell et al., 1969: Maddrell el al., 1971 ). Diuretic hormones therefore act by stimulating ion transport, but achieve this in different ways (see section 5.2). The proximal segment of R. prolixus MTs actively reabsorbs KCI from the lumen via an apical membrane H+/K+-ATPase similar to that found in the gastric mucosa (Haley et al., 1997: Haley and O'Donnell, 1997). The osmotic permeability of this segment is about 30% lower than the distal tubule, and is reduced further by serotonin (O'Donnell et al., 1982). The absorbate is therefore hyperosmotic to the luminal contents and this region functions as a diluting segment (Maddrcll and Phillips, 1975).

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

286

LUMEN

........ ~, Na* H* K*

Cl-

Cl-

PC

......

K*

cr

K a"

CI'K* Na* K* CI"

Na + HAEMOLYMPH FIG. 2 A generalised model for Malpighian tubule 1on transport. The principal cell (PC) V-ATPase generates a proton gradient across the apical membrane that drives K+/Na + transport into the lumen via alkali cation/proton antiports. Cations are taken up across the basolateral membrane via ion channels and by secondary (cation/ CI cotransport) and/or primary (Na+/K + ATPase) active transport. Chloride moves into the lumen through a "shunt' pathway that may be paracellular, or through a series arrangement of ion channels in the apical and basolateral membranes of principal cells or stellate cells (SC) when these are present. Arrows with circles indicate primary (black circles) or secondary (grey circles) active transport, while arrows through cylinders represent membrane ion channels.

2.3

TRANSPORT PROCESSES IN THE H I N D G U T T H A T ARE CONTROLLED

The hindgut epithelial mechanisms responsible for selective reabsorption and their cellular location (epithelial models) have been well characterised only in the desert locust (reviewed by Phillips et ell., 1986, 1988; Phillips and Audsley, 1995), including more recently the key role of hindgut segments in whole-body acid-base balance, ammoniagenesis, and nitrogen excretion (Harrison and Kennedy, 1994; Phillips et al., 1994). In both hindgut segments of the desert locust, the dominant transepithelial active transport mechanism is an unusual (compared with vertebrates) electrogenic C1 p u m p located in the apical membrane (reviewed by Phillips el al., 1996). When the transepithelial potential (TEP) is clamped at zero millivolts (mV), the change in applied current (i.e.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

287

short-circuit current, Isc) required to maintain the clamp is a direct continuous measure of the increase in CI- transport rate after stimulation by cyclic AMP or neuroendocrine tissue extracts. Chloride exits the rectal cells passively via a basolateral conductance pathway (putative channel). Potassium, the major cation absorbed, follows CI- passively by electrical coupling via cation channels with different properties in apical and basolateral membranes. The levels o f N a + in the primary urine and hence in fluid entering the locust hindgut lumen are quite low (20 mM), and active reabsorption of this cation is therefore quantitatively less important. Sodium enters rectal cells passively by several mechanisms (a conductance pathway; in exchange for NH~ and H+; and by cotransport with glycine: Black et al., 1987) down a very large electrochemical potential gradient at the apical border, and is actively removed from cells basolaterally by a Na+/K+-ATPase pump. Stimulation of fluid transport in the absence of, or against, osmotic concentration differences across both hindgut segments is dependent on salt transport, especially Cl (Proux eta/., 1984; Lechleitner el al., 1989a,b). An electrogenic H + pump (V-type ATPase) in the apical membrane causes equal rates of acid secretion into the lumen in both hindgut segments, and this is associated with passive exit of base equivalents ( O H - and HCO;-) to the haemocoel side (Thomson and Phillips, 1992; see also Phillips el al., 1994). While locust ilea and recta share many common epithelial transfer mechanisms, some important differences have been observed. Fluid transport (J,) is always close to isosmotic in the ileum and can be increased four-fold by stimulants, as compared to a less than two-fold increase for rectal J,.. The rectum can extract a hypo-osmotic absorbate to concentrate the lumen content to final osmotic concentrations several times that of the haemolymph: that is, this segment can create strongly hyperosmotic excreta. This is thought to be achieved by solute recycling at elaborate lateral intercellular channels that are not observed in the ileum (reviewed by Phillips et al., 1986; lrvine et al., 1988). Moreover, a high-capacity proline pump (Meredith and Phillips, 1988) in the apical membrane of the rectum (but not the ileum) can drive considerable additional water extraction even in the absence of luminal Na +, K ~ and CI (Lechleitner and Phillips, 1989). Proline recovered from the rectal lumen also provides the principal substrate tk~r cellular respiration and ammoniagenesis leading to apical NH + secretion in exchange for luminal Na + (Chamberlin and Phillips, 1983; Thomson el al., 1988). Water and CO~ from cellular respiration provide the H + and HCO 3 for acid-base transport and pH regulation, which is aided by ammonia production that traps H + to form NH~. In contrast, several neutral amino acids recovered apically from the primary urine serve the same functions as substrates for cellular respiration, the ammoniagenesis and acid-base equivalents in the ileum (Phillips et al., 1994). The ileum is the predominant site of Na + reabsorption and NH~ secretion in the locust excretory system, and is the only site where there is some evidence for

288

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

endocrine control of these processes (Irvine et al., 1988; Phillips et al., 1994). In summary, the role of the locust ileum and rectum is functionally analogous to that of the proximal and more distal tubular segments, respectively, of the mammalian kidney. 2.4

FLUID REABSORPTION ACROSS THE CRYPTONEPHRIC COMPLEX

In the cryptonephric complex of Coleoptera and larval Lepidoptera, the rectum is closely associated with the distal part of the MTs, which lie within a perinephric space that is separated from the haemolymph by an impermeable perinephric sheath. The cryptonephric tubules of T. molitor contact the perinephric sheath at regions called 'boursouflures' or 'bulges'. Here the perinephric sheath is greatly reduced and forms a ~blister' over an underlying leptophragma cell (Ramsay, 1964). Potassium chloride is transported from the haemolymph into the cryptonephric tubules across the leptophragma cells without an accompanying movement of water, thereby establishing a high osmotic concentration (as much as 6 . 8 0 s m in T. molitor larvae, close to water-saturation for KCI) in the tubule lumen (O'Donnell and Machin, 1991). The perinephric space also has a high osmolarity consisting largely of non-ionic solutes (melting point depression up to - 1 0 C ) with an N a + - K + ratio similar to that of haemolymph (Ramsay, 1964), unlike the distal parts of the cryptonephric tubules, which absorb K ~ to the near exclusion of Na + (O'Donnell and Machin, 1991). The high osmotic concentration in the tubule accounts for the osmotic withdrawal of water from first the perinephric space and ultimately the rectal lumen. This exceptional osmotic gradient may be necessary to drive water transport across at least two membranes and can account for the uptake of water vapour from air of >89% relative humidity (RH) (Coutchi6 and Machin, 1984), The cryptonephric complex of lepidopteran larvae differs from that of T. molitor in that it is restricted to three longitudinal bands running the length of the rectum, corresponding with the three pairs of MTs (Ramsay, 1976). Between each band, a patch of 'normal' epithelium separates the luminal contents from the haemocoel. Additionally, boursouflures appear to be absent in the lepidopteran cryptonephric complex, which nevertheless is assumed to function in a manner similar to that of T. molitor, although high osmotic concentrations are not established. About 90% of the fluid entering the rectum of day 1 fifth-instar larvae is reabsorbed (Reynolds and Bellward, 1989), a quarter of which enters the cryptonephric tubules and is recycled to the midgut via the straight ascending and descending tubule segments (Moffett, 1994). At the same time, K + in the tubule lumen is returned to the midgut in exchange for Na + (Moffett, 1994). The remaining fluid is reabsorbed across the 'normal' epithelium, which is composed of cells with extensive infoldings of the apical and basolateral membranes and associated mitochondria (Reynolds and Bellward, 1989).

INSECT DIURETIC AND ANTIDIURETIC HORMONES

3

289

The functions of diuretic and antidiuretic hormones

3.1

INTRODUCTION

Mammalian physiologists use the terms diuresis and antidiuresis to describe periods of increased and decreased urine output, respectively. Terrestrial insects normally need to conserve water, and antidiuresis is therefore expected to be the norm, interrupted by periods of diuresis associated with increased dietary intake of water, increased metabolic water production or a need to reduce the haemolymph volume prior to flight.

3.2

POSTPRANDIAL DIURESIS

Haematophagous insects consume a blood meal up to l0 times their unfed body weight, which severely restricts their movement and imposes a considerable salt and water load, because vertebrate plasma is rich in NaC1 and is hypoosmotic to haemolymph. The excess salt and water are voided in a postprandial diuresis (see Fig. 3), during which MT secretion may increase up to 1000-fold (Maddrell, 1963) in response to DHs released into the circulation. The stimulus for DH release is distension of the anterior abdomen, which is detected by stretch receptors in the tergal sternal muscles (Maddrell, 1964b). The response

100-

A

B 100 -

Peak

Post-peak

Late

t= 75-

E 50-

~.. 50-

0

O o c:

°

0

30

60 T i m e {min)

90

120

25-

0iJ

I i

25

5O

75

Time (rain)

FIG. 3 A. Water loss from a blood-lEd female A. aegypti recorded using a flowthrough humidity meter. Peaks of increased water loss (horizontal arrow) that are superimposed upon cuticular and respiratory losses (vertical arrow) represent urine output. During the peak phase of diuresis, urine drops or almost constant volume are voided every 30s. The initial peak marked with an asterisk is when the insect is introduced into the chamber. B Calculated rates of urine output over the first 70 rain of the postprandial diuresis. The divisions into peak-phase, post-peak phase and late phase are as used by Williams et al. (1983). G. M. Coast unpublished data.

290

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

is extremely rapid and diuretic activity is detectable in the circulation within 15s of the onset of feeding (Maddrell and Gardiner, 1976a). Between blood meals there is no circulating DH and MTs secrete very slowly. Excretory water loss may be further reduced by fluid reabsorption from the hindgut under the control of ADHs. A postprandial diuresis is also evident in xeric insects, and faecal water content as a percent of dry weight increases five-fold within 1 h of feeding S. gregaria on fresh grass (Norris, 1961). The duration of the meal is controlled by input from stretch receptors in the wall of the foregut (Bernays and Simpson, 1982), and they most likely provide the stimulus for DH release, because excretory water loss is increased after the foregut is artificially distended with agar (Bernays and Chapman, 1972).

3.3

POST-ECLOSION DIURESIS

Newly emerged butterflies and moths have a large haemolymph volume to facilitate wing expansion. Once the wings are expanded, however, the haemolymph volume is reduced by c.50% to lessen the weight of the insect before the first flight. |n the cabbage-white butterfly, Pieris brassicae, this is due largely to the excretion of a considerable volume of urine in the 3 h after ecdysis, and MT secretion can reach 100 nL rain 1 during this post-eclosion diuresis (Nicolson, 1976). Other flying insects, such as A. ae~o'pti (Gillett, 1982), may also use a post-eclosion diuresis to reduce the haemolymph volume after emergence. Since the post-eclosion diuresis forms part of a sequence of events linked to eclosion behaviour, it will be interesting to know whether hormones that orchestrate this behaviour, notably eclosion triggering hormone, eclosion hormone and crustacean cardioactive peptide (CCAP), stimulate DH release.

3.4

EXCRETION OF EXCESS METABOLIC WATER

Male bumblebees fed artificial nectar (50% sucrose) must excrete daily an amount of water equal to their entire body water content (Bertsch, 1984), which comes from their diet and from metabolic water production. Much of it is lost while they are foraging, and flying bees are observed to urinate, a process described as 'squirting'. A DH may therefore be released during flight, although there is no direct evidence for this. The requirement for flying insects to void excess metabolic water probably extends beyond nectar-feeders. In a study of ventilation in flying locusts, Weis-Fogh (1967) demonstrated that a swarm of insects flying at an altitude of 3 km at 23c'C are in positive water balance if the ambient relative humidity (RH) exceeds 35%, and may therefore need to excrete excess water.

INSECT DIURETIC AND ANTIDIURETIC HORMONES 3.5

291

CLEARANCE OF TOXIC WASTES

Nicolson and Hanrahan (1986) found a DH in CC extracts of the tenebrionid beetle, Onymacris plana, which lives in the Namib Desert, one of the driest environments on Earth. The DH increases MT secretion 60-fold, but faecal water loss is minimal, because urine mostly flows forward into the midgut and is returned to the haemolymph (Nicolson, 1991). Hence, the beetle 'DH' does not stimulate diuresis per se, but might instead have a clearance function (Nicolson, 1991), which is enhanced at high urine flow rates (Maddrell, 1981). 3.6

RESTRICTING METABOLITE LOSS

Paradoxically, very high rates of tubule secretion can also be a means of reducing the loss of haemolymph metabolites during diuresis (Maddrell and Gardiner, 1980). After a blood meal, R. prolix,~s' eliminates c,140/*1 of urine containing trace amounts of amino acids, although the haemolymph concentration of fl-ee amino acids is c.35 raM. A major contributing factor is the low permeability of the distal tubule to amino acids, which makes it advantageous for fluid secretion to be as high as possible (Maddrell and Gardiner, 1980). According to these authors, if R. prolixz~s MTs secrete at 30 nL rain -I , which is greater than the observed maximum for most insects, about 30% of haemolymph amino acids would be lost while eliminating 140/~1 of urine, but this falls to <5% when fluid secretion is 300nLmin -I, which can be achieved in ~,ivo.

4 4. ]

Isolation and structural characterisation of active factors IN'FRODUCTION

With the notable exception of serotonin, the factors identified as having diuretic and/or antidiuretic activity are all neuropeptides and are listed in Table 1. Here, and elsewhere in this review, we have followed the species abbreviation used by the National Center for Biotechnology Information (NCBl)/Swiss-Prot, in which the first three letters of the genus is added to the first two letters of the species name. Thus, a 31-residue DH from Drosophila melanogaster is called Drome-DH31. This avoids anomalies with the peptide nomenclature scheme suggested by Raina and G/ide (1988), and we strongly recommend that all newly discovered neuropeptides be given a similar five-letter prefix. Hewes and Taghert (2001) assembled a list of 22 D. melanogasler neuropeptide genes, but consider this list to be incomplete. Assuming a 1:1 ratio of neuropeptide genes to peptide G protein-coupled receptor (GPCR) genes @.45) in the D. melanogaster genome, there are c.20 orphan peptide receptors, but there may be less, because some genes encode multiple peptides (Hewes and

292

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D, A. SCHOOLE'

e-, e~

-.g

=

.~

e,o

.~.~.~.~

?

.< .~

.~.~.~.~..~

.....

~.~

.,,-j .'m

q=

<

q::2 ¢,,.

L;

g, e¢ o c~

,,m

>-

>.>~>~>->-

>.

.g>, eO

;4

.D>

z

E "r-'

°

=Te. q:

P~ :>,

.z:;

,j

z "E

o

o

o

.c,.

e-

.~

~E .<

0 m

...-~

d~<
~

INSECT DIURETIC AND ANTIDIURETIC HORMONES

293

Taghert, 2001). Thus, Vanden Broeck (2001) lists 50 fruit-fly neuropeptides that were identified by BLAST T M (Basic Local Alignment Research Tool) searches of D. melanogaster genome databases for putative peptide precursor genes using as a query the sequences of known neuropeptide families. Insect neuropeptides can be grouped into c.20 families, some of which are very extensive, such as the adipokinetic/red pigment concentrating hormone (AKH) family, whereas others contain highly conserved peptides, notably proctolin and CCAP. Other neuropeptides are present as multiple isoforlns within a single species; for example there are 13 chemically distinct allatostatins encoded on a single precursor protein in Diploptera punctata (Donly el al., 1993). Significantly, some insect neuropeptides, such as the corticotropin releasing-factor (CRF)-related diuretic peptides (see section 4.2.2) are similar to those of vertebrates, indicating a long evolutionary history as these two groups diverged c.600 million years ago. In the following section, we describe the isolation, purification and structure of neuropeptides having diuretic and/or antidiuretic activity. It is worth emphasising here, although it will later become apparent, that diuretic and antidiuretic peptides have been identified from very few species. This small subset cannot be representative of the enormous diversity of insects, and it is likely there are novel peptides with diuretic or antidiuretic activity still to be identified. 4.2

P U R I F I C A T I O N A N D C H E M I C A L S T R U C T U R E OF N E U R O P E P T I D E S T t t A T ACT ON M A L P I G t n A N TUBULES

Of central importance to a successful isolation scheme for any bioactive peptide is an assay technique capable of detecting small amounts of the desired peptide hormone with high sensitivity and specificity. The very large number of identified myotropic peptides attests to the ease and rapidity of bioassays for such factors. For DHs, isolations have been guided by functional bioassays based on the ability to elevate fluid secretion in vivo, assays to enhance secretion by MTs in vitro, assays based on ability to elevate second messengers (usually cyclic AMP) produced by MTs in vitro, and immunoassays based on antisera to homologous peptides. To date, only one successful identification of diuretic factors has utilised fluid secretion by isolated MTs (the Ramsay assay) for monitoring purification, the isolation of the helicokinins (Blackburn et al., 1995a; see section 4.2.4). Attempted isolation of a diuretic peptide from A. domesticus using this assay led to unsuccessful results, perhaps due to the lack of selectivity of the Ramsay assay compared with the more specific second messenger assays (Coast el al., 1990b). The explosion in knowledge of insect peptide hormone structures began in the mid 1980s, when use of HPLC techniques for neuropeptides became common. A number of modes of chromatography have been utilised for isolation of both diuretic and antidiuretic factors from insects. The most

294

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

common is reversed-phase liquid chromatography, more conveniently abbreviated RPLC than RP-HPLC. ~Normal phase' LC refers to the use of a relatively polar stationary phase eluted with solvents of increasing polarity. For peptides, the 'weakest' normal phase eluent usable without causing precipitation of bioactive materials is about 80% acetonitrile; the column is eluted with a gradient of gradually increasing water content, i.e. the opposite of RPLC. Ion exchange LC has proven very useful for purification of many peptides, in part because of its higher sample capacity than RPLC columns, but also because the mechanism of separation is totally different from either reversed- or normal-phase LC. Another separation technique, hydrophilic interaction chromatography, was extremely selective for purification of one of the two CRF-like DH isolated from the yellow mealworm Tenebrio molitor (Furuya et al., 1995). This technique seems highly similar in mode of separation and selectivity with the normal-phase steps used in the isolation of Locmi-DH, Peram-DP, and Achdo-DP by Kay et al. (1991a,b, 1992). Prior to purification by RPLC (or other mode) of separation, prepurification steps are necessary to remove fats and other impurities in extracts that would otherwise bind irreversibly to the expensive columns used. Diuretic factors have been extracted from insect tissues with 1% acetic acid containing 20 mM mineral acid plus protease inhibitors (Schooley et al., 1987; Kataoka et al., 1989; F'uruya et al., 1995, 1998), or with 87-90% methanol containing 1-5% acetic acid, with the balance being water (Kay et al., 1991a,b, 1992; Clottens et al., 1994). Use of a totally non-aqueous extraction solvent consisting of 90% methanol, 9.9% acetic acid and 0.1% 2-(methylthio)ethanol proved to give far 'cleaner' extracts than the aqueous acid extraction mentioned above; this was used successfully in isolating two DH from Hyles lineata (Furuya et al., 2000a), two from Diploptera punctata (Furuya et al., 2000b), and one from Zootermopsis nevadensis (Baldwin et al., 2001). Most likely the high methanol solvents yield cleaner extracts than the aqueous solvents due to the poor solubility of proteins in high-methanol concentrations. In most separations, crude extracts have been pre-purified by passage through reversed-phase solid-phase extraction cartridges, either commercial ones such as the Waters Sep-Pak Cis, or home-made cartridges prepared with bulk wide-pore reversedphase supports such as Vydac C4 or Cts. Schooley et al. (1987) demonstrated a 25% loss of immunoreactive "AVP-like insect diuretic hormone (AVP-IDH)' on passage of extracts through Sep-Pak Cjs cartridges; losses would be higher for larger peptides. Other techniques used for pre-RPLC purification include use of gravity flow ion exchange supports such as SP-Sephadex (Kataoka et al., 1989). Acetone precipitation has been used as a technique for pre-purifying extracts in a number of isolations of CRF-like DH (Kay et al., 1991a,b, 1992; Clottens et al., 1994). Recovery of biological activity was not reported. The difficulty of peptide purification depends crucially on the abundance of the peptide and the tissue source used: surgical removal of neurohaemal tissue can greatly reduce the number of chemical purification steps required. For

INSECT DIURETIC AND ANTIDIURETIC HORMONES

295

isolation of Manse-DPIl, Blackburn et al. (1991) painstakingly dissected clusters of medial neurosecretory cells from the protocerebrum of M. sexta or 130 CC; extraction of either tissue gave Manse-DPIl in an essentially pure state in a single RPLC purification step. The amount of peptide was, however, minute, making the sequence determination difficult. Most peptide identifications have utilised whole heads of insects, which require far more purification steps. 4.2,1

Arginine vasopressin-like insect diuretic hormone

Proux el al. (1982) demonstrated that a factor in the suboesophageal ganglia of L. m~gratoria stimulated the clearance of injected amaranth from the haemolymph of locusts #7 vivo, a measure of primary urine secretion (Mordue, 1969), and regarded this material as identical to a factor that cross-reacts with arginine vasopressin (AVP). A total of 51 000 dissected SOG (suboesophageal ganglia) and T G (thoracic ganglia) were extracted with aqueous acid and prcpurified using Sep-Pak Cts cartridges. All purification steps were monitored using a radioimmunoassay (RIA) for AVP. On the first RPLC separation, two fractions were observed which were immunoreactive. Only the slower eluting of these two fractions was able to stimulate amaranth clearance in L. migratoria. The inactive, faster eluting factor (called F1) was isolated to homogeneity in two additional RPLC steps. In the second RPLC step, use of heptafluorobutyric acid (HFBA) instead of the usual trifluoroacetic acid (TFA) as ion pairing reagent gave an extremely selective purification: this was not unanticipated because the C-terminus of AVP, Cys-Pro-Arg-Gly-NH> is strongly basic and should therefore form a strong ion pair. The active factor (F2, slower eluting) required four additional steps for isolation to homogeneity. Drying down samples in the presence of bovine serum albumin (BSA) to remove acetonitrile from the mobile phase was found, using RIA to quantify recovery, to lead to appreciable losses of sample. Esch el a/. (1983) advocate never drying down samples of bioactive neuropeptides during RPLC, but instead diluting fractions three to four-fold with water to decrease the acetonitrile concentration so that the sample will absorb to the column in the next RPLC step. Modification of the instrument to allow this approach was not possible with the instrument used, so a vacuum centrifuge was used to partially remove acetonitrile between purifications. However, this still led to rather poor recovery of F2, so that no estimates were made for the quantity of F1 and F2 per locust head (Schooley el al., 1987). Amino acid analysis of the two peptides gave identical amino acid compositions, suggesting they were homologues of AVP containing Leu and lie, but no Tyr or Phe. Each sample was reduced and carboxymethylated (RCM) prior to sequence analysis. Purification of the RCM peptides showed them to have identical retention on RPLC: sequence analysis of each showed that both peptides had the same primary sequence: C L I T N C P R G . This peptide was

296

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

synthesized in both the C-terminal amidated and acid forms, which were reduced and carboxymethylated. The RCM-amide was found to co-elute with RCM-F1, while the RCM acid form had different retention properties, establishing the sequence as CLITNCPRG-NH2. Some of the remaining native F1 and F2 were analysed by size exclusion LC with a calibrated column; the amounts of peptide injected were too small to detect by UV, but were easily detectable in numerous timed fractions by RIA. This analysis showed F2 to have about twice the M,- of F1. Consequently, difficult specific syntheses of F2 were conducted as both parallel and antiparallel dimers. These were analysed by RPLC and the antiparallel dimer (the faster eluting of the two) was found to co-elute with F2. Bioassays using semi-isolated locust MTs including the ampullae and a portion of gut indicated that only the antiparallel dimer had biological activity. This established the 'AVP-like insect diuretic hormone' to be an antiparallel dimer of primary sequence C L I T N C P R G - N H 2 (Proux el al., 1987).

4.2.2

CRF-related neuropeptides

4.2.2.1 Isolation and puro?cation. M. sexta DH (Manse-DH) was isolated from 420 g of trimmed pharate adult heads (10 000 animals). The isolation was monitored with a slow, difficult assay in which aliquots of purification fractions were injected into decapitated, newly eclosed Pieris rapae butterflies (Kataoka et al., 1989). This species undergoes a post-eclosion diuresis (see section 3.3), which is blocked by ligaturing the neck of pharate adults, but is restored on injection of material with diuretic activity. A full 5nmol of peptide was obtained from 10000 animals and was easily sequenced, corresponding to about 0.5 pmol/head. The purification scheme utilised four RPLC purification steps and one ion-exchange LC step. The 41-residue peptide sequence has a 40% identity to sauvagine, a frog skin peptide, which at that time was thought to be the amphibian homologue of CRF. Interestingly, synthesis of the Cterminal free acid form of Manse-DH (as opposed to the amidated form) gave a product whose biological potency was reduced 1000-fold compared with Manse-DH in the P. rapae (and other) bioassays (see section 5.5.2). In 1991, Blackburn et al. isolated a second DH from M. sexta, which had escaped detection by Kataoka et al. (1989), using an in vivo bioassay with adult M. sexta. This peptide has only 30 residues and seems to represent a paralogue of Manse-DH. When aligned as shown in Fig. 4, only seven residues are identical between these peptides. The only major difference in biological potency between these two DH reported to date is that Manse-DH can promote fluid reabsorption from the everted rectal sac of M. sexta larvae, whereas ManseDPII cannot (Audsley et al., 1995). A further 11 CRF-related DHs have been isolated and sequenced (see Table 2 and Fig. 4), and an additional member of this family is encoded in the D. melanogaster genome (see below).

INSECT DIURETIC AND ANTIDIURETIC HORMONES

297 +

+ Musdo

DP

Pe~ am

gP

-

goone

DH+

-

Dippu

DH+

Locmi

DH Achdo D P Te m o D H ~ Manse

-

T G

S G

P

+ T G

A V

P

-

T G

. . . . -

XelJla C R F sauvacjlne

-

DPiI DH,

-

S

N KIP

T G P - M G MG - • G A Q AGALGESGA

DH

~; ...... D.

-

-

H v• ] l i DH: 1 Catco U I i
Manse Hvlli

-

.

.

R MIP -

.

.

.

- R M P

-NDDP . D D P - - A E EQPG

VNPL DVL RQRL L L E I ARRQMIKE NT VNPL DVL RQRL L L E I ARRQMKENS V N P L D V L RQ Rr~]L E [ ~ A R R Q M RL~EEN T VNPLDVLRQRLLLEIARR MRQSQ S VNPLDVLRQRLLLEIARR MRQSQ S VNPLDVLRQRLLLEIARR MRQTQ-NMI S VNP[~]DVL RQRLLLE IARR AE I V[~_~L DVL RQ R LFM-~E[]~]R~JRMLM ~E L QG tVN~VLDVLR RL L L E I A R K K A KEG A I[D~IP]~V L RQ L LE RKVHALRAI I D L T~JMS~LVVLRQ L LE RKVQAL RA

-

.

.

.

: : : : : : : sly

liD LD I~D V N VN I~Z

L T F H LIL

RIT

L ~ q E r ~ A R T Q SIQIR L T F H LIE RIE V l L I E M ' , _ & R I ' ~ I E O q - A L S [ ] ~ L I L ~ J K~r~- M TrE~ - - . m-l~[ ~ ] K ~ E K E " P A VIUI h L l u HIKIY M E ~, VIAIQ N P ~A V E ~ J L QLQHLRJY M E K VIALQ N ~ A[~] ~ K"~W E Q~ RIAR1K~]V

+

RQV RQV RQV DQI DQI EQI S R I - -N - - - -

E~_A E - QF~-IA H - - QQ~A A - - --7_ _ . - K -

FIG. 4 The sequences of all 13 chemically isolated CRF-related DH are aligned, together with one sequence deduced from the D. melanogaster genome. Also shown are sequences of four members of the vertebrate CRF superfamily of peptides: sauvagine from the frog Phyllomedusa sauvagei, Xetmpus laevis (Xenla-) CRF, urotensin 1 from the white sucker CalOSlOmUS commersoni (Catco-), and urocortin (rat). A lower-case 'a' at the end of the sequence signifies an amidated C-terminus: OH is shown for the two T. molitor DH with free acid termini. This alignment was performed with Clustal W (http://searchlauncher.bcm.tmc.edu/). Residues identical in at least 33% of the sequences are boxed. Four residues are completely conserved in all CRF-like DH and are indicated by an arrow over their positions. For three of the four, this complete identity extends also to members of the CRF family shown. Insect species abbreviations are defined in the text.

Lehmberg el al. (1991) and Kay el al. (1991b) isolated and sequenced independently a DH from L. migratoria (Locmi-DH). They arrived at the same sequence, although Lehmberg el al. utilised a direct enzyme-linked immunosorbent assay (ELISA) with Manse-DH antiserum for monitoring isolation, whereas Kay el al. monitored their isolation by measuring elevation of cyclic A M P produced by MTs of L. migratoria in the presence of 0.1 mM 3-isobutyll-methylxanthine (IBMX). The latter assay had earlier been used to isolate a 46 amino acid CRF-related peptide (Achdo-DP) from 1000 heads of A. domestitus (Kay et al., 1991a). Later Patel et al. (1995) provided unequivocal evidence of a hormonal role for Locmi-DH, the first proof that peptides of this family can in fact be termed diuretic hormones. Kay el al. (1992) used cyclic A M P production by L. m~gratoria tubules to isolate a peptide (Peram-DP) from 600 heads of adult PeHplaneta americana, employing the same purification steps as used for Achdo-DP. Peram-DP was shown to also be active on tubules of A. domesticus and P. americana; the latter tubules are difficult to work with and hence were not used for the isolation assay. Clottens el al. (1994) reported the identification of a 44 amino acid peptide, Musdo-DP, from both the house-fly Musca domestica and the stable fly Smmo.',ys cah'itrans. They started with 6.35 kg of whole house-flies (c.444 500 animals) or 3.6 kg of stable flies (c.662400 animals) because of evidence of occurrence of diuretic activity in thoracic ganglia, abdominal nerves, and

G. M. COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLE

298

F-

,>-.

0

0

.j

o

.g q...

L

~o

tO

©

TO r.-

o

{o

LE

¢-

~ . ~ o 0 rG

~9

C,'~ H

.<

-4.4

INSECT DIURETIC AND ANTIDIURETIC HORMONES

299

brain in other fly species, Extracts were filtered and pre-purified using two successive CI~ solid-phase extraction steps to remove massive impurities, followed by a seven-step RPLC isolation scheme for the peptide from M. domestica, and an eight-step RPLC isolation for S. calcitrans. Both isolations were monitored using a bioassay based on the secretion of cyclic AMP from MTs of adult M. sexta (see Table 2). Later, Audsley et al. (1995) advocated use of M. sexta tubules as a general, heterologous assay for CRF-like DHs. Interestingly, analysis of the material isolated from M. domestica gave the sequence of a 44 amino acid peptide, Musdo-DP, but the Mr determined by mass spectroscopy (MS) was 16Da higher than the calculated mass. These researchers concluded that the mass difference was due to oxidation of the single Met residue. Consequently greater attention was paid to isolating the S. calcitrans DH quickly. In fact, two peaks were observed in the last isolation step for Musdo-DP from the latter organism; both peaks gave the same sequence following Edman degradation, but the faster eluting, minor peak proved to be the unoxidised form, while the slower eluting peak had 16 Da higher M,.. The use of water miscible antioxidants during isolation, such as thiodiglycol, 2-(methylthio)ethanol, or even methionine, will prevent such oxidation (Lehmberg et al., 1991). Once authentic, synthetic peptide was available, it was tested in a secretion bioassay with MTs of M. domestica, and found to double fluid secretion; the same result was obtained with the 'natural', oxidised form of the peptide. Two CRF-related DHs have been isolated from T. molitor. These are of some interest in that they are the only members of the CRE-like DH family to have their C-termini in the non-amidated form. Attempts to isolate these peptides using MTs of M. sexta were unsuccessful. In hindsight, this is likely to be explained by the fact that these beetle DH have a free acid at the carboxyl terminus. As mentioned earlier, M. sexta MTs have little or no intrinsic response to Manse-DH with a C-terminal carboxyl group. Attempts to utilise the EL1SA for Manse-DH used in isolating Locmi-DH (Lehmberg et al., 1991) gave only false positive results for a very broad peak of proteins eluting at higher acetonitrile concentration than that required to elute all other CRF-like DH. Successful isolation of DH t¥om this species required development of an assay based on quantification of cyclic AMP secreted by MTs of newly eclosed adults of T. molitor. Extracts of 8400 heads were processed through seven steps of RPLC and one step utilising hydrophilic interaction LC (HILC). This separation gave a somewhat poor recovery of biological activity, perhaps due to precipitation of peptide because of the need for injecting sample dissolved in 80% acetonitrile. The last isolation step proved unnecessary, because a single component was observed. Edman sequence analysis of the pure material showed it to be a 37-residue peptide whose alignment with the other CRF-like DH requires extension of the C-terminus with an Asp-free acid residue (see Fig. 4). While this peptide had at least 10 000-fold lower activity in elewtting cyclic AMP than Manse-DH in M. sexta tubules, it is highly potent in

300

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

T. molitor tubules, with an ECs0 of 2.6 nM for elevation of cyclic AMP. Interestingly, Manse-DH is only 17-fold less potent on T. molitor tubules (Furuya et al., 1995). In the first isolation step of Tenmo-DH37, a less abundant, more hydrophobic zone of biological activity was detected. Another 20000 heads of T. molitor were collected and processed through the first step of the RPLC isolation. The more retained active fraction was combined with the saved active fraction from 8400 heads and processed through an additional seven RPLC steps. The HILC purification step was not utilised, despite its remarkably high selectivity, because of the apparent loss of activity at this step. The peptide was sequenced, revealing that it was a 47 amino acid peptide terminating in a Leufree acid residue and lacking the Asn extension of the first T. molitor DH. This peptide is termed Tenmo-DH47 (Furuya et al., 1998). Tenmo-DH47 has an ECs0 of c.l.6/JM, 600-fold less potent than Tenmo-DH37; while originally named Tem-DHI, the authors renamed this peptide based on the number of amino acid residues, largely because calling the larger peptide DH2 (or DPII) would have inverted the relative order compared with Manse-DH/-DPII. Subsequently the ECs0 values of the two peptides have been determined for fluid secretion in a Ramsay assay with the free portion of larval T. mo/itor tubules; the values of 0.12nM for Tenmo-DH37 and 26nM for Tenmo-DH47 (Wiehart et al., 2002) are considerably lower than the values based on cyclic AMP secretion. Also in the larval fluid secretion assay, the relative differences in potency are less (c.200-fold) than in the adult cyclic AMP assay, A single DH was isolated from heads of Zootermopsis nevadensis, a damp wood termite, using elevation of cyclic AMP secreted by tubules of M. sexla as criterion for function. Extraction of 45.3 g of heads (c.9800 heads from mixed nymphs and adult females) gave material that was pre-purified by C4 solidphase extraction. It was then purified on a strong cation-exchange LC column, revealing two peaks of activity. The faster eluting peak (less basic) was purified to homogeneity in three RPLC steps and sequenced. It proved to be a 46residue peptide differing at only three residues for Peram-DP. Regrettably, no estimate for the amount of recovered peptide is available. The slower eluting peak from ion exchange proved to be faster eluting (more hydrophilic) on reversed-phase than Zoone-DH. It was further purified, but on the third RPLC step all activity was lost. Another 20 000 termite heads were collected from the wild, and the material was processed through ion exchange LC followed by three successive steps of RPLC. Similarly, the activity was again lost on the third RPLC step, although a different column was used. A possible explanation for this phenomenon comes from the known instability of Asp Pro bonds at low pH: nine of the 14 CRF-related DHs contain an Asn Pro (NP) bond in a conserved domain near the N-terminus (see Fig. 4), regarded as essential for receptor activation and binding (see section 5.5.2). In two other DHs, this partial sequence is Ala Pro, in two others Leu Pro, and in one other Asn Set. A single base change in a gene encoding Asn Pro can change it to

INSECT DIURETIC AND ANTIDIURETIC HORMONES

301

Asp Pro, a partial sequence in proteins notorious for its lability to cleavage under acidic conditions. The non-isolable second factor is not the only DH which has proven extremely difficult to isolate by RPLC (the pH of the mobile phase is usually c.2); similar problems were encountered in repeated attempts to isolate a DH from heads of the honeybee Apis mellilera, using again MTs of M. sexta for bioassay of fractions (E. Lehmberg and D. A. Schooley, unpublished observations). No estimate of the amount of Zoone-DH recovered was reported. The probable existence of a second CRF-related DH in Z. twvadensLs" is of considerable interest. To date, no homologue of a "short" CRF-related DH has been isolated from a hemimetabolous species. Two CRF-related DH were identified from the white-lined sphinx, H r l e s lineata (a sphingid moth closely related to M. se.vta). Each differs from Manse-DH and Manse-DPll at a single residue (see Fig. 4). They were isolated to homogeneity from only 500 moth heads, using release of cyclic AMP from M. se.v/a tubules as the bioassay to guide purification. The two factors separate in the first, semi-preparative RPLC isolation step. The slower eluting peptide (Hylli-DH41; the number signifies the count of residues) has a Gin at residue 27 in place of a His in Manse-DH. The faster eluting peptide (Hylli-DH30) has a Glu in place of Asp at position 5 in the sequence of Manse-DPll. These peptides have not been synthesised to date because of the extreme similarity to the M. sexta homologues. The most recently isolated CRF-related DH is from the Pacific beetle cockroach, Diploptera punctata. Dissected brain CC complexes from 1040 animals were used as the tissue source. The extracts were screened for biological activity using M. se.vla tubules challenged with test tYactions; release of cyclic AMP was measured with a modified Gilman assay (Gilman, 1972) as usual. Only one fraction had biological activity on M. sexta tubules. However, because of the evolutionary distance between the species from which the extract was made and that used for assay, the fractions were also assayed using tubules of the grasshopper Schislocerca americana, as usual monitoring release of cyclic AMP. A second, faster-eluting fraction was detected with the S. americana assay, in addition to the fraction active on M. sexta. The factor active on both M. se_vta and grasshopper tubules was isolated using M. se.vta tubules because this assay was easier than with grasshopper tubules: only two additional steps of RPLC purification were required to obtain homogeneous peptide. Similarly, the same two RPLC steps sufficed for isolating the factor active only on S. americana to homogeneity. Sequence analysis of the peptide active on both species showed it to be a 46-mer, christened Dippu-DH4~,. Curiously, this peptide is less similar to Peram-DP, the other cockroach DH, than is Zoone-DH (see Fig. 4). Dippu-DHar, was shown to have an EC50 of 13 nM in D. pmu'lata or 110 nM in L. migraloria, a species that responds poorly to heterologous DHs (Coast el al., 1994; Audsley el al., 1995). A huge surprise came upon sequencing the peptide active only on S. americaml: it is a 31-residue peptide with little or no sequence similarity to

302

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

any CRF-related DH! A BLAST search revealed no similarity with known neuropeptides; some proteins with partial similarity were observed. However, a manual search of the Peninsula Labs, Inc. Catalogue of bioactive peptides showed that the only known neuropeptides with Arg Pro-amide at the Cterminus belonged to the calcitonin family (see Fig. 5). Chicken calcitonin was found to have significant biological activity on D. punctata tubules: while the ECso was about 30-fold higher than Dippu-DH~l, calcitonin elicited a higher stimulation of tubule secretion. A synthetic sample of Dippu-DH~j was synthesised and tested: on D. punctata tubules, it had a slightly lower ECs0 (9.SnM) than Dippu-DH46 and lower maximal stimulation. When tested together, each of the peptides was found to potently synergise the effects of the other (see section 8.3). Table 2 summarises the isolations of CRF-like D H known to date. Of some interest is the abundance of peptide recovered per whole animal or per head. The abundance of these peptides varies by almost 1000-fold between the species investigated to date (with the exception of C. salinarius, whose sequence has been published but without details of the isolation; Clark et al., 1998b). There does not seem to be an obvious correlation between size of insect and abundance of peptide; A. domesticus and P. americana have c. 1 pmol of DH per head, yet A. domesticus is a rather small insect, not terribly much larger than the flies containing 3 6 fmol D H per head. Most hemimetabolous insects studied contain rather large amounts of DH, yet attempts to isolate CRF-like D H from several thousand heads and thoraces of R. prolixus have been fruitless to date (J. J. Hull and V. Te Brugge, personal communication).

Galga-CT Oncke-CTl Oncki-CT3 Homsa-CT Suscr-CT Dippu-DH31 Forpo-DH? Drome-DH31

FIG. 5 Shown above is an alignment of calcitonin from five vertebrate species, and two complete and one incomplete sequence of "calcitonin-like DH" from insects. The sequences are from chicken (Gallus gallus, Galga-CT), chum salmon (Oncorhynchus keta, Oncke-CT1), coho salmon calcitonin 3 (Oncorhynchus kisutch, Oncki-CT3), human calcitonin (Homsa-CT), porcine calcitonin (Sus scrq/~l, Sussc-CT), D. punctata DH~j (Dippu-DH30, a partial sequence from F. polyctena (Forpo-DH?), and D. melanogaster (Drome-DH30. Residues identical in the calcitonin family and the CTlike DH are boxed, and similar sequences (PRIFT criteria; see Cornette et al., 1987) are shaded. There are nine residues identical (29%) and an additional seven (23%) similar between Dippu-DH31 and the calcitonins, not counting the downstream Gly giving rise to the C-terminal amide. While this is not as high as the similarity between the CRF-related DH and the CRF superfamily, chicken calcitonin is only 30-fold less potent than Dippu-DH~) on D. punctata MT.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

303

4.2.2.2 Structures q/" CRF-related DH. Insects are the most diverse and successful order of living organisms, and this is reflected in the CRF-like DH, which are apparently the most structurally diverse family of peptide hormones known; gaps are required in all but one sequence (Achdo-DP) for optimal alignment (see Fig. 4). In contrast, alignment of the complete vertebrate CRF superfamily (not shown) shows no gaps and extremely high sequence conservation. The vertebrate family consists of eight highly similar CRFs from 11 species, plus paralogous sequences: three urotensin 1 and five urocortin sequences, and sauvagine. When performing an alignment such as that shown in Fig. 4, inclusion of more than a minimal number of members of the vertebrate peptides in the data to be aligned biases the results, appearing to favour the highly conserved sequences, resulting in peculiar "phylogenetic trees" as a result of the algorithm used. In the alignment of the DH, the sorting of the DH sequences into what we call the short DH (sorting below the CRF family) and the long DH (sorting above the members of the vertebrate family) is of significant interest. To date, short DH sequences are known only from the Endopterygota, but there is evidence they are likely to exist in Exopterygota (see section 4.2.2.1). There are only four invariant residues in the CRF-related DH; the residues are non-adjacent and indicated by arrows over the alignment shown in Fig. 4. Those conserved are a Ser and a Leu residue in the N-terminus, and C-terminal Asn and Leu residues. The regions of high structural similarity are rich in amino acids having highly degenerate codons (Arg, Leu, Ser). Perhaps not surprisingly, to date, only one precursor protein for a CRF-related DH has been cloned that for Manse-DH (Digan el al., 1992). The precursor protein for Manse-DH encodes a 19 residue signal sequence for secretion, followed by a 61 residue propeptide including a dibasic cleavage site (Lys Arg), followed by the 41-residue sequence of Manse-DH proceeded by a Gly (amidation signal) and then a dibasic: cleavage site (Lys Arg) followed by an 18 residue propeptide. The N-terminal propeptide contains two Gly-Arg sequences, which may be recognition sites for a monobasic residue-processing enzyme coupled with an amidation signal. It is there[ore possible that other bioactive peptides may be processed from this precursor protein. It would be of obvious interest to know precursor peptides for other members of this family, but they present challenges in molecular cloning studies. An in silico cloning of the D. me/anogasler genome by BLAST analysis identifies a gene (Dh; CG8348) localised at 85E2 on chromosome 3R thai encodes a putative CRF-related DH. The sequence out to residue 33 differs from Musdo-DP by a single amino acid substitution, Ser ~1 for Thr 31. but thereafter it is very different and lacks a C-terminal processing site. To form the InRNA that encodes the full sequence of Drome-DH, another intron/exon excision is needed (J. Vanden Broeck, personal communication). The full coding sequence is in an EST deposited in GenBank (BF499889), which shows Drome-DH is identical to Musdo-DP (Vanden Broeck, 2001; see Fig. 4) apart from the aforementioned substitution and is flanked by Lys Arg processing sites with a C-terminal amidation signal.

304

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

The existence of paralogous sets of short and long CRF-like DH is analogous to the existence of highly conserved CRF sequences in vertebrates, accompanied by the paralogues sauvagine, urotensin I, and urocortin. Four different CRF receptors are knowm CRFI and three splice variants of CRF2 (CRF2~, CRF2fi, and CRF2y). The CRF1 receptor prefers CRF as a ligand and is localised predominantly on pituitary cortieotropes, whereas CRF2 receptors are localised in different parts of the body and prefer urocortin as a ligand (Dautzenberg and Hauger, 2002). CRF and urocortin are believed to play rather different roles in vertebrates. At the moment, only two CRF-like DH receptors have been cloned, one each from M. sexta (Reagan, 1994) and A. domesticus (Reagan, 1996), but the presence of two sets of paralogous sequences suggests that two types of receptor may coexist and be associated with different roles as in vertebrates. In this context, it is worth noting that two putative CRF-like DH receptors, which appear to be paralogues, are encoded in the fruit fly genome (Hewes and Taghert, 2001; see section 5.2.3.1).

4.2.3

Calcitonin-like peptides

The isolation of the first calcitonin-like peptide (Dippu-DH~l) has already been mentioned together with the isolation of Dippu-DH46 (see section 4.2.2). It appears that this peptide constitutes but the first example of a family of DH (Fig. 5): in silico 'cloning' (using computer hardware and software) of the D. mehmogaster genome with Dippu-DH31 as a query resulted in the location of a putative peptide with 71% sequence identity to Dippu-DH3b flanked by appropriate cleavage and amidation sites. The gene (Dh31; CG13094) is localised at 29D1 2 on chromosome 2L. Synthesis of the peptide and extensive biological testing showed that it functions as a diuretic in D. melanogaster (Coast et al., 2001). The only other similar peptide to be chemically isolated and partially sequenced is described in the thesis of Laenen (1999), who isolated a peptide from the Belgian forest ant, Formica polyctena, based on its ability to stimulate MT writhing in L. migratoria: sequence analysis of this peptide was only possible out to 29 residues, but all are identical with the first 29 residues of Dippu-DH3~ (see Fig. 5)! The MS analysis revealed that the sequence was incomplete; the Mr is lower than that of Dippu-DH31. Because this peptide lacked diuretic activity on tubules of F. polyctena, it was not investigated further. Nevertheless, it is likely that these peptides constitute a new family of DHs, because in several species of insects they have potent activity and have been isolated from diverse orders (Dictyoptera, Diptera and Hymenoptera). Unpublished studies show that peptides immunologically related to Dippu-DH31 exist in T. molitor (V. C. Lombardi and D. A. Schooley, unpublished observations) and also in A. domesticus and R. prolixus (J. J. Hull and D. A. Schooley, unpublished observations; see section 6.1.7).

INSECT DIURETIC AND ANTIDIURETIC HORMONES

4.2.4

305

Diuretic/myotropic kinin neuropeptides

Insect kinins (originally called myokinins) were first isolated from whole head extracts of the Madeira cockroach (Leucophaea maderae; leucokinins; LeumaKs) and the house cricket (A. domesticus; achetakinins; Achdo-Ks) on the basis of their myotropic activity in a cockroach (L. ma&'rae) hindgut assay (see Holman el al., 1990; Holman et al., 1991). For details of the extraction and high-performance liquid chromatography (HPLC) methodologies used in the isolation of these peptides, see Holman and Hayes (1997). Reasoning that peptides acting on the hindgut might have other actions on the excretory system, Leuma-Ks and Achdo-Ks were tested on isolated MTs from A. aegypti (Hayes et al., 1989) and A. domesticus (Coast et al., 1990a), respectively. In both species, kinins had diuretic activity and, in the mosquito, they were also shown to depolarise the TEP of perfused tubules. The ECsos of the five kinins from A. domesticus in the diuretic assay ranged from 18 pM for Achdo-K-V to 324pM for Achdo-III (Coast et al., 1990a). Other kinins have been isolated from orthopteran, dictyopteran and dipteran insects using either a myotropic assay or an ELISA with antisera raised against leucokinins. The isolation of kinins from a lepidopteran insect, the corn earworm, Helicoverpa zea, is of some interest because an assay for diuretic activity was employed. Blackburn et al. (1995) used c.2000 abdominal ventral nerve cords (VNC) from adult moths as a tissue source. Homogenised tissues were pre-purified on a Sep-Pak Cts, and then fractionated on RPLC. Fractions were assayed using a Ramsay assay with adult M. sexta MT. Three fractions with diuretic activity coincided with peaks having strong absorbance at 280nm, which gave second derivative UV spectra characteristic of Trp, one of three amino acids totally conserved in the insect kinins. Subsequent purification steps were monitored solely based on the characteristic UV spectral behaviour of Trp, rather than any bioassay. The fastest eluting peak was purified to apparent homogeneity in a single RPLC step, and sequenced by Edman degradation (Helze-K-III; see Table 3). The other two factors, apparently present in smaller quantity, required two additional steps for purification to homogeneity. Edman sequencing did not give a clear result tk~r the presence of the probable Trp residue, so these two factors were additionally sequenced by coupled MS MS techniques. The deduced sequences, Helze-K-I and Helze-K11 (see Table 3) were synthesised, like Helze-K III, and bioassayed on MT of M. sexta. No ECs0 values were determined, but rather 'threshold values' the lowest concentration at which a statistically significant effect on fluid secretion could be measured. These were c.7 pM for Helze-K-l, 0.6 pM for Helze-K-II, and 6 pM for Helze-K-II1. In addition to the identified insect kinins, there is immunocytochemical evidence (see section 6.1.5) for their presence in A. mellffera and in R. prolixus, although interestingly the kinin-like material in R. prolixus does not stimulate tubule secretion hut is active on the hindgut (Te Brugge et al., 2002).

306

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

TABLE 3 Kinin sequences with conserved residues in the C-terminal pentapeptide shown in bold type Species L. maderae

P. americana

A. domesticus

L. migratoria C. salhunqus

d. aegypti

M. domestica D. melanogaster H. zea

P. vannamei L. slagnalis

Peptide

Sequence

Leuma-K-I DPAFNSWGa Leuma-K-ll DPGFS SWGa Leuma-K-lll DQAFNSWGa Leuma-K-IV DASFHSWGa Leuma-K-V GSGFS SWGa Leuma-K-VI pESSFHSWGa Leuma-K-Vll DPAFSSWGa Leuma-K-VIll GASFYSWGa Peram-K-1 RPSFNSWGa Peram-K-2 DASFS SWGa Peram-K-3 DPSFNSWGa Peram-K-4 GAQFS SWGa Peram-K-5 SPAFNSWGa Leuma-K-Vll DPAFS SWGa Leuma-K-Vlll GASFYSWGa Locmi-K AFHSWGa Achdo-K-I SGADFYPWGa Achdo-K-ll AYFS PWGa Achdo-K-IIl ALPFSSWGa Achdo-K-1V NFKFNPWGa Achdo-K-V AFHSWGa Locmi-K AFS S WGa Culsa-K-I NPFH SWGa Culsa-K-ll NNANVFYPWGa Culsa-K-llI TKYVSKQRFHSWGa Aedae-K-I NSKYVSKQKFYSWGa Aedae-K-II NPFHAWGa Aedae-K-lll NNPNVFYPWGa Musdo-K NTVVLGKKQRFHSWGa Drome-K NSVVLGKKQRFHSWGa Helze-K-I YFS PWGa Helze-K-ll VRFS PWGa Helze-K-IIl KVKFSAWGa Penva-I ASFSP YGa Penva-2 DFSAWAa Lymst-K PSFSSWSa

Reference Holman el al. (1986a,b,c) Holman el al. (1986a,b,c) Holman et al. (1986a,b,c) Holman el ell. (1986a,b,c) Holman et al. (1987a,b) Holman et al. (1987a,b) Holmanetal.(1987a,b) Holman et al. (1987a,b) Predelel al. (1997) Predelel al. (1997) Predelet al. (1997) Predelet al. (1997) Predelet al. (1997) Predel et al. (1997) Predelel al. (1997) Predelet al. (1997) Holman e t a / . (1990) Holman et al. (1990) Holmane:al. (1990) Holman et al. (1990) Holman et al. (1990) Schoofsel a/. (1992) Hayesel a/. (1994) Cady and Hagedorn (1999a,b) Cady and Hagedorn (1999a,b) Veenstra(1994) Veenstra(1994) Veenstra(1994) Holman et al. (1999) Terhzazel a/. (1999) Blackburne: al. (1995) Blackburn et al. (1995) Blackburn et al. (1995) Nieto et al. (1998) Nieto eta/. (1998) Coxet a1. (1997)

Insect kinins are characterised by the C-terminal sequence P h e - X a a l - X a a 2T r p - G I y - N H 2 , where Xaa 1 is Asn, His, Phe, Ser, Tyr, and Xaa 2 is Ser, Pro or, less frequently, Ala (see Table 3). There is some departure from this in kinins identified from the white shrimp P e n a e u s v w m a m e i (Nieto e t a l . , 1998), with T y r replacing T r p in P e n v a - K - I a n d G l y - a m i d e being replaced by A l a - a m i d e in Penva-K-2, a n d the p o n d snail, L y m n a e a s t a g n a l i s (Cox e t a l . , 1997), which terminates in Ser-amide (see Table 3). Veenstra e t a L (1997) isolated a e D N A encoding the A e d a e - K prepropeptide from an A . a e g y p t i a b d o m i n a l ganglia e D N A library. The 2 2 8 a m i n o acid prepropeptide has an 18-residue signal sequence and c o n t a i n s single copies

INSECT DIURETIC AND ANTIDIURETIC HORMONES

307

of Aedae-K-l, 2 and 3, each followed by an amidation signal (Gly-Lys-Arg). Other possible products from the prepropeptide have no resemblance to known peptide sequences. The gene encoding Drome-K (CG13480; pp) was identified in low-stringency TBLASTN searches of Genbank and the Berkeley Drosophila Genome Project (BDGP; Terhzaz et al., 1999) and is localised at 70E3-70F4 on chromosome 3L. The Drome-K sequence is flanked by typical dibasic (LysArg) cleavage sites and a C-terminal amidation signal. There is insufficient sequence information to know whether other peptides are encoded on the same gene. Based upon MS data, an identical kinin is present in the flesh fly, Neohellieria bullata, whereas a peptide identical to Musdo-K occurs in the horn fly, H a e m a t o b i a irritans, and the stable fly, S l o m o x y s calcitrans (R. Predel and R. J. Nachman, unpublished observations). Thus the kinins of cyclorrhaphan flies appear to be highly conserved, as are their CRF-related peptides (see Fig. 4). 4.2.5

Cardioaccelerato W peptide 2h/periviscerokinins

A number of cardioacceleratory peptides (CAPs) are present in the central nervous system (CNS) of the tobacco hawkmoth M . sexta, and are involved in regulating heartbeat during wing inflation after emergence and during flight (Tublitz and Truman, 1985; Tublitz and Evans, 1986; Tublitz, 1989). Initial attempts to purify the active factors from VNC extracts of pharate adults using gel filtration chromatography yielded two peaks of activity (CAP1 and CAPe) when tested on the abdominal heart of a pharate adult male (Tublitz and Truman, 1985). With RPLC these fractions were resolved into at least two CAPls (la and lb) and three CAPes (2a, 2b and 2c) (Cheung et al., 1992). Of these, CAPea (=crustacean cardioactive peptide" CCAP) and CAPeb have been fully sequenced (Cheung el al., 1992; Huesmann el al., 1995). ManseCAP>, is an octapeptide with blocked N- and C-termini (see Table 4), which TABLE 4 The sequence of Manse-CAPeb compared with that of putative CAPehlike peptides from D. melanogaster and with identified PVKs. Bold type shows sequence identity to Manse-CAPxb Species M. sexm D. melanogaster P. [llll~'FiCdlla L. nladerae

L. migratoria

Peptide M anse-CAPeb Drome-CAP:~-I Drome-CAPeb-2 Peram-PVK- 1 Perarn-PVK-2 Leuma-PVK-1 Leuma-PVK-2 Leuma-PVK-3 Locmi-PVK- 1

Sequence

Reference

pELYA g PRVa GANMGLYAFPRVa ASGLVA F PRVa GASGL I PVMRNa GSSSGL 1S M PRVa GSSGL1 P FGRTa GSSGL I S MPRVa GSSGM I PFPRVa AAGL FQ F PRVa

Huesmann et al. (1995) Vanden Broeck (2001) Vanden Broeck (2001) Predel et al. (1995) Predel et al. (1998) Predel et al. (2000) Predel et al. (2000) Predel et al. (2000) Predel and Giide (2002)

308

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

in addition to its cardioacceleratory activity, stimulates secretion by fruit-fly MTs (Davies el al., 1995), although it has no effect on hawkmoth tubules (N. Tublitz, personal communication}. Using in silico cloning, Vanden Broeck (2001) identified the D. melanogasler CAP2b gene (CG15520; capa), which localises to 99C8-99D1 on chromosome 3R. The gene encodes two putative CAP2b-like sequences (Table 4) with the same C-terminal as Manse-CAP2b (AFPRV), each followed by an amidation signal. The 151-residue prepropeptide begins with a 16 amino acid signal sequence and, in addition to the two CAP2b sequences, contains a 15-residue peptide flanked by cleavage sites (Lys Arg) and a C-terminal amidation signal. The C-terminus of this putative peptide (FGPRL-NH2) is characteristic of pyrokinin/pheromone biosynthesis activating neuropeptides (PBAN), which are known to have myotropic activity on cockroach hindgut (Holman et al., 1986b). It will be interesting to know whether this peptide has any effect on MT secretion. Manse-CAPzh-like peptides have also been identified in extracts of perivisceral organs from P. americana (Predel el al., 1998), k. maderae (Predel et al., 2000) and the migratory locust, L. mig, ratoria (Predel and Gfide, 2002). Unfortunately, these CAP2~-like peptides were christened periviscerokinins (PVKs), which, although reflecting their tissue of origin does not recognise their sequence similarity to Manse-CAP2b (see Table 4). The C-termini of Peram-PVK-1 and Leuma-PVK-1 are dissimilar from Manse-CAP2b, although their N-termini are strikingly similar to the CAP2b-like peptides of both species (see Table 4). 4.2.6

Tenebrio molitor and Leptinotarsa decemlineata A D F

Antidiuretic factors that inhibit MT secretion have been characterised from two coleopteran species. Lavigne el al. (2001} reported the purification of a factor from L. decemlineata through five RPLC steps using only two different columns; the protocol used involved minimal changes of parameters between successive purifications, and utilised evaporation steps between RPLC purifications likely to result in poor recovery (see section 4.2.1). While no peak was visible in the final chromatogram, they reported its apparent molecular size (estimated by dialysis) as 25-50 amino acids. Antidiuretic factors have been isolated and identified recently from pupal heads of T. molitor (the same stage of T. molitor used for isolation of Tenmo-DH3v and -DH47); the hydrophobicity of these factors (as reflected by retention behaviour on RPLC colmnns) and their size (13 and 14 amino acid residues) suggests they may be related to the L. decemlineata factor. The two ADF from T. molitor were isolated based on their ability to elevate cyclic GMP production by MTs (assayed using a commercial enzyme immunoessay (EIA)), rather than using a fluid secretion assay as did Lavigne et al. (2001). Fluid secretion assays for these ADF are difficult, because one

INSECT DIURETIC AND ANTIDIURETIC HORMONES

309

must quantify the decrease in basal secretion. The extraction procedures developed allowed a highly selective pre-purification from 1500 head equivalents. After fat removal with dichloromethane, heads were extracted with 90% aqueous methanol, which recovers a lot of protein but little activity. However, extraction of the pellet with a p H 4 acetate buffer gave good recovery of biological activity and relatively few impurities. This extract was pre-purified on a weak cation exchange cartridge (Toyopearl CM650M); the activity was retained upon loading the column with the 20 raM, pH 4 acetate buffer, but eluted on changing to a 2 0 m M , pH 7 acetate buffer. This highly enriched eluate required only three steps of RPLC on narrow-bore columns to obtain each factor in a pure state. The factors were sequenced, synthesised, and assayed for their effects on T. molitor tubules in secretion assays (laboratory of Dr S. W. Nicolson), The larger peptide, T e n m o - A D F a , has the sequence V V N T P G H A V S Y H V Y - O H and a calculated pl of 6.89 (Compute p I / M W tool at www.ExPASy.ch: the only ionic sites are two His residues and the free N- and C-termini). It has an exceptionally potent effect on T. molitor tubules (ECs0 10fM), but shows receptor desensitisation or internalisation at high (l nM and above) concentrations (Eigenheer et al., 2002). The smaller peptide, T e n m o - A D F b , has the sequence Y D D G S Y K P H I Y G F - O H and an ECs0 of 240pM in a fluid secretion assay, 24000-fold less potent than its congener. T e n m o - A D F b is also more acidic than its congener. (The former has a calculated pl of 5.17, which is more consistent with the poor solubility in acidic extracting solvent.) BLAST searches of these two sequences revealed that T e n m o - A D F a has an interesting similarity to the endothelins: 57% identity to rabbit big endothelin I, or 64% allowing for the conservative substitution of Val for Leu (see Fig. 6). However, it seems inappropriate to term T e n m o - A D F a an endothelin-like peptide, because the similarity is entirely to that part of endothelin which is removed in the proteolytic processing of big endothelin I to the potent vasoconstrictor endothelin 1. In fact, the N-terminus

. . . . 3880 7enmo J~J:)Fa

- A G Y H A P L V H[~Y A Y S A P[~F R A A T L S T V ~ A [ ~ I s [ ~ H V y . . . . . . . . . . . . . . . . . . . . . . . . . . VlV N T P G HIAIVlSIYIH V Y . . . .

Homsa ETZ Tenmo ADFb TmPCPg.2

. . . . . .

.

AGP

.

.

.

C S C SLSSJL M D K E CLVJY F C H L D I I ~,lV N T P]ELHJVlVlPJY GJL G S P - R S .

.

.

VAYAAVP

.

.

.

.

.

.

.

.

AGSGLEGQWI

.

.

.

.

.

PD

.

I NEKL

GI F

Y D O G S Y K P H

I IY

YDDGSYKPH

IYL~F

-

-

-

FIG. 6 Sequence alignment of T. molitor (Tenmo) ADFa and b, T. molitor cuticle proteins CAA03880 (40 C-terminal residues only) and TmPCP9.2 (40 C-terminal residues only), and rabbit (Oryctolagus cuniculus, Orycu-ET1) and human (HomsaETI) big endothelin I, aligned using Clustal W. Identical residues are boxed. Big endothelin I is cleaved by endothelin converting enzyme at the W V bond into the vasoconstrictor endothelin 1; the similarity to Tenmo-ADFa occurs in the inactive part of big endothelin I, on the C-terminal side of the cleavage site. The similarities were determined using the BEAST algorithm against the non-redundant database, but setting the expect value from 10 to 10000, which is important R~r short sequences.

310

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

of the cleaved, inactive piece of endothelin I is at the amino terminal residue of ADFa. Interestingly, Tenmo-ADFa is identical, except at residue 4, to the 14 C-terminal residues of a T. molitor cuticle protein (CAA03880) and TemnoA D F b is completely identical to the 13 C-terminal residues of T. molitor putative cuticle protein 9.2 (TmPCP 9.2; Baernholdt and Andersen, 1998; see Fig. 6). Because of the far lower potency of Tenmo-ADFb, and its disturbing complete identity to a cuticle protein of this species, its structure was not submitted for publication until immunocytochemical evidence was available that it appears to be produced mainly in two pairs of bilaterally symmetrical cells in the protocerebrum (Eigenheer et al., 2002b). This evidence is consistent with a bonaJide neuropeptide role for Tenmo-ADFb, although its relative lack of potency compared with Temno-ADFa may mean its primary role is something other than antidiuresis. About 33 fmol of A D F a was recovered per head (Eigenheer et al., 2002a), compared with c.200 fmol Tenmo-ADFb (Eigenheer et al., 2003). In contrast, the same pupal heads of T. molitor contain c.45 fmol per head of Tenmo-DH37, although head extracts have only a diuretic effect (Wiehart et al,, 2002). This is again consistent with the desensitisation phenomenon observed in the response to the ADF, in contrast to relatively low desensitisation with CRF-like DH. 4.2.7

M. sexta antidiureticJactors

Liao et al. (2000) demonstrated the presence of antidiuretic activity in brain CC-CA complexes of M. sexta that stimulates fluid uptake from everted rectal sac preparations of larval M. sexta. Fractionation of an aqueous extract of 300 neuroendocrine complexes on a polymeric RPLC column gave two zones of activity that stimulated fluid reabsorption. The more abundant, slower eluting of these was christened Manse-ADFB. This factor appears to have much in common with the action of Schgr-ITP on the ileum of S. gregaria in that it promotes active CI- transport (see section 5.4). Interestingly, basal reabsorption from everted rectal sacs is blocked by addition of bafilomycin A~ or by amiloride, drugs which block MT secretion (see section 2.2), but these have no affect on the ADFB-stimulated fluid reabsorption (see section 5.4). This blockage of basal reabsorption allows improvement of the everted rectal sac assay for isolation; inclusion of amiloride in the bathing medium makes the effect of ADFB more evident so smaller doses can be used. Manse-ADFB was purified to partial homogeneity from an 80% methanol extract of 10 000 frozen larval heads of M. sexta; the choice of solvent was based on poor recovery of activity from tissues extracted with acidic solvents. The extract was concentrated to remove methanol, defatted, and pre-purified by absorption to a ToyoPearl QAE-550C anion-exchange cartridge. The 0.25 M NaC1, pH 8, eluent was separated by two successive semi-preparative RPLC steps, then by an anion-exchange LC step, followed by three additional RPLC steps. The apparently homogeneous product had an M,- of 8770.6 Da, in

INSECT DIURETIC AND ANTIDIURETIC HORMONES

311

the same range as the neuroparsins but c.200 Da higher than Schgr-ITP. Only partial sequence data were obtained (Liao, 2000). Based on this a cDNA clone from a brain CC CA c D N A library was isolated and sequenced: the peptide encoded by this clone was unfortunately a M. sexta homologue of cytochrome bc~ subunit H, rather than an ADF. The deduced protein sequence matched the partial protein sequence, showing that the putative ADF sequenced was an impurity, probably masking a far lower quantity of bioactive peptide (S. Nagata and D. A. Schooley, unpublished data). No attempts have been made to isolate the less abundant ADFA, and neither was it characterised to the degree that ADFB was. 4.2.8

Mosquito natriuretic peptide ( M N P )

Three fractions were isolated from a head extract of A. aegypti by RPLC on the basis of their effect on the TEP of isolated perfused MTs (Petzel et al., 1985). Fraction I depolarised the TEP, but had no effect on fluid secretion, although it increases tritiated water (THO) loss and urine output from intact flies, possibly by inhibiting fluid uptake from the hindgut (Wheelock et al., 1988). Fraction I! also depolarised the TEP, whereas the response to fraction Ill was biphasic, with the TEP first depolarising and then hyperpolarising. Fractions II and Ill both have diuretic activity and selectively stimulate secretion of NaCl-rich urine. Fraction Ili was the more potent, and its diuretic and natriuretic activity was indistinguishable from that of exogenous cyclic AMP, although the latter only hyperpolarised the TEP (Petzel et al., 1985). All three fractions contain peptides with Mr estimated by gel-filtration chromatography of 2425 (fraction 1), 2721 (fraction II) and 1862Da (fraction III) (Petzel et al., 1986). Fraction II1 was named Mosquito Natriuretic Peptide (MNP) and was shown subsequently to stimulate cyclic AMP production in isolated tubules (Petzel et al., 1987). 4.2.9

F. polyctena antidiureticjitctor ( F o p A D F )

The MTs of forest ants (F. polyctena) recently collected from the nest fi'equently fail to secrete and have a closed lumen (Van Kerkhove et al., 1989). One day later, tubules from the same batch of ants have an open lumen and are secretory, leading to the suggestion that an A D F is released in response to stress (Laenen et al., 2001). Using a 15% trifluoroacetic acid (TFA) extract of 150 000 ant abdomens, gaenen et al. (2001) isolated an ADF that reversibly inhibits MT secretion by 70%. Biological activity was lost towards the end of the purification by RPEC, but the presence of an ADF in a 15% TFA extract of haemolymph from ants with non-secreting MTs suggests it has an hormonal function. The active factor was christened FopADF, and Laenen et al. (2001) suggest that it is released when the poison gland is emptied. Since this gland contains an estimated 10% of the total body

312

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

water, it may be necessary to inhibit MT secretion to conserve water reserves, which are then replenished by drinking. 4.3

PURIFICATION AND CHEMICAL STRUCTURE OF NEUROPEPTIDES T H A T STIMULATE LOCUST H I N D G U T

4.3.1

Introduction

Early studies identified the retrocerebral complex, consisting of the median neurosecretory cells (MNCs) of the pars intercerebralis (PI) and their axons projecting to storage and release sites in the CC, as the major source of stimulants of locust hindgut. However, initiation of stimulant release from the CC using K + depolarisation in the presence of Ca 2+ has not been successful to date (J. E. Phillips and J. Meredith, unpublished observations). Stimulatory activity is also present in ventral ganglia VG4 to VG7 of locusts (Lechleitner and Phillips, 1989; Audsley and Phillips, 1990). The locust VG stimulant has different properties from that of the CC with regard to the time course for changes in the hindgut Isc, solubility, acid lability and heat stability. Bilgen (1994) partially purified the acid-labile VG factor, which had an approximate mass by size-exclusion chromatography of 37 000 Da, i.e. several times that of factors from the CC discussed below. Three neuropeptides have been purified fully or partially from the CC of locusts (reviewed by Phillips and Audsley, 1995; Phillips et al., 1998a,b): Neuroparsins, Ion Transport Peptide (ITP), and Chloride Transport Stimulating Hormone (CTSH). Only the first two have been fully sequenced. Neuroparsins and CTSH were bioassayed on recta, whereas ITP was bioassayed originally on ilea. 4.3.2

Neuroparsins

Herault et al. (1985) reported that the CC and the glandular lobe (GCC) of the L. migratoria CC both contain a rectal ADH factor, each differing in size and extraction properties. The GCC factor was not purified, but Herault and Proux (1987) report that GCC extracts cause a sharp peak in rectal tissue cyclic AMP levels coinciding with elevated J,.. This stimulation is mimicked by forskolin, a stimulant of adenylate cyclase. The factor from the CC was identified as neuroparsins, because all antidiuretic activity in crude CC of L. migratoria was abolished by an antibody to this neuropeptide (Fournier and Girardie, 1988). Neuroparsins are two proteins (NpA, NpB) isolated and sequenced from CC of L. migratoria by Girardie et al. (1989, 1990). NpB is a homodimer of a 78-residue polypeptide (8188 Da). NpA is identical to NpB except for an additional heterogeneous Nterminus, the longest of which has 83 residues. NpB is thought to be formed from NpA by cleavage of the terminal amino acids. However, Hietter et al.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

313

(1991) have proposed a revised monomeric structure for neuroparsins involving three disulphide bridges. Fournier (1991) has reviewed pharmacological and direct evidence that NpB acts on rectal J,. by stimulating the inositol phosphate (IP) cascade, resulting in elevation of cystolic Ca 2+. In summary, they conclude that neuroparsins are the only A D H in L. migratoria CC and that these peptides act via the 1P Ca 2+ second messenger system rather than via the cyclic AMP pathway. They did not study actions of purified or synthetic Nps on rectal solute transport processes. However, rectal Jv in the absence of (or indeed against) an osmotic difference across this epithelium must of necessity be driven secondarily by solute transport. Since stimulation of rectal J, is abolished in CI--free saline, antidiuretic factors such as Nps presumably must first act by stimulating the predominant ion transport process, CI- absorption. Neuroparsins might possibly also increase osmotic and cation permeability to enhance fluid transport. Some comment on the rectal Jv bioassay used by Fournier and Girardie (1988) is in order, because their results appear to differ from those reported later. These workers pre-incubate everted rectal sacs in CI -free saline for 1 h (Fournier el al., 1987), a treatment known to cause drastic loss of most CI from this tissue (Williams et al., 1977). They then restore preparations to normal C1 -containing salines, with or without stimulants or other test agents on the haemocoel side. This ionic change in itself stimulates considerable increase in Jv of the controls over the next hour. The effect of stimulants was assessed from a small additional increase in Jv in the same 1 h time period. Since the major ion pump in locust rectum is an apical C1 pump, restoring CIto C1 -depleted cells introduces a transient situation in which increased C1entry (with K +) into rectal cells with accompanying fluid would be expected to cause cell swelling. This in turn could trigger well-known cell volume regulatory mechanisms in which elevated cell Ca 2+ initiates KC1 and hence fluid exit from these cells. Thus it is not clear whether neuroparsins stimulate long-term steady-state J,. (i.e. as measured over several hours when C1- is always present) or some short-term and transient (1 h) cell volume regulatory response. Subsequently, Jell; and Phillips (1996; see also Jeffs, 1993: Phillips et al., 1998b) observed no effect of Nps over several hours on either rectal J,., or on rectal and ileal Isc of the desert locust even at high doses. An unlikely explanation might be the presence of different major stimulants in the CC of these two locust species. However, reciprocal bioassays of CC extracts from L. migratoria and S. gregaria on rectal and ileal 1so and J, in these two locust species have been more recently conducted (Macins et al., 1999). Corpora cardiaca extracts from either species were equally effective in several hindgut bioassays on both species of locust, suggesting similar stimulants are present. L. migratoria CC must therefore contain a major stimulant of S. gregaria rectal lsc and Jv other than Nps because the latter had no action on S, gregaria hindgut (Jeffs, 1993; Jeffs and Phillips, 1996). Moreover, the deduced amino acid sequences of the

314

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

ITP neuropeptide from the CC of S. gregaria and L. migratoria are identical (Macins et al., 1999), while their ITP-like (ITP-L) neuropeptides differ only by one neutral amino acid substitution (see section 4.3.4.6). A neuroparsin cDNA from L. migratoria was recently expressed using a baculovirus vector (Girardie et al., 2001) and the peptide found to have a small stimulatory effect on rectal J,, using the questionable bioassay discussed above. Further studies of neuroparsin actions on other well-characterised transport bioassays are required to clarify the conflicting reports on its stimulatory role on hindgut. 4.3.3

Chloride Transport Stimulating Hormone ( C T S H )

The CI -dependent l~c across flat sheet preparations of S. gregar& recta has been used as a bioassay to partially purify a neuropeptide stimulant from the CC (Spring and Phillips, 1980a,b). This preparation maintains a 10-fold increase in Isc for >8 h after stimulation with cyclic AMP or CC extracts. An active factor (CTSH) eluted as a single peak with an apparent Mr of about 8000Da using a size-exclusion (BioGel P-30) column (Phillips et al., 1980; Phillips el al., 1982; Phillips et al., 1986). Biological activity is destroyed by trypsin digestion, indicating that the active factor is a peptide. The estimated concentration of CTSH required to cause maximum increase in rectal Is~ is <7nM. CTSH is water, saline, and ethanol (80%) soluble, but biological activity is rapidly lost below pH6.0, which has prevented separation by RPLC. CTSH activity is largely in the CC (80%); the small amount of CTSH activity (20%) in GCC could be in nerve tracts that are known to penetrate from CC into the locust GCC. Reciprocal bioassays indicate that the CTSH fraction is different from other known or putative hormones from locust CC, such as AKH, or DHs that stimulate MT secretion. Proux et al. (1985) provided some evidence that CTSH is produced in the PI region of the brain and is transported down the paired nervi corpori cardiaci 1 (NCC I) for storage in the CC. Spring and Phillips (1980c) report that cardiatectomy reduces CTSH-like activity of the haemolymph. Haemolymph from recently fed locusts contains more CTSH-Iike activity than that from starved animals, suggesting that it is released from CC when the insect feeds (Phillips et al., 1982). Chamberlin and Phillips (1988) report that CC extracts containing CTSH cause a peak in rectal tissue cyclic AMP levels at the time o f l ~ increase. Moreover, the high l~c observed immediately after removing recta from locusts is also associated with elevated rectal cyclic AMP levels (i.e. cyclic AMP levels and l~c both decline over the first 2 h). Finally, agents that elevate tissue cyclic AMP (forskolin, theophylline) also mimic the actions of CTSH. Thus CTSH appears to act on specific ion transport mechanisms via cyclic AMP. Both CC extracts and cyclic AMP have been shown also to stimulate unidirectional 42K+ fluxes in the absence of external C[- by opening apical K + channels. Corpora cardiaca extracts and cyclic AMP also inhibit active proton secretion across the

INSECT DIURETIC AND ANTIDIURETIC HORMONES

315

rectum (reviewed by Phillips et al., 1986, 1988, 1994). However, neither rectal Na + absorption nor ammonia secretion are affected by cyclic AMP, and extracts of all neuroendocrine tissues tested had no influence on Na + transport across locust rectum (Black et al., 1987; Thomson et al., 1988). Given their common source and size, CTSH might conceivably be similar or identical to neuroparsins, which stimulate rectal Jv in L. migratoria (Fournier and Girardie, 1988). Several observations argue against this possibility. Firstly, Fournier and Girardie (1988) isolated the ADH of L. mi~,,ratoria CC (i.e. neuroparsins) using 0.1 M acetic acid, which completely destroys CTSH activity fi'om S. gregaria CC. Secondly, neuroparsins act via the IP-Ca 2+ second messenger system, whereas CTSH and CC extracts from S. gregaria appear to act via cyclic AMP. Indeed stimulation of rectal I~ in S. gregaria is unaffected by removal of extracellular Ca 2+ for several hours, or by addition of Li 2+, which is known to inhibit IP turnover. Moreover, external Ca 2+ in the presence of a Ca 2+ ionophore does not stimulate rectal l~ in S. gregaria (Jeffs and Phillips, 1996). Finally, neuroparsins supplied by the Bordeaux group did not stimulate rectal 3",. (long-term in Cl--saline) and Is~ bioassays using S. ~gregaria as previously discussed. Further progress in sequencing CTSH awaits new separation methods for acid-labile peptides, or expression cloning of the gene for this neuropeptide using the very specific rectal 1~ bioassay.

4.3.4

Ion transport peptide (ITP)

4.3.4.1 Purification, partial sequencing and actions. Using ileal l~c as the bioassay, Audsley and Phillips (1990) surveyed the whole CNS of S. gregaria for stimulatory activity. Proteinaceous factors were detected in the brain and CC, and also in VG4 to VG7, that stimulated in a dose-dependent manner. Similar results were observed using ileal Jv as the bioassay (Lechleitner et al., 1989a,b). Both CC and VG extracts stimulated ileal J,. four-fold, but only if external C1- was present. As for the rectum, CC and VG factors caused different time courses for the increase in ileal l~c. Moreover, the VG factor is much more heat labile, and its biological activity is destroyed by extraction with 0.2 mM acetic acid, unlike that in the CC. Audsley et al. (1992b) used RPLC and the ileal Isc bioassay to isolate the predominant stimulant in CC, which they named Ion Transport Peptide (Schgr-ITP). A second more hydrophobic factor, which has not yet been isolated, accounts for 30% of the total stimulation by crude CC extracts, A third fraction had little effect on ileal I~c but stimulated Jv, presumably by acting on a solute transport process other than CI- (Audsley, 1991). Purified Schgr-lTP has an unblocked N-terminus and an Mr of 8652± 3 Da. A partial N-terminal amino acid sequence (33 residues) was published (Audsley et al., 1992b, 1994). This partial Schgr-lTP sequence was 44-59% identical to hyperglycaemic (CHH), moult-inhibiting (MIH), and vitellogenesis-inhibiting

316

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

(VIH) hormones of crustaceans: locust ITP was the first member of this large neuropeptide family to be reported outside crustaceans. Purified Schgr-ITP at a concentration of 2.5nM had the same range of actions as crude locust CC extracts on the ileum; namely it caused large increases in l~c (i.e. CI transport, 10-fold), Na + transport (2-fold), K + permeability (3-fold) and isosmotic fluid absorption (4-fold), and inhibited acid secretion almost completely at high doses (Audsley et al., 1992a, 1994; Phillips and Audsley, 1995). ITP-like activity that co-eluted with Schgr-ITP during the first HPLC purification step was also detected in locust haemolymph. This is the only evidence to date that ITP might normally be released into the haemocoel to influence ileal transport activities in vivo. Given the similar molecule size and source of Schgr-lTP and CTSH, are these actually different neuropeptides? Auds[ey et al. (1992a, 1994) tested purified ITP on locust rectal transport activities. At maximum doses, SchgrITP caused only partial (40%) stimulation of rectal l~c as compared with crude CC extracts. Moreover, Schgr-lTP did not cause significant increases in rectal Jv or K + permeability, as do crude CC extracts and cyclic AMP. Audsley et al. (1992a) concluded that different neuropeptides in locust CC must stimulate the rectum (CTSH) and ileum (ITP), especially given that the biological activity of CTSH but not ITP was destroyed by extraction with acetic acid. King et al. (1999) came to the same conclusion using synthetic ITP on S. gregaria ileal and rectal bioassays. 4.3.4.2 Complete I T P amino acid sequence deduced j~om e D N A . Meredith et al. (1996) used the partial amino acid sequence of Schgr-ITP to produce degenerate oligonucleotide primers for residues 2 8 (sense) and 24-30 (antisense). These were employed with a locust brain cDNA library and the polymerase chain reaction (PCR) to clone cDNA that exactly encoded the known partial N-terminal amino acid sequence of ITP. This nucleotide sequence was extended by anchored PCR to the start and end ot" the ITP message using the 5' and 3' RACE (rapid amplification of complementary ends) system. The resulting cDNA of 517 base pairs (bp) coded a complete open reading flame for an ITP prepropeptide of 130 amino acid residues (Meredith et al., 1996). This consisted of a 55-residue leader sequence (signal peptide) with a methionine start signal and a dibasic cleavage site preceding the start of the known ITP sequence from Audsley et al. (1992b). The final sequence (Gly-Lys-Lys-stop) suggested C-terminal amidation at a second dibasic cleavage site to give a complete ITP sequence of 72 amino acid residues. The deduced prepropeptide sequence is shown in Fig. 7. This deduced ITP sequence agrees completely with the known partial sequence of the HPLC-purified ITP for residues 1 34, and also with additional unpublished sequence data (Audsley, 1991) for residues 35-45, 47, 49 and 54. Two PCR products coding for amino acids - 5 5 to +23 and 1 to 75 ITP (see Fig. 7) were used to screen a

INSECT DIURETIC AND ANTIDIURETIC HORMONES

317

-50

Met-His -His-Gln-Lys-Gln-Gln-Gln-Gln-Gln-Lys-Gln-Gln-Gly-40

-30

Glu-Ala-Pro-Cys-Arg-His-Leu-Gln-Trp-Arg-Leu-Ser-Gly-Val-20

Val-Leu-Cys-Val-Leu-Val-Val-Ala-Ser-Leu-Val-Ser-Thr-Ala-io

1

Ala-Ser-Ser-Pro-Leu-Asp-Pro-His-His-Leu-Ala-Lys-Arg-Serio Ehe-Phe-Asp-lle-Gln-Cys-Lys-Gly-Val-Tyr-Asp-Lys-Ser-Ile2o

Phe-Ala-Arg-Leu-Asp-Arg-Ile-Cys-Glu-Asp-Cys-Tyr-Asn-Leu30

40

Phe-Arg-Glu-Pro-Gln-Leu-His-Ser-Leu-Cys-Arg-Ser-Asp-Cys5o

Phe-Lys-Ser-Pro-Tyr-Phe-Lys-Gly-Cys-Leu-Gln-Ala-Leu-Leu60 Leu-

I le-Asp-

70 Glu-Glu-Glu-Lys-

Phe-Asn-

Gln-Met-Val-Glu-

I le-

72

Leu-Gly-Lys-Lys-STOP FIG. 7 The deduced prepropeptide sequence of Schgr-ITP, numbered relative to the start of ITP at number 1, with dibasic cleavage sites indicated in bold.

locust brain cDNA library. Six positive ITP clones were identified and sequenced. All of these clones confirmed the above ITP sequence derived using PCR and degenerate primers (Meredith et al., 1996). The complete deduced ITP sequence shares 39-42% sequence identity with C H H of the crab Carcinus maenas, the lobster H o m a r u s americanus and the woodlouse Armadillidium vulgare, and 29 and 30% respectively with lobster VIH and crab MIH. Including neutral amino acid substitutions, the percentage similarity of ITP to C H H homologues (Lacombe et al., 1999) is 67 71%, and to MIH and VIH it is 53 54%. The six cysteines at positions 7, 23, 26, 39, 43, 52 are conserved in all members of this peptide family. Based on disulphide bridge locations determined for crab C H H by Kegel et al. (1989), Meredith et al. (1996) predicted identical bridges at 7-43, 23 39 and 2 6 5 2 in ITP. Crustacean CHH and MIH, or indeed whole sinus gland extracts of crabs, do not stimulate locust ileal l~c. The M,. of ITP (oxidised form with three disulphide bridges) predicted from its cDNA is 8558 Da, which is less than that determined initially by MS for

318

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A, SCHOOLEY

HPLC-purified ITP (Audsley et al., 1994). Possibly, a potassium salt of ITP may have been originally analysed by MS (suggested by D. A. Schooley; see Meredith et al., 1996). 4.3.4.3 S y n t h e t i c I T P and its biological actions. King et al. (1999) synthesised ITP (SynITP) in the chemical form proposed by Meredith et al. (1996), with amidation of residue 72 and with three disulphide bridges based on crab CHH. This is the only member of this large family of arthropod neuropeptides to be synthesised. Of particular note was that during the oxidation step of the synthesis only one combination of disulphide bridges was formed without the assistance of any cellular constituents. The biological activities of this synthetic ITP were consistently similar to those of ITP purified from locust CC by Audsley et al. (1992b, 1994). Dose-response curves for natural and synthetic ITP were not significantly different, given experimental uncertainties, and indicated that the ECs0 for stimulation of ileal l~c is 1.1 to 2.3nM. The time courses of ileal I+c at different doses were similar, as were the increase in opencircuit TEP (three-fold maximum) and the 50% decrease in transepithelial resistance. Using a different bioassay, synthetic ITP at maximum doses caused a four-fold increase in fluid transport (J,+) across everted ileal sacs, identical to that caused by purified ITP. Synthetic ITP only partially stimulated (40% of maximum) rectal I~ and there was no significant effect on rectal J,,. Thus, the effects of synthetic ITP on locust rectum are similar to those of the native peptide. Moreover, western blots probed with an antibody specific to the C-terminus of ITP indicated co-migration of synthetic and purified ITP from CC (Ring et al., 1998; Macins et al., 1999; Wang et al., 2000; Zhao, 2000). This extensive series of bioassay and antibody studies with synthetic and purified ITP provides strong support for the chemical structure of ITP deduced by Meredith et al. (1996). 4.3.4.4 A n I T P - l i k e ( I T P - L ) e D N A in locusts. Meredith et al. (1996) probed a locust ileal mRNA library with the ITP cDNA sequence primers. An ITP-L clone was sequenced that was identical to the brain cDNA for ITP except for an additional 121 bp insert at amino acid position 40 of ITP, suggesting alternative C-termini splicing of genomic DNA. ITP-L has an open reading frame (134 residues) that is four residues longer than that of ITP (Fig. 8). All six cysteines are conserved. The unique C-terminus of ITP-L shares only 14 of the last 36 amino acid residues in common with ITP+ most of the difference being over the last 20 residues. Using reverse transcription PCR (RT-PCR), ITP-L mRNA was detected in many tissues (flight muscle, hindgut and MTs) that have no stimulatory effect in the locust ileal I+c bioassay. In contrast, ITP mRNA was restricted to the brain and CC which do stimulate ileal I+~. Ring et al. (1998) have expressed IPT-L in Sf9 cells using a baculovirus vector and detected a secreted product using an antibody specific to the unique C-terminal of ITP-I. This expressed form of ITP-L had no stimulatory action on ileal l~c, but did inhibit stimulation of ileal I~ by SynITP.

INSECT DIURETICAND ANTIDIURETICHORMONES

319

95

Arg Lys Asp Cys PheThr Ser Asp Tyr Phe Lys Gly Cys lie Asp G A~

GAC TGT VI-C ACA TCT GAC TAC TTC A ~

GGA TGC ATC GAT

327 110

Val Leu Leu Leu Gin Asp Asp Met Asp Lys lie Gin Ser Trp lie G-I-I- CTA C-FI- CTC CAA GAT GAC ATG GAT AAG ATA CAG TCT TGG ATA 370 125

134

kys Gin lie His Gly Ala Glu Pro Gly Val END AAA CAA ATA CAT GGG GCA GAG CCA GGG G-I-I-TAG 415

444

ITP 56

95

56

'95

ITP-L

I 30

.... -~1.34

FIG. 8 The nucleotide and deduced amino acid sequences of the unique ITP-L insert (i.e. alternate C-terminal ending) in a cDNA open reading frame from S. ,gregaria (redrawn fi'om Meredith el al., 1996). Amino acMs identical in the C-termini of ITP-L and ITP are indicated in bold type. Numbering of residues is from the start of the prepropeptides, i.e, including a common signal peptide of 55 amino acids. Shaded boxes represent dibasic cleave sites at the start and end of the ITP and ITP-L peptides. Black boxes represent stop codons. Open boxes indicate identical sequences in ITP and ITP-L prepropeptides, while the hatched area indicates the location of the 1TP-L unique C-terminus preceding the ITP C-terminus. Alignment between the two cDNA open-reading frames is indicated by broken lines. A similar arrangement was observed by Macins et al. (1999) for L. migratoria.

4.3.4.5 Expression q / I T P . Meredith et al. (1996) cloned the complete open reading frame for the ITP propeptide in baculovirus. After several days, insect Sf9 cell cultures transfected with such baculovirus secreted a stimulant of ileal 1~ (60 maximum doses per 3 mL), whereas cells transfected with wild-type baculovirus did not. This expression of ITP occurred relatively late (on the third day) when cell cultures were beginning to deteriorate, which may explain why the expressed stimulant has a larger molecular mass than native 1TP as estimated by western blots (Ring et al., 1998). Sequencing of this expressed ITP revealed that the last 11 amino acid residues of the leader sequence were still attached to the N-terminal of the ITP peptide, indicating that processing of the ITP propeptide was different in Sf9 cells. This extended form of expressed ITP did cause full stimulation of locust ileal I~c but only at higher (270-fold)

320

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

concentrations than native ITP and with a 10-fold slower time course (Ring et al., 1998; Pfeifer et al., 1999). Pfeifer et al. (1999) designed a plasmid transfection vector with an early promoter and a Zeocin-resistance gene to permit selection of stably transformed insect ceils. Of several insect cell lines from five species that were tested, transfected Drosophila Kcl cells secreted an ITP (KcITP) with the highest biological activity and was correctly cleaved from the prepropeptide at the N-terminus. Unlike Sf9 expressed ITP, KcITP stimulated locust ileal I~ with the identical time course and maximum effect as Schgr-ITP and SynITP. Transiently transformed Kc I cells produced two forms of ITP (A, B), whereas stably transformed Kcl cells produced only one (A). The ratio of A:B forms in the former cells was consistently 3:1. Form A co-eluted with SynlTP, while form B has reduced electrophoretic mobility. Form 13 appears to be a stereoisomer because the initial 22-amino acid sequence is identical to that of A. Both A and B are also detected by an antibody to the C-terminal sequence of SynITP. Moreover, as discussed later (see section 5.5.5), single substitutions of alanine for phenylalanine at either position 2 or 3 of KclTP abolishes expression of the B form (Zhao, 2000). Stereoisomers involving enzymatic conversion of phenylalanine from the L- to D-isomer at position 3 have been reported for crustacean homologues of ITP (Soyez et al., 1994). Approximately 10-fold higher concentrations of KcITP than SynITP are required for maximum stimulation of ileal l~c because the Kcl system (like all insect and most mammalian cell expression systems tested to date) fails to process the C-terminal amidation message Leu-Gly-Lys-Lys ( L G K K ) to Lamide (Wang et al., 2000). In support, amidation of KcITP biochemically increased K c l T P specific activity near to that of SynlTP, but not completely because enzymatic amidation is only 50% efficient (Wang et al., 2000). Even in large insects such as the locust, biochemical purification of ITP is exceedingly slow and expensive, while yielding inadequate amounts of pure peptide for extensive physiological studies and antibody production. Significant progress in this field is likely to depend on good expression systems like the Kcl-plasmid vector one of Pfeifer et al. (1999), both for structure function studies and for production of ITP homologues from other insects using cDNA obtained by PCR or from genome projects. In summary, the methodology is now available to make ITP homologues from other insects for physiological studies. This represents a promising future area of research. 4.3.4.6 I T P sequence evolution among insects. ITP homologues are widespread in insects. The ileal Isc bioassay is very specific in that no other purified insect peptide or crustacean ITP homologue tested to date stimulates it. Positive locust ileal l~c bioassays and western blots probed with Schgr-lTP antibodies indicate homologues very similar to Schgr-ITP in brain-CC of all orthopteran and closely related groups (e.g. cockroaches, stick insects) tested to date. The ITP sequence in L. m~ratoria is identical to that of S. gregaria

INSECT DIURETIC AND ANTIDIURETIC HORMONES

321

(Macins et al., 1999). Slower rates of ileal I~c increase with brain-CC extracts from cockroaches and crickets suggest some sequence change from locusts. In contrast to all orthopteroid (i.e. exopterygotes) insects that have been tested, the brain-CC extracts of all but one endopterygote species tested (from Orders: Diptera, Lepidoptera, Hymenoptera, Hemiptera) failed to stimulate locust ileal I~ or show a band of correct M,- (8.7 kDa) on western blots probed with SchgrITP antibodies (Meredith et al., 1996; Macins el al., 1999; Phillips et al., 2001). However, Endo et al. (2000) recently obtained a eDNA from the silkworm ( B o m b y x mori) that encodes a lepidopteran ITP that has 63% sequence identity with Schgr-ITP, although this peptide has not yet been produced and bioassayed. Phillips et al. (2001) also reported a very different ITP homologue in Diptera using information from the fruit-fly genome (Adams et al,, 2000). A large putative gene (CG13586) near the tip (60D13) of the right arm on chromosome 2 encodes the N-terminus and C-terminus of a putative ITP homologue separately in exons 9 and 11, with 2 introns and exon 10 (i.e. 1.2 kb) interposed. Phillips et al. (200 I) used primers based on these nucleotide sequences and RT-PCR on total RNA from D. mehmogaster brain-CC to obtain a processing intermediate in which exons 9 and 11 are spliced together following the cysteine at residue 40 (i.e. equivalent to Cys > in Schgr-ITP) to yield a deduced prepro Drome-ITP sequence (108 residues). The deduced amino acid sequence for the fully processed ITP homologue (Drome-ITP; 73 residues) is compared with Schgr-lTP/Locmi-lTP and Bommo-lTP in Fig. 9. All these insect ITP homologues share several common features. Dibasic cleavage sites at both ends (not shown at N-termini) and amidation signals

splice site Bonm~o-IT? S-FFTLECKG V F D A A I F A R L DRICDDCFNL FREPQLYTLC 1

i0

20

30

S c h g r - I T P S - F F D I Q C K G V Y D K S I F A R L DRICEDCYNL FREPQLHSLC D r o m e - I T P SNFFDLECKG IFNKTMFFRL DRICEDCYQL FRZTSIHRLC

40

50

60

B o m m o - I T P R A E C F T T P Y F KGCHESLYLY DEKEQIDQMI

70

DFV-amide

~R~

S c h g r - I T P RSDCFKSPYF KGCLQALLLI DEEEKENQi~" Z I L - a m i d e

(GKK)

D r o m e - I T P K Q Z C F G S P F F NACIEALQL}~ E E M D K Y N E W R D T L - a m i d e

(GRK)

FIG. 9 Deduced 1TP amino acid sequences for B. mori (Bommo-lTP) S. gregaria/L mi~ratoria (Schgr-ITP) and D. melano~aster (Drome-lTP). The asterisks indicate residues different from Schgr-ITP.

322

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

(GKR, G K K , GRK) predict processed peptides that are 72 or 73 residues long, with a terminally amidated hydrophobic amino acid (Leu or Val). The latter is essential for receptor stimulation (see section 5.5.5; Wang eta/., 2000) and is also common to 12 crustacean CHH homologues (Lacombe et al., 1999). The insect ITP homologues in Fig. 9 share the sequence SFF at the N-termini although Drome-ITP contains an Ash insert Ser-Asn-Phe-Phe (SNFF). A phenylalanine at position 3 is common to nearly all CHH homologues. Zhao (2000) has shown that Phe 2 and Phe 3 in Schgr-ITP are essential for receptor activation and are also involved in binding (see section 5.5.5). The six cysteines in each of these ITP homologues are conserved relative to all crustacean members of this neuropeptide family (Lacombe et al., 1999). All three ITP homologues share 40% sequence identity, with three-quarters of this in the N-termini preceding the splice site at position 40 of Drome-lTP (39 in Schgr-ITP and Bommo-lTP), The computer program of Ito et al. (1997) predicts that the Schgr-ITP N-terminus consists of two short /~ strands (residues 2 4 and 36 39), an ot-helix (residues 13 30), and two short regions of random coil (residues 6 12 and 31 34). In Schgr-ITP, the C-terminus is predicted to consist of a terminal a-helix (residues 53 72) after the last cysteine with the remainder (residues 40 52) as random coil. The ITP sequence divergence between the endopterygote fruit fly and silk moth (48% identical) is in fact larger than between them (52% and 63% identity, respectively) and the exopterygote locusts. Only six sequence changes (residues 5, 6, 11, 42, 54, 70; three of them neutral) from Schgr-ITP are common to B. mori and D. melanogaster. Moreover, ileal l,~cbioassays and western Nots probed with Schgr-ITP antibody both indicate an ITP homologue in rose beetle (Pachnoda simuata. Scarabaeoideae) brain-CC indistinguishable from that of locusts (data in Phillips eta/., 2001). This indicates great ITP sequence divergence between coleopterans and both dipterans and lepidopterans. In a recent (1999) US Patent (US 5989861-A), P. Lal, N. C. Corley, K. J. Guegler and C. Patterson report the identification of a eDNA from a human breast tissue library that encodes a human Schgr-lTP homologue (HITLP: AAE37516). Their deduced sequence for this human protein is 100% identical to Drome-ITP. This is indeed a remarkable coincidence given the great ITP and CHH sequence diversity observed within arthropods as documented above. Clearly confirmation of this obserwttion on a human ITP homologue in a peer-reviewed paper is needed. The recently deduced amino acid sequences of signal peptides in D. mehmoj~aster and B. mori (Endo et al., 2000; Phillips et al., 2001) preproITP homologues have negligible sequence identity to those of locusts (Fig. 10), which share an 80% identical sequence (Macins et al., 1999). Nevertheless, all have the requisites for signal sequences as predicted by the computer program SPScan (Nielsen et al., 1997), including terminal dibasic cleavage sites (KR or RR). Moreover, Drosophila Kcl cells express and secrete Schgr-lTP (only lacking C-terminal amidation) when transfected with the locust preproITP

INSECT DIURETIC AND ANTIDIURETIC HORMONES

323

Drome-ITP

MCSRNIKISV

VLFLVLIPIF

AALPHNHNLS

KR

Bommo-ITP

MHLSSVQFAW

AALVALAVSA

AGALPSSAPH

HVERR

Schgr-ITP

MHHQKQQQQQ

KQQGEAPCRH

LQWRLSGWL

CVLWASLVS

TAASSPLDPH

HLAKR

HG. l0 Deduced signal peptide amino acid sequences of the D. me/anogaster, B. moll and £'. gregaria ITP prepropeptides.

message (Pfeifer et al., 1999), indicating that Kcl cells can recognise and utilise quite different insect signal sequences. Evidence from both S. gregaria and D. melanogaster (i.e. evolutionary distant insect species; Phillips et al., 2001) indicates that N- and C- terminal sequences of preprolTP peptides are coded in separate genomic messages (e.g. exons 9 and 11 for DroslTP) that are spliced together in a processing intermediate as indicated by the vertical line in Fig. 9. The genomic message in both cases encodes a signal sequence ending in a dibasic cleavage site preceding the ITP N-terminus (residues 1-39/40). A second genomic message codes the ITP C-terminus (residues 40/'41 to 72/73 amide). In both species, an alternative form of C-terminus is observed, namely in ITP-L of locusts (Ring et al., 1998; Macins el al., 1999) and in putative exon 10 of the D. melmtogaster gene (Fig. l 1). The genomic coding of the alternative (ITP-L) C-terminal sequence occurs before that of the 1TP C-terminal sequence in locusts (Fig. 8: Ring et al., 1998). This arrangement appears to be conserved in CG13586 where an alternate C-terminal sequence (ITP-L) coded in exon l0 precedes the C-terminal sequence of ITP in exon 11 (Phillips et al., 2001). A cDNA obtained by RT PCR confirms the presence of a ITP-L processing intermediate in brains of both larval and adult D, melanogaster in which exons 9 (Fig. 9) and 10 (Fig. 11) sequences are spliced together, with an additional eight residue sequence (positions 80-86 in Fig. 11) of unknown genomic source added at the end (J. Meredith, unpublished observation). In both locusts and D. mehmogaster, the two cysteines of the alternate C-terminal sequence are located four and 13 residues past the splice site (Fig. 11), i.e. at a location common to all other arthropod members of this protein family. There is 38% sequence identity for the first 21 residues of C-terminal sequences in these two insect ITP-Ls. We

Drome-ITP

40 KANCFVHET_F

Schgr-ITP

RK~TSDYF

50 GDCLI
60 70 D__EEISQLQHY LKVINGSPYP

80 FHKPIYH*

KG~ID~.LLQ DDMDKIQSWI KQIHGAt~GV*

H G . I 1 Deduced alternate (ITP-L) C-terminal sequences of D. melam~j~aster and S. gregaria. C o m m o n amino acid residues are in bold and underlined. The asterisks indicate stop codons.

324

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

propose that the shared N-termini of these alternate forms (ITP and ITP-L) permit ITP-L peptides to bind the ITP receptor (e.g. involving the required Nterminal FF sequence), while the absence of a final u-helix (residues 53 72 of Scbgr-ITP), and of terminal amidation in ITP-L, block receptor activation, thereby explaining the observed antagonistic action of Schgr-lTP-L discussed earlier (see also section 8.4). If the function of the last 20 residues in ITP-L is simply to disrupt the specific activation topography of ITP C-termini, evolution of very different ITP-L terminal sequences in locusts and flies (Fig. 11) is not surprising.

5

Cellular actions

5.1

INTRODUCTION

Ion transport has been extensively investigated in a limited number of insects using in vitro preparations of MTs and hindgut (ileum and rectum), which has allowed the cellular activities of diuretic and antidiuretic factors to be defined. Here we review what is known of the receptors, signalling pathways, and effects on M T and hindgut transport, for factors identified as having diuretic and/or antidiuretic activity. 5.2

5.2. l

REGULATION OF MALPIGHIAN TUBULE SECRETION

Serotonin (5-hydroxytt?'ptamine," 5-HT)

Serotonin stimulates secretion by MTs from many insects and was initially thought to be a DH mimic, but is now known to be a functional DH in R. prolixus (Maddrell et al., 1991) and in A. aegypti larvae (Clark and Bradley, 1997). It stimulates secretion by R. prolixus distal tubules up to 1000-fold via a cyclic AMP-dependent mechanism (ECs0 30-40nM; Maddrell et a/., 1993), whereas the response of L. migratoria tubules is only 25% of the maximum obtained with a CC extract and is mediated by a different second messenger, possibly Ca 2+ (ECs0 10 -100riM; Morgan and Mordue, 1984). The response of A. aegypti tubules to serotonin varies with the stage of development. In larvae, it acts via cyclic AMP to stimulate maximal secretion (ECs0 c. 100nM; Clark and Bradley, 1996; Clark and Bradley, 1998), whereas the response of adult tubules is c.25% maximum (ECso I # M ) and both cyclic AMP and inositol trisphosphate (IP3) production are increased (Veenstra, 1988: Cady and Hagedorn, 1999b). 5.2.1.1 Receptors. Serotonin activity is blocked by the vertebrate 5-HT2 receptor antagonists ketanserin (R. prolixus: Maddrell et al., 1991) and

INSECT DIURETIC AND ANTIDIURETIC HORMONES

325

methiothepin (A. aegypti; Clark and Bradley, 1997). However, 5-HT2 receptors couple to Gq and activate phospholipase C (PLC) rather than adenylate cyclase. Of the serotonin receptors cloned from D. melanogaster (Colas el ell., 1997), only the 5-HT7Dro receptor couples positively to adenylate cyclase. However, although Pietrantonio el al. (2001) have cloned a 5-HTvD,.o-like receptor from an A. aeg),pti M T cDNA library, it is expressed in tracheolar cells and not in the tubule epithelium. 5.2.1.2 Mode qfaction. Serotonin stimulates secretion by R. prolixus distal tubule up to 1000-fold, and the Na + K + ratio of the secreted fluid increases from 0.7 to 1.8 (Mad&ell, 1969). In contrast, it stimulates KC1 reabsorption from the proximal tubule and lowers the osmotic permeability (Maddrell and Phillips, 1975; O'Donnell el ell., 1982). As a result, the Na + K + ratio of the secreted fluid increases to c.40 in the proximal tubule and its osmotic concentration falls to as low as 75mOsm (Maddrell and Phillips, 1975). Stimulation of distal tubule secretion with either serotonin or exogenous cyclic AMP is accompanied by a triphasic change in the TEP, which first goes more negative and then hyperpolarises before going negative again. These changes are attributed to sequential activation of (1) an apical CI-conductance; (2) the apical V-ATPase: and (3) a basolateral Na+/K+/2C1 cotransporter (Fig. 12A: O'Donnell and Maddrell, 1984; lanowski and O'Donnell, 2001; M. J. O'Donnell, personal communication), lanowski el al. (2002) made use of double-barrelled ion-sensitive microelectrodes to measure the electrochemical gradients for Na +, K + and CI across the principal cell basolateral membrane, and confirmed the central role of a bumetanidesensitive Na+/K+/2CI cotransporter in supporting serotonin-stimulated secretion. A Na+/K+/2C1 cotransporter has been cloned from M. se.vta tubules (Reagan, 1995b) and a similar transporter (53% sequence identity) is encoded in the D. mehmogaster genome (CG2509) on the right arm (83A5 6) of chromosome 3, which suggests they are widespread in insects. The cloned transporter shares between 42% and 46% sequence identity with Na+/K+/ 2C1 cotransporters from the shark, rat, rabbit and mouse and, significantly, has a conserved protein kinase A (PKA) phosphorylation site (Reagan, 1995b). Serotonin has also been shown to inhibit the Na+/K + ATPase activity of R. proli.vus tubule homogenates (Grieco and Lopes, 1997), but this is unlikely to contribute to its diuretic activity, because ouabain has no effect on secretion by serotonin-stimulated tubules (lanowski er al., 2002). Fluid secretion by MTs from larval A. aeg37)ti is stimulated four-fold by serotonin and exogenous cyclic AMP (Clark and Bradley, 1998), but the Na * K + ratio of the secreted fluid is unchanged (Clark and Bradley, 1996}. This contrasts with the marked natriuretic effect of cyclic AMP in adult tubules (Williams and Beyenbach, 1983; see section 5.2.3.2) and is consistent with the differing requirements for Na + homeostasis in larval and adult mosquitoes; fl'eshwater larvae must conserve Na +, whereas adult females need to excrete

326

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

A

B -"'"4,

H"

Na*

K" c,-

Na*m.---..

K*

"

2d_

....

seroton in

I-I*

CI-

Cl

.t

cAMP

C a 2*

CRF -related diuretic peptide

FIG. 12 A Proposed model for the action of serotonin on R. prolixus distal tubule (based upon Ianowski and O'Donnell, 2001). Serotonin acts via cyclic AMP to stimulate transepithelial secretion of NaCI and KC1 by activating an apical C1conductance, the V-ATPase and the basolateral Na~/K+/2C1 cotransporter. Serotonin also inhibits the basolateral Na +/K + ATPase, although this is unlikely to contribute to its diuretic activity. B Composite model showing how CRF-related diuretic peptides can act via cyclic AMP to: (1) stimulate the apical V-ATPase (Tenmo-DH); (2) open a basolateral Na" conductance (Culsa-DP); or (3) activate basolateral Na+/K+/2CI cotransport (Achdo-DP). Additionally, Culsa-DP and Achdo-DP open a Ca 2~-activated C1 conductance (4) that may be transcellular or paracellular. The symbols have the same meaning as in Fig. 2.

excess Na + from the blood meal. The change from the larval to the adult phenotype appears to be associated with the appearance of a cyclic AMPdependent, bumetanide-sensitive Na+/K+/2CI - cotransporter in adult tubules (Hegarty et al., 1991), which is absent in larvae (Clark and Bradley, 1996).

5.2.2

Arginine vasopressin-like insect diuretic hormone (A VP I D H )

AVP IDH stimulates cyclic A M P production and has a small effect on fluid secretion by L. migratoria MTs attached to a short segment of the gut (Proux et al., 1987, 1988; Proux and Herault, 1988). However, it has no effect on secretion or cyclic A M P production by fully isolated tubules, although they do respond to Locmi-DH (Coast et al., 1993). The preparation used by Coast et al. (1993) did not include the MT ampullae in which there are endocrine cells containing Locmi-DH (Montuenga et al., 1996; see section 6.1.4). Conceivably AVP IDH acts by releasing Locmi-DH from these cells to stimulate tubule secretion and cyclic A M P production, but this has not been investigated.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

327

5.2.3 CRF-related diuretic peptides Using conspecific assays, CRF-related peptides have been shown to stimulate secretion by MTs from A. domesticus (Coast and Kay, 1994), L. migratoria (Patel et al., 1995), D. punctam (Furuya et al., 2000b), M. domestica (Iaboni et al., 1998), T. molitor (Wiehart et al., 2002) and M. sexta (Audsley et al., 1993, 1995; Blackburn and Ma, 1994) at low nanomolar concentrations. Diuretic activity varies considerably, from a maximum response in A. domesticus (Coast and Kay, 1994), equivalent to that obtained with a CC extract, to <25% maximum in M. domestica (laboni et al., 1998), and probably depends upon whether anion or cation transport is the rate-limiting step in tubule secretion. 5.2.3.1 Receptors. Reagan (1994, 1996) cloned receptors for Manse-DH (Manse-DHR) and Achdo-DP (Achdo-DPR) and expressed them in COS-7 cells where they bind their respective ligands with high affinity and couple positively to adenylate cyclase. The Manse-DHR has broader specificity than the Achdo-DPR (Reagan, 1994: Reagan, 1996), which is consistent with data from fluid secretion and second messenger assays (Coast et al., 1994: Audsley et al., 1995). Thus, Manse-DH, Manse-DPlI, Peram-DP, Achdo-DP and Locmi-DH all bind the Manse-DHR with high affinity (ICs0s 1 12nM: Reagan, 1994), and are potent stimulants of cyclic AMP production by transfected COS-7 cells (ECs0s 0.5 5 nM). Since these peptides vary in length and have only five residues in common (see Fig. 4), secondary structure rather than linear sequence must determine biological activity. The Manse-DHR and Achdo-DPR are 51% identical and have between 41% and 48% sequence identity with the products of two D. melanogaster genes (CG8422 and CGI2370; see Fig. 13) localised at 5 IA4 and 49B2, respectively, on chromosome 2R, which appear to be paralogues (Hewes and Taghert, 2001). The insect receptors are 29 37% identical to vertebrate CRFI and CRF2 receptors, and belong to Family B (the secretin-like receptors) Group I of GPCRs. Conserved regions lie close to or within the seven transmembrane (TM) domains (see Fig. 13), and all have five Cys residues in the extracellular N-terminus, which are important for ligand binding and activation of CRFI receptors (Spiess et al., 1998). A well-conserved region in the middle of the N-terminus (see Fig. 14) corresponds with the ligand selective domain of Xenopus laevis CRFI receptors (Spiess et al., 1998). The eight invariant residues within this region may form a scaft\~ld holding the five amino acids identified as being critical for selectivity (see Fig. 14) in an appropriate geometry for ligand binding. The third intracellular loop, which is implicated in G protein binding (Dautzenberg et al., 2001), is 100% conserved in CRF receptors, whereas over the same region the insect receptors share 72 86% sequence identity and are 50 59% identical to the vertebrate sequence.

328

G. M . C O A S T ,

I. O R C H A R D ,

Achdc DER Manse DHR

MEDAWEADA

Drome CRFRCG123';'(

.

X e n l a CRFR;;

- - - - - - - - - - - - - - - - - - -,.wo.,.u

...........

Manse-DHR

D=ome CRFRCG8422 Drolqe CRFRCG]231( xenla CRFR1 Xenla CRFR2

.

.

.

.

J. E. P H I L L I P S A N D

LEAAEAAAAAAQAVAADAP .

.

.

.

.

.

.

.

.

.

.

.

.

TTTPSPRTLQRORACEA .

.

.

.

.

.

F Y L S F

N L S G ND LYR N F t G FE

.

.

~M~Y S ~ _R F HA F FP E HL

......... Man~e DHR

%A I _E]I}F L YY F K V S L V VF ~

..................

. , ,uu

.

.

.

.

N~-~ST . D I¥, ~ V T . H,Cl ~ R T~ -- --- AI r ~ L F ~ E A E r ~ D E D A E K TN DA A H t P pIFISIVIP E F ~ , V I V E DH SD DR HQNSG I PVMpD FISIP N V E A GRGD AQ QE LKQEKK TKVHYH

LRCLRNTIHT

~ V~F~c~.~,.~

I GW G~L,P

V

W.~ I IF~H~G

K TA

~', ~ ~ , ~ , H . ~ . ~ . .

CS

~

~

L Mr~G D A MA[~JE C L .

.

LED

.

.

F

M ceT~°~

D

FLHSES

A NI~EV G~H~ C SIlL ~I L P T Y I T L .E. A AG F I I NFL HS I

V

N

V V A C

Yv EN

VL

NWSDE

~ sv

Q

,s,~

L~IF] ,I ,L]R N L L T F L V I TTE T F R N V T L~ F ~r ~ M . ~ Q T L S H E A H~IS I.~ TI-l.lrl-~--~lm ~ I M LQMID HN IH H4 G L Y L ~ A ~ G [ ~ . ] K V r ~ L Q~-/'/~I I GLYLYM E A N LK TT K G N FN AS G LYL L~LT.I-.~VlQIT~E.SISIDINJ_~F G.i~Y I I I YIA L G Y L T V H T I Y I S ITID~K~L RIK W M F I C ICL~JHHL~JAA : W M ITIyI£1 TIDI R IIIK KHW5V F L F

DNAGESH

~.E

.

TDS

MS

~LLIVIvV AILUJ IIF LLISRLb ~ J T R C L R N~T~ I H F L F L F R C L R N,I IiII ,II~ H~.,I~ . . ~ LI I A I I V I I A I F L LIF LIC L R S I I R C L R N I H3 . . . . . . . . . . . . L~]S V[~AE~ L M~:~F Y L Z FWMF~V E M . . . . . DHR QDQTS MI V CM F...y. L T N F F W M VE Drone CRFRCC8422 V~S - I AILII~T~L F H ~ F I T I L T FWM VE S,, . . . . . . . . . . . . . n QAG--r~l T I I I V l I I M FJ~IY F Y ~ . T N F FWM Fr~E ............ N V V WlCIR L V T I I I A HIN Y F YIV~T FW M F E ............ N E V -WlCIR C I T T I yIN Y FIV v I T N FE F W M . . . .

.

VIQ KIC.ilILIE I L YI G I IQIY DIDITIQIN AIS V I L ~ I C U i ~ I ~ E I L QIG.J.IHIY D ~ K I N A T[ VILIP C F I E I E I F K I G I V H I Y ~ T T I D I N A TI A RIP C l P ~ Y F HIGIV QIYIN ] T TIGINIV Y E RIP CIpIDIS F NIG IIRIYINIT TIRINIV y

. .

Ez .................. ............. ...........

Achdo DHR

ro~,.~

D N YICIPIAIY I F DIG.L. -I L C W I D P I TIP W j ~ T I L A S . ~ H I C I L T QI F DISI II -[L C W P R TIA.~ilGITIL A I I / I Q I C I P £ SIF DIS V -II C W P J i . T I N I A G I S I L A GILIAICINIAIS I IDIMIIIG TIC W PI~LTJAIA GIQ.t&V G P yICISIAIT IIDIQI IIG TIC, W P RIS LIA GIE~LIV

. . .. .. ........... . Drome CR]2RCG8422 Orome CRFRCGI 237( X e n l a CRFRI

rJrcme CRF RCGI237[ Xenla CRFRI ...........

.

----------------,:~;;~;::~:,-,:~,-L~X~D~

A R K A S T KAS NS S F

D. A . S C H O O L E Y

M V L L I K HF~P~IA~D~]

: : : : :.~E * ~G, L F N

F~

O.

~ , R CWFGKKAGV T C W i G K E PGKIYYIIIDD~FL~ j

G L YDNE . . . . . . . . [I CG~WW]~IllPP[~pP[; I V T[~WA A IIc~KKLr~y E N E K . . . . . . . .

H6

Achdo DHR

Drome CRFRCS8422 brome CRFRCGi237

Hr~A~TETClA HKA IYOGP F K V

............... .......... ......... Drome CRFRCG8422

LV C

G ~

Q

~ ~

~

,

M

~

F

c-

T A S

S

,

v

R AN7 N YA~O~] t[~]Lk ~ ~ TT ~A A ~ H F ~ T G H MIF~A V LIR A I V I L ~ . S T O G FISIVIS[L F Y C F L N S E V R]NIAI

ViG H S WFA

s S

A

L

E R R

TV

L FFINLUVRR" N ~ h K L

F FV T V

Xei11a CRFRI CRFR2

MWV ~ , TK L ~ J ~ b i ~ 7 1 Q r s ~ m - ~ ] T ~ RIMWVLI TKLRSANTL R I MWV L I TK LRSA I,~L~I~T K L

~

Xenla

.......... MaSSe EHR Drome CRFRCS84EE Drome CRFRC:]t237: Xenla C2FRI Zenla CRFR2

V A

SGL

G G

E I DV

R ] Q I

I I

FN S F

S V S

~---'~H

M S~I_~T TF~G G G R Rr~t L~]G H~K Dwsr~R[~]E VE G GGGRR GA Y KDWS R I S._TJWRIOITIR,..~IIQ L N Q N I R I R Y T T K I S I F S K G G G FT S HRNSS vh~ ~ AKMN ~ R- W A N- - o s ~, T HI ~ K ~s"~ R A R V A AIRIKUtJWHIRWlQIDIH H S L R V R V A I R I A MISI . . . . I T

S S

S I RC LQHR[ . . . . ESJ_R LTVRS DLVRD I AE RP L TSYYGRGKR t :: R. ' HS- [ K Q S S A -, R I HS I KQTAAV - -

ILRHIH

DHR .Achdo ..... ~.~

:::::::::::::::::::::::::::::::::::::::::::::::::::

Drome CRFRCG8422 brott e CRFRCGI237( X e n l a CRFR1 Xen!a CRF[42

ESCVSSA TTTTLVGQHAP LS LHRGS NNA LH TMP TLAANAMSSGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

TLSVMPR .

.

FIG. 13 The aligned sequences of CRF receptors from X. laevis (Xenla-CRFRI and CRFR2), M. sexta Manse-DH (Manse-DHR), A. domesticus Achdo-DP (AchdoDPR) and two putative receptors (CG8422 and CG12370) from D. melanogaster, The alignment was performed with Clustal W, and residues identical in 50% of the sequences are boxed. Bars over the protein sequences indicate the seven putative transmembrane domains of the Manse-DH receptor. The C-terminal 15 residues of CG8422 are not shown.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

Tupbe-R2B Homsa-R2 Musmu-R2

329

R'N A Y R E C LIEINNGTV~ G T V~ i- -GI E ¥ 7 NI KI- ]A W I i i ~ p p I Y F NIG VIKIY NIAIT R N A Y R E C L E NNGGTTW

TW GTW Y F N GF~JlVJKKYN { { R A Y R E C L E N G N R N A Y R E C L EIN NG GTT~~ YFN NG W R Nr-V]YRE C[-FEIN GTTW X e n l a R2 S F i i U _ J v V i i i i i N N I G I Y R E C L A I NN G W G TTW N G-IW Oviar-Rl L V F F s Nl lv A,N Tupbe-Rl F F YIG v R Y N T T N NIGIY R E C L AIN N GISIW GISIV~ N GISIW Ratno-Rl F F YIG v R Y N T T N NIGIY R E C L AIN N GISIW GISIV~ Homsa-Rl F F YIG v R Y N T T N NIGIY R E C L AIN Galga R1 Y F Y G V R Y N T T N NIGIY R E C L AIN N GISIW GISIV~ Xenla-Rl Y F H G VI-Q-IYN T T 6 NlvlY R E cFH LIN N GlSlW NG Dmome CG123V) I P C I F E E F K GVr_~YI-~_._I. D NA-~7RI-~CIF PIN GTZWW TW A c h d o DP R L~]QCIFF D E L NIGI I!_RYI D I R c a AITIRIMIClY SINN G GTW Drome-CG8422 D E L QIGII HIYID S S K N AITIRIFICIH AINN G W G TTV~ Manse-DH-R KICI F K E L YIGII QLY_JDDF" C NAISIRILIC L D--"~V---[_~_ - ~ v-!~_

lili

FIG. 14 Sequence alignments of a conserved 27-amino acid region in the middle of the extracellular N-terminal domain of insect and vertebrate CRF receptors. This region corresponds with the ligand-binding site in X. laevis CRFI receptors and critical residues for the binding of CRFI are shown in bold. The alignment was performed with Clustal W and identical residues are boxed. Abbreviations: Tupbe, Tupaia helangeri; Homsa, Homo sapiens; Musmu, Mus musculus; Xenla, Xem~lmS laevis: Bosta, Bos taurus: Oviar, Ovis aries; Ratno, Rattus norve~icus; Galga, Gallus gallus.

5.2.3.2 Mode ol'Action. CRF-related diuretic peptides stimulate cyclic A M P production by isolated MTs (see section 4.2.2.1) and, in A. domesticus, intracellular levels of the cyclic nucleotide increase prior to any change in ion transport or fluid secretion, which is consistent with a second messenger role (Coast and Kay, 1994). Exogenous cyclic A M P has been shown to stimulate cation transport by activating the apical V-ATPase and/or by increasing cation entry across the basolateral membrane of principal cells via a conductance pathway or cation/C1 cotransport (reviewed by Nicolson, 1993: Beyenbach, 1995), and it is likely that CRF-related diuretic peptides act in a similar manner (see Fig. 12B). It is worth remembering that an increase in V-ATPase activity will not influence the Na + K + ratio of the secreted fluid (see section 2.2), but that this can change following stimulation of cation entry across the basolateral membrane. Exogenous cyclic A M P stimulates the activity of the apical V-ATPasc in D. melanogaster tubules, which hyperpolarises the apical membrane potential I~, and increases the proton gradient driving cation transport into the lumen via cation (K +, Na+)/H + antiports (O'Donnell et al., 1996). Potassium, the major cation transported, enters principal cells via an electroneutral K+/C1 cotransporter driven by the favourable electrochemical gradient for C1 uptake (Linton and O'Donnell, 1999). Recent work (U. I. M. Wiehart, S. W. Nicolson

330

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

and E. van Kerkhove, personal communication) shows that Tenmo-DH37 acts via cyclic A M P (see section 4.2.2.2) to stimulate V-ATPase activity in the free portion of mealworm MTs, although in normal saline it has no effect on (V,) because events at the apical membrane are masked by rapid K + entry through channels in the basolateral membrane (Weltens et al., 1992). Stimulation of the apical V-ATPase is revealed when these channels are blocked with Ba 2+, and both V, and the basolateral membrane potential (Vb) then hyperpolarise in response to Tenmo-DH37 (and exogenous cyclic AMP) as increased proton transport into the lumen makes the cell interior more negative (U. I. M. Wiehart, S. W. Nicolson and E. van Kerkhove, personal communication). How cyclic A M P activates the V-ATPase is unclear, however, because there is no evidence for the p u m p being phosphorylated by protein Kinase A (PKA) (Wieczorek et al., 2000). Exogenous cyclic A M P stimulates secretion of NaCI-rich urine in adult A. aegypti tubules, and the Na + K + ratio of the secreted fluid increases from unity to c. 10 (reviewed by Beyenbach, 1995). This is attributed to the activation of a basolateral membrane Na + conductance (Sawyer and Beyenbach, 1985), which makes more Na + available for transport into the lumen. The additional Na + is accompanied by C1 , which moves through a shunt pathway into the lumen to preserve electrical neutrality. The apical membrane potential does not change significantly, and there is no evidence for stimulation of the V-ATPase, although this cannot be excluded. Sodium entry through a basolateral membrane conductance (channel) is believed to be the rate-limiting step in Na + transport and can be inhibited by 0.1 and 1 mM amiloride, an epithelial Na + channel (ENaC) blocker (Hegarty el al., 1992; Beyenbach and Masia, 2002). However, at these concentrations amiloride blocks a variety of transporters and Beyenbach and Masia (2002) were unable to exclude effects on a Na+-dependent, electrogenic transport process e.g. a Na+/amino acid cotransporter. The tubule expression of amiloride-sensitive ENaC has not been investigated in A. aegypti, but these channels are not expressed in D. melanogaster tubules (Giannakou and Dow, 2001). Importantly, Hegarty et al. (1991) showed that bumetanide is able to block the diuretic and natriuretic activity of exogenous cyclic AMP, but does not prevent changes in tubule electrophysiology associated with activation of the basolateral Na ~ conductance. Thus activation of the conductance pathway alone cannot account for the stimulation of NaC1 transport and fluid secretion. In contrast to the actions of exogenous cyclic AMP, the CRF-related peptide from C. salinarius (Culsa-DP) has little effect on Na + transport or fluid secretion by A. aegypti tubules (Clark el al., 1998b), but this is not unusual in a cross-species assay (Coast et al., 1994; Audsley et al., 1995). However, 100rim Culsa-DP causes a biphasic change in the TEP of perfused tubules (Clark and Bradley, 1998), which the authors attribute to the sequential opening of: (1) a Ca2+-dependent CI shunt, and (2) the cyclic AMP-activated basolateral Na + conductance (see Fig. 12B). Culsa-DP might therefore activate

INSECT DIURETIC AND ANTIDIURETIC HORMONES

331

both second messenger pathways, as suggested for some vertebrate CRFI receptors, which can couple to G~ and Gq (see Dautzenberg and Hauger, 2002). Likewise, the diuretic activity of Achdo-DP has been attributed to the stimulation of cation transport by cyclic AMP and the opening of a C1conductance pathway via a Ca2+-dependent mechanism (see Fig. 12B; Coast and Kay, 1994). Further research (G. M. Coast, unpublished observations) suggests that Achdo-DP stimulates secretion by activating a cyclic AMPdependent Na+/K+/2C1 cotransporter, because all three ions are needed in the bathing fluid (see Fig. 15A), and the concentrations of K - and Na+ required for half-maximal stimulation (4 and 30mM, respectively; see Fig. 15B, C) are comparable with the ion requirements of the human Na+/K+/ 2C1 cotransporter (Payne and Forbush, 1995). As a result, the Na+ K + ratio of the secreted fluid increases from 0.4 to 1.2 after stimulation with Achdo-DP (G. M. Coast, unpublished observation).

5.2.4

Diuretic/myolropic kinin neurope7)tides

In conspecific assays, kinins stimulate secretion by MTs of A. domeslicus (Coast el al., 1990a), L. m~watoria (Thompson et al., 1995), M. domestica (Holman et ell., 1999), and D. mehmogaster (Terhzaz et al., 1999) at nanomolar or sub-nanomolar concentrations. Aedae-K-1 and -3 are somewhat less potent, while Aedae-K-2 has no effect on fluid secretion (Veenstra et al., 1997), but stimulates IP3 production by A. aegypti tubules (Cady and Hagedorn, 1999b). Diuretic activity varies considerably, from a maximum response in M. domestica, equivalent to that obtained with an extract of the thoracico-abdominal ganglion (Iaboni et al., 1998: Holman et al., 1999), to c.25% maximum in A. domesticus (Coast el a/., 1990a). 5.2.4.1 Receptors. Ligand-binding studies show kinin receptors are present in MT plasma membranes (Chung et al., 1995), and proteins of 53 57kDa have been labelled in mosquito and cricket tubules using a photoaffinity kinin analogue (Pietrantonio el ell., 2000: J.-S. Chung, T. K. Hayes, A. Strey, G. J. Goldsworthy and G. M. Coast, unpublished observations). Specific binding to a 54 kDa protein in A. aegypli membranes is saturable, with a Kd of 13 nM and a Bm,,x of c.2 p m o l m g I membrane protein (Pietrantonio el al., 2000). Cox et al. (1997) cloned a kinin receptor (Lymst-KR) from a brain cDNA library of the pond snail Lymnaea stagnalis, which when expressed in CHO-K1 cells could be activated by an endogenous kinin (Lymst-K) and by Leuma-Ks IV, V and VI, as evidenced by an increase in cytosolic Ca 2+. The Lymst-KR is 41% identical to a putative kinin receptor from the Southern cattle tick Boophilus microplus (Boomi-KR; Holmes et a/., 2000), and using these sequences as queries in BLAST searches of the BDGP identifies a gene encoding a putative Drome-KR (CG 10626) localised at 64D2 on chromosome

332

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

,,0 A A

"7r -

\\\\

.m

,N\'~

E 500 -

"7

E E

\\\\ \\\\

r-

250 -

.£

\.\\.\

,,\-..-.-.,.,

O

,x.\-~

o

Controls

Na+-free

K+-free

C

B ¢.-

CI--free

100-

100-

75-

75-

so-

50-

25-

25-

0-

0-

.o

U)

E

5

0

10

20

[K +1 (mM)

30

0

30

60

90

120

[Na +1 (mM)

FIG. 15 A The effect of omitting Na +, K - or CI from the bathing fluid on basal (open bars) and Achdo-DP-stimulated (50nM: hatched bars) secretion by A. domesticlts tubules. Vertical lines represent ±S.E (standard error). All three ions are needed for Achdo-DP to stimulate secretion. B and C The effect of varying bathing fluid [K ~] (with 100mM Na~; B) and [Na +] (with 9.3mM K+: C) on the diuretic activity of 50 nM Achdo-DP. Symbols represent means 4-S.E. The concentrations of K + and Na ' required for half maximal stimulation of tubule secretion are 4mM and 30 mM, respectively. G. M. Coast unpublished data.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

333

3L. The D r o m e - K R is 46% identical to the B o o m i - K R and 36% identical to the Lymst-KR (see Fig. 16), and they are related to neurokinin receptors in Family A Group 111 of the GPCR superfamily (Hewes and Taghert, 2001). The greatest similarity is over or near to the seven TM domains and in the short extracellular loop between TM2 and TM3 (64% identical; Fig. 16). 5.2.4.2 Mode o/actio#~. Kinins have been shown to open a Ca2+-activated anion conductance, which allows more CI- into the tubule lumen and hence depolarises the TEP (Hayes et al., 1989; Pannabecker et al., 1993; O'Donnell el o/., 1996). With more C1 available, additional Na + and K + can be transported into the lumen, and kinins therefore have a non-selective effect on NaC1 and KC1 secretion. O'Donnell et al. (1998) and Terhzaz et al. (1999) used the GAL4-directed expression of aequorin in D. melanogaster tubules to show kinins produce a rapid (
[ ,'~me

K

:,>,

K

t

r, >,:,e ~:P : :,,~ ;t v,r E~; ,r,J }<~ E ~ :):he KR

[ ,;{,mS ~1 [ £ o::]e K~

5<:(ui D

m,

<:<:,

r

KI2 Kl<

t

~ e

,'>,

K~

r,yn st ~k

><1:2

Xt

l: : : m e KF) ,fn~r

>¢

> < , ,ri:

>I

_ .

r~lT S L P G M T L D P S[~]P D P~.J~ M LI ELQ o ~ ~ ~,~,~ or~ o ~ D ~

~s O,E s~s~

' ] V I ~ I F f ~ V ~

I

~

Y I v [S PF D~LYPr~[. ' N N L. s. .LR.L S . LS L[~A E EEA [ ~ A NDV ~s ~ ~ TT % ~ O S r ~ E ~

F,.[~-~--LJ'~sF~'M1Q[~'-'N1FL~

FQA , . ~ @ , ~ P L FS I P FQ F Q A A L L O R W V LIA NIF MISISIL PIP H4 T~ L~ C~ T~ VA ~ C P F G Mll AI L L IPlFIK AIGC SIKIKIRAIA I L L ~ s t l t W I A V G I G

F V Q r ~ L S V N V S["~F T L ~ FV QVIV TIV NIL TII F T I p~F F V

L w ~ T ~ S F E Y D

ER

A

V

D

I A N L A,..~A D L I J - ~

IA I V

N H

H A L~] ..... L~'~ RFRE N E TY NV NN I V I P . . . . . . IRICD

WE~

V G IF]R.I~A W ~ I F - ~ H V [ ' ~ L['Q'-~_E, FF-I'I ~ L TF~cF'~A[~-'~z , L KF~]~sF~--~'~G A~ E K N LI.,~.IDIDIQ L Q S,.E.~Y TIL VIF VIQ Y LIVIPIF C Vl IISty_.V.,Y I I,..~I~,AIV RI..L,W GITIRIAIP G,.~A,..Q]D SIRDI APDNWLIDIFH LYIY NIT LI[ VC.IFIQY LILIP LIV I I I I T Y CIYICIRII[~&]WH IIWG S RIR[.p_.~] IAIHV T T E H6 H7 A G I ~L K N K K K~V IKML A A I~ ] C W L P ~ R I r ~ I [ ~ W F C,.~T H I TL NK KV KM t IVV I~ GLCWLPLQ YN L Y VET IV F ~ ] KE~ ' ] K D Hy ~ y I I F CD DVR C V CWl P HN I N L H INI IWFISSN L A M SN O YNP yIAI I Y N EIRIF K R E FIA 7 % - - [ ~ l t ~ MSN CYNP YG NEKFKR FNK AA . . . . . . . . . . . . [~] ...... E FH Q VM

F P

G Hr~Iy S P K F ~ A ~ I Y E Q E [ I ~ KFK S DAH F MHTR K A~T~v K y y TIEIYIFIS E DIA NIl

VSM H F LS FK KAST tI~ ] l S T Y A r ~ S ~1 M~ ] I R S N L FNGS A ~ ] AL R G G V N NQV G ~1 PG ~1 H M P ~1 V H I~1 S G A N S G I C R A N TINIGHC P A N R H G A V G T TS T E T T R S M S R S CK TRRRRQT

E] N G S S G Q N N N DERRE TSS

-

FIG. 16 The aligned sequences of kinin receptors Dora L. smg,nalL~ (Lymst-KR), B. microl~h~,~" (Boomi-KR) and a putative receptor (CGI0626; Drome-KR) from D. melam~gaxter. The sequence alignment was perforined with Clustal W, and residues identical in at least two sequences are boxed. Bars over the protein sequences indicate the seven putative transmembrane domains of the Lymst-K receptor. The C-terminus of the putative Drome-KR is considerably longer than the other sequences, and the last 73 residues have been omitted.

334

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY •" " 4 , H+

A

I

Na* K*

CI

! H*~

Cl-

-~ "~Iv',Ca2+"~ SOCC

Leu ma-K-Vlll

K+ ~ . . . . . ,H.

B

.... "

cGMP

NO

,,

Ca 2* ",: C/tM -->-{ ~ iP3 ~f ',,

CAP2 b

Ca 2+

,, C a 2+

SOCC

FIG. 17 A Proposed models for kinin activity on stellate cells (SC) of D. melanogaster (Drome-K) and principal cells (PC) of A. aegypti (Leuma-K-VlI1). Kinin stimulation results in the release of Ca 2~ from intracellular stores, which may 3+ be augmented by Ca- uptake through store-operated calcium channels (SOCC; A. aeg/ptO. A rise m the mtracellular [Ca- ] opens a Cl conductance pathway that is transcellular (through stellate cells) in the fruit fly and paracellular in the mosquito. B Proposed model for the activation of soluble guanylate cyclase by Manse-CAP2t, in D. melanogaster principal cells (based upon Rosay et al., 1997 and MacPherson et al., 2001). CAP,b binds a receptor to activate phospholipase C (PLC), leading to increased levels of lnOSltOl trlsphosphate (IP3) and C a - release lrom mtracellular stores. The emptying of these stores opens store-operated calcium channels (SOCC) and the resultant rise in intracellular [Ca 2+] activates a Ca 2 ~-calmodulin (CAM)sensitive nitric oxide synthase (NOS) increasing production of nitric oxide (NO). NO in turn activates a soluble guanylate cyclase (sGC) to increase production of cyclic GMP (cGMP), which stimulates cation transport, and hence fluid secretion, by activating the apical V-ATPase. Cyclic GMP may also activate cyclic nucleotidegated Ca 2" channels to further raise the intracellular [Ca2+]. The symbols have the same meaning as in Fig. 2.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

335

- > Br > C1 ) of the type 1 channels (the most frequently observed) is the same as that of unstimulated tubules, compared with Br > CI > 1 after kinin stimulation (Yu and Beyenbach, 2001b). O'Connor and Beyenbach (2001) concluded the type I channels could be a route for transepithelial Ctlnovement under control conditions, but found no evidence for them being Ca ~+-actlvated. Kinins are believed to act on principal cells in A. aegypti (see Fig. 17A) and the effect of Leuma-K-VIII on the TEP, transepithelial resistance and CI diffusion potentials in short perfused tubule segments that lack stellate cells is identical to its effect on longer segments (Yu and Beyenbach, 2001c). Indeed, principal cells must mediate kinin activity in A. domesticus tubules, because stellate cells are absent (Hazelton et al., 1988). It is significant that Terhzaz et al. (1999) occasionally saw a Ca 2+ response in D. melanogas'ter principal cells following stimulation with Drome-K, because it shows that these cells can have kinin receptors. Kinins stimulate IP3 production by A. ae~ypti tubules (Cady and Hagedorn, 1999b) and Yu and Beyenbach (2000, 2001a) suggest the emptying of IP~-sensitive stores triggers Ca 2~ uptake through store-operated channels in the principal cell basolateral membrane, although this has still to be confirmed by direct measurement of cytosolic Ca 2+ levels. Interestingly, Culsa-K-II ( N N A N V F Y P W G a ) also stimulates cyclic AMP production by A. ae,~ypti tubules, whereas the ahnost identical peptide Aedae-K-lll ( N N P N V F Y P W G a ) does not (Cady and Hagedorn, 1999b), as shown previously for other kinins (Coast, 1995; Thompson et al., 1995: Terhzaz e: a/., 1999). There could therefore be different populations of kinin receptors coupled to different G proteins. The C1 shunt activated by Leuma-K-VIII in A. aegypti tubules has been characterised electrophysiologically (Pannabecker e: al.. 1993) although its location, paracellular versus transcellular, remains uncertain. Nevertheless, data obtained for CI diffusion potentials in response to concentration gradients imposed from the apical and basal surface of kinin-stimulated tubules can best he explained by: C1 moving through a paracellular pathway that contains only one barrier to diffusion, namely the septette junctions (Beyenbach, 1995: Yu and Beyenbach, 2001b). In support of this, Leuma-KVIII increases the tubule permeability of .4. aegypli tubules to inulin and sucrose (Wang el al., 1996), both of which move through a paracellular pathway into the lumen (Maddrell and Gardiner, 1974). Indeed, an X-ray microanalysis study' of ion transport by MTs of the black field cricket, Teleogryl/us oceanicus, concluded that Br (and by inference C1 ) crosses the epithelium through the intercellular clefts (Marshall and Xu, 1999), although this was under basal conditions. It is worth noting that the direction of C1 movement through a paracellular pathway depends upon the transepithclial electrochemical gradient. In R. pro/i.vus distal tubule, this gradient is towards the haemolymph side (lanowski and O'Donnell, 2001), which could explain why kinins have no effect on fluid secretion (Te Brugge et al.. 2002).

(I

336

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

5.2.5 CAP2~,/PVK-2 Manse-CAPxb has no effect on secretion by adult M. sexta tubules (N. Tublitz, personal communication), but has diuretic activity in D. melanogaster (Davies et al., 1995), M. domestica and S. calcitrans (G. M. Coast and R. J. Nachman, unpublished observations). The related peptide, Peram-PVK-2, also has diuretic activity in M. domestica and S. calcitrans. Rosay et al. (1997) used GAL4-directed aequorin gene expression to show that CAP2b increases the cytosolic [Ca 2+] of principal cells in D. melanogaster tubules. External Ca ,-+ is required, and is believed to enter through store-operated channels in the basolateral membrane (MacPherson et al., 2001). The rise in cytosolic Ca 2+ initiates a second messenger cascade that culminates with nitric oxide (NO) activating a soluble guanylate cyclase and stimulation of the apical V-ATPase by cyclic GMP (see Fig. 17B; Davies et al., 1997; O'Donnell et al., 1996; reviewed by Dow and Davies, 2001). Stimulation of M. domestica tubules by Manse-CAP~b is, however, unlikely to involve activation of the V-ATPase by cyclic GMP, because the cyclic nucleotide does not mimic its diuretic activity and both V, and the TEP are depolarised (C. S. Garside and G. M. Coast, unpublished observations). In contrast to its diuretic activity in dipteran insects, Manse-CAP2b acts via a cyclic GMP-dependent mechanism to reduce secretion by R. prolixus distal tubules stimulated with serotonin (Quinlan et al., 1997). There is no evidence for the involvement of NO, which suggests that CAP2b activates a membranebound guanylate cyclase. Based upon the antagonistic effects of cyclic GMP and cyclic AMP on distal tubule secretion, Quinlan and O'Donnell (1998) suggest that cyclic GMP and CAP2b reverse the diuretic activity of serotonin by activating a cyclic AMP phosphodiesterase, thereby speeding up the degradation of the second messenger. 5.2.6 Calcitonin-like peptides Calcitonin-like peptides stimulate secretion by MTs from D. punctata (DippuDH31; Furuya et al., 2000b), D. melanogaster (Drome-DH3L; Coast et al., 2001) and L. migratoria (Dippu-DH31; Furuya et al., 2000b) at nanomolar concentrations. In the fruit fly, this is attributed to the cyclic AMP-dependent activation of the apical V-ATPase and the resultant increase in K + transport (accompanied by CI-) into the lumen (see section 5.2.3; Coast et al., 200l). Intriguingly, however, although Dippu-DH3~ was isolated using a cyclic AMP assay to monitor activity (see section 4.2.2.1), its effect on secretion by cockroach and locust MTs is only 50% that of Dippu-DH47 and Locmi-DH, respectively (Furuya et al., 2000b). Indeed, Dippu-DH31 acts synergistically with these CRF-related peptides, which suggests that it targets a different cyclic AMP-dependent effector system or activates some other second messenger pathway.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

5.2.7

337

Tenebrio A D F a ( T e n m o - A D F a )

T e n m o - A D F a acts via cyclic G M P to inhibit secretion by c.75% in the free portion of larval tubules with an EC50 of 10 fM (Eigenheer el al., 2002: see section 4.2.6), which is remarkably low for a neurohormone. Its antidiuretic activity resembles that of Manse-CAP2b in the control of R. prolixus tubules (Quinlan el al., 1997), and CAP2b reduces secretion by T. molitor tubules, but is c. l0 s times less potent than Tenmo-ADFa. The availability of Tenmo-ADFa and Tenmo-DH3v have allowed, for the first time, the study of endogenous neuropeptides with antagonistic action on MTs. Wiehart et al. (2002) show that not only do cyclic AMP and cyclic G M P act antagonistically in controlling T. molitor MTs, but that T e n m o - A D F a can reverse the effect of Tenmo-DH37 applied first to tubules. Conversely, tubules inhibited with Manse-CAP2b will increase their secretion on addition of Tenmo-DH37. The effects of the peptides are mimicked by treatment of tubules with the second messengers. Interestingly, use of 100nM Tenmo-DH37 to stimulate tubules, followed by 100nM Tenmo-ADFa, barely led to return of tubules to control levels of secretion. In this context, it is interesting to note that cyclic AMP levels in tubules stimulated with 10nM Tenmo-DH37 in the absence of any phosphodiesterase inhibitor show a greater decrease on treatment with 1 pM TenmoADFa than with 1 nM, indicating a receptor desensitisation or internalisation mechanism operating at "high" levels of ADFa (Eigenheer et al., 2002; see section 4.2.6). T e n m o - A D F a and cyclic G M P are thought to function by activating a cyclic AMP phosphodiesterase, as proposed for the antidiuretic activity of Manse-CAP2b (Quinlan and O'Donnell, 1998), but this requires more study (Wiehart el al., 2002).

5.2.8

Partially characterised factors acting on Malpighian tubules

MNP acts via cyclic AMP to stimulate Na + transport and fluid secretion by opening a basolateral Na + conductance in the principal cells of A. aegypti tubules (Petzel el al., 1987; reviewed by Beyenbach, 1993; Beyenbach, 1995). It has a biphasic effect on the TEP of perfused tubules (Petzel et al., 1985), which resembles the response to 100 nM Culsa-DP (Clark et al., 1998a: see section 5.2.3.2), but based upon its estimated Mr (see section 4.2.8), MNP is unlikely to be a CRF-related peptide. The antidiuretic factor (FopADF) isolated from F. polyctena reversibly inhibits secretion by ant tubules, and has a similar effect on the free portion of tubules from T. molitor larvae (Laenen et al., 2001). Since cyclic G M P (and cyclic AMP) stimulates secretion by ant tubules (De Decker, 1993: Laenen, 1999), F o p A D F activity differs from that of Manse-CAP>, and T e n m o - A D F a (see sections 5.2.5 and 5.2.7). Both ~, and Vb depolarise in response to F o p A D F , which can be duplicated using Ba 2+, to block K + channels, and

338

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

2,4 dinitrophenol, to inhibit the apical V-ATPase, giving some indication as to how it may act (Laenen et al., 2001). 5.3

REGULATION OF H I N D G U T ACTIVITY

Figure 18 (redrawn from Phillips et al., 1994) summarises the major transepithelial transport processes demonstrated in S. gregaria hindgut and indicates which of these are stimulated or inhibited h7 vitro by exogenous cyclic AMP (and/or 1/zM forskolin), saline extracts of locust CC or VG4-VG7. These are shown in the context of measured in situ composition and volume of MT and hindgut fluids and final excreta in starved dehydrated locusts. Nearly all M T fluid is recovered (see urate as a volume indicator) in the hindgut, which is the dominant site of acidification (body acid-base regulation) and which contributes substantially to ammonia (i.e. nitrogen) excretion, so that ammonia concentration is three times that of urate in the final excreta, contrary to earlier dogma. Virtually all Na +, bicarbonate and proline is reabsorbed. Hanrahan and Phillips (1984) and Phillips et al. (1986, 1988) used doublebarrelled ion-sensitive glass microelectrodes and electrophysiology to show that elevated intracellular cyclic AMP acts directly on the apical C1 pump and on both apical K + and basolateral CI- conductances (ion channels) to stimulate KCI absorption across locust rectum, while apical H + secretion was greatly inhibited (reviewed by Phillips et al., 1994) and Na + reabsorption was unaffected. Chamberlin and Phillips (1988) measured elevation of cyclic AMP levels in rectal tissue in parallel with increases in Isc caused by extracts of CC: several treatments and conditions associated with high l~c were all accompanied by increased tissue titres of cyclic AMP, whereas increases in cyclic G M P were delayed. Similarly, Richardson (1993) used single-barrelled electrodes to study electropotentials, resistances, and specific ion conductances (by external ion substitutions) across apical and basolateral membranes of locust ilea stimulated with exogenous cyclic AMP (Fig. 19; see Phillips el al., 1998a). This second messenger caused a large K + and a small Na + conductance increase in the apical membrane without any drop in V,. The latter result could be explained if cyclic AMP also directly stimulated an apical C1pump to balance enhanced cation entry into the epithelial cells. In contrast to the situation in the rectum, the basolateral membrane resistance of the ileum is very low relative to the apical membrane. The ileal basolateral membrane has high conductances for both K + and C1 even in the unstimulated state. Not surprisingly therefore, no increase in basolateral CI- conductance was measured after stimulation with cyclic AMP. The ileal response to cyclic AMP therefore differs from that of the rectum in three respects: (1) an apical Na + conductance appears; but (2) there is no change in the large basolateral C1- conductance; and (3) cyclic AMP has no effect on ileal acid secretion (JH) whereas strong inhibition was observed for the rectum. However, crude

INSECT DIURETIC AND ANTIDIURETIC HORMONES

L

339

MIDGUT /

/ ] SECRETEisosmotic KCl-richfluid (5#1 h"1) [ MALPIGHIAN TUBULES ~ with (mrnol I-1):30 Na+; 5 NH4+; / Isosmotic \ " I 6 urate (as a precipitate;38 proline,5.5 HCO 3/ pHTI ~ MECHANISM

CONTROL

/ / ~

H+~ r - . ) ~

OH- (HCO3-)

N~+.~ ILEUM

~)~ //~

I

~¢ CC, ITP * cyclic AMP

Na+(+Jv) KCl-drivenJv

~k cyolioAMP,CC, ITP ~k cyolicAMP,CC, ITP

\,SZcOtiC/

L NW.

Na+

NO

\Hyperosmetic /

~ ~

pHi6 ~ . . ~ EXCRETEhyperosmotiepellets II ~ I (25 Yowater; high KCI) Y with (mmolkgq): 263 NH4÷" + I 85 urate (as a precipitate)<1 total CO2" low Na

FIG. 18 Diagram of the desert locust excretory system, summarising major transepithelial mechanisms, including solute-driven active fluid absorption (Jr), which have been demonstrated and characterised in detail. Control of specific transport mechanisms (stimulation by upward and inhibition by downward arrows) by cyclic AMP, crude corpora cardiaca extracts (CC), and either partially purified (CTSH) or fully sequenced (ITP) neuropeptides from CC has been established using isolated hindgut preparations in vitro. Relevant solute concentrations in Malpighian tubule fluid (primary urine) entering the hindgut (collected in sire) and in faecal pellets excreted by locusts starved for one day are shown, as well as changes in luminal pH (acidification) and osmotic concentration during passage through the gut of these animals. Well over 90% of the water in primary urine is reabsorbed in the hindgut, leading to a 15-fold increase in urate concentration. The ratio of NH + to urate increases from near 1:1 in primary urine to 3:1 in the final excreta due to hindgut secretion of NH +. Redrawn from Phillips et al. (1994).

340

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

LUMEN

HAEMOLYMPH

CELL

~"= CI"

Cl'--ll

~.J f

~..::-@. .. ........ cAMP~(-- " @ Schg r-IT P

K+ { Na*[

I L

~-

W p~,.-

K*

D>-

K+<-..

I L-> Na*

NH4*<'-(I I---

H+ ,-ll i.--.~....@ ...... .......? ~

Schgr-ITP

FIG. 19 Proposed model for control of ion transport across desert locust ileum. Ion Transport Peptide (Schgr-lTP) acts via cyclic AMP (at least partially) at the apical membranes to increase both K ~ and Na + conductances and to stimulate the electrogenic Cl pump directly. Schgr-ITP must act via another second messenger to inhibit the apical H* pump. The symbols have the same meaning as in Fig. 2. Redrawn from Phillips et al. (1998a,b): based on electrophysiological studies by Richardson (1993).

extracts o f c o r p o r a cardiaca strongly inhibited ileal H + secretion, indicating that a second messenger other than cyclic A M P must be involved in the case of the ileum. The only fully characterised A D H that unequivocally has antidiuretic activity on the hindgut is Schgr-lTP. Purified Schgr-ITP mimics the actions o f locust CC extracts on the ileum (see section 2.3), but has no effect on total a m m o n i a secretion. Thus a single neuropeptide (Schgr-ITP) can account for all the actions o f crude CC extracts on locust ileum. Exogenous cyclic A M P has the same range o f actions on ileal transport processes as Schgr-ITP, except that it does not inhibit acid secretion, but does stimulate a m m o n i a secretion. The results are consistent with cyclic A M P being the second messenger for SchgrITP actions except for the control of acid secretion (Fig. 19). In support o f this, ITP causes an elevation of cyclic A M P levels in ileal tissue within 1 h (reviewed in Audsley e t al., 1994), but the time course for this change and the effects on other potential second messengers have yet to be investigated.

INSECT DIURETIC AND ANTIDIURETIC HORMONES 5.4

341

FLUID UPTAKE FROM THE C R Y P T O N E P H R I C COMPLEX

Fluid uptake from the cryptonephric complex of M . s e x t a larvae varies with the water content of the diet, which suggests it is under endocrine or nervous control (Reynolds and Bellward, 1989). Using an inverted rectal sac preparation, Audsley et al. (1993) showed that Manse-DH stimulates fluid uptake from the cryptonephric complex. This antidiuretic effect contrasts with the diuretic activity of Manse-DH on tubules from adult moths (Audsley el al., 1993) and, by inference, on the free portions of larval tubules, although this has not been confirmed. Manse-DH increases cyclic AMP production by the cryptonephric complex and is believed to act on the cryptonephric tubules, because its activity is abolished when the structural integrity of the complex is destroyed by cutting the perinephric sheath (Audsley et al., 1993). The activity of ManseDH is blocked by bumetanide and by removing Na +, K ~ or C1- from the haemolymph side of the preparation, which suggests that it stimulates Na+/K+/2CI cotransport (Audsley et al., 1993). The bumetanide-sensitive Na+/K+/2CI - cotransporter cloned from ,1//. se.vta MTs has a conserved PKA phosphorylation site (Reagan, 1995b; see section 5.2.1.2) and is a likely target for Manse-DH acting via cyclic AMP. Stimulation of the transporter will increase KCI and/or NaC1 secretion from the haemolymph into the lumen of the cryptonephric tubules, and thus promote the osmotic withdrawal of water from the rectal contents (Audsley el al., 1993). Interestingly, the shorter CRF-related peptide isoform (Manse-DPII) has no effect on fluid uptake from the cryptonephric complex (Audsley el al., 1995), and the cryptonephric tubules may therefore express a different receptor from that cloned by Reagan (1995b). In this context, it is noteworthy that of the two CRF-related peptides identified from T. m o l i t o r pupae, Tenmo-DH47 is between 200 and 600 times less potent than Tenmo-DH37 in stimulating secretion by the free portion of larval tubules (Furuya et al., 1995; Furuya et ctl., 1998; Wiehart el ell., 2002), and it may therefore have some other function. One possibility worth exploring is that Tenmo-DH47 acts on the cryptonephric tubules as shown for ManseDH. Two additional antidiuretic factors (ADFA and ADFB) are present in extracts of brain/CC/CA from M . s e x t a larvae that stimulate fluid uptake from an inverted cryptonephric complex (Liao et al., 2000; see section 4.2.7). The more potent (or more abundant) is ADF-B, which like Manse-DH stimulates cyclic AMP production by the cryptonephric complex, although its antidiuretic activity is not inhibited by bumetanide (cf. Manse-DH: Audsley el al., 1993). The ion requirements for Manse-ADFB activity also differ from those of Manse-DH (Liao et aL, 2000). Chloride is required on the luminal side of the complex, and CI channel blockers DIDS (4-4'-diisothiocyanatostilbene-2,2'-disulphonic) and DPC (diphenylamine-2-carboxytic acid) added to either surface inhibit its activity. Significantly, addition of either Bafilomycin A~ or amiloride to the luminal side of the complex, to inhibit MT

342

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

V-ATPase activity and proton/H + antiports, respectively, blocks the basal rate of fluid uptake, but does not prevent stimulation by ADFB (Liao et al.+ 2000). From this, Liao et al. (2000) concluded that ADFB does not act on the cryptonephric tubules, but increases fluid uptake by stimulating an apical membrane C1- transporter in the 'normal' rectal epithelium (Fig. 20). The activity of Manse-ADFB therefore resembles that of ITP and CTSH, which act via cyclic AMP to stimulate an apical electrogenic C1 pump in the locust ileum and rectum, respectively (see sections 4.3.3 and 5.3). Significantly, the solubility properties of Manse-ADFB, which has not been fully characterised (see section 4.2.7), resemble those of CTSH and ITP, and they may be similar peptides. Since Manse-DH stimulates secretion by the free portion of the MTs and reabsorption from the cryptonephric complex, it may have little impact on excretory water loss, and function instead to increase fluid recycling through the excretory system, thereby accelerating the clearance of toxic wastes. On the other hand, Manse-ADFB may be more important in regulating faecal water content. Importantly, KCI reabsorbed from the lumen of the rectum across the +normal' epithelium is available for transport into the cryptonephric tubules and can therefore be returned to the midgut via the straight ascending and descending segments (Moffett, 1994). 5.5

5.5.1

STRUCTURE/ACTIVITY STUDIES

Serotonin

A range of substances chemically related to serotonin has been tested on MTs of both R. prolixus and C. morosus (Maddrell et al., 1971). Serotonin is a potent stimulator of secretion by the MTs of both species, and activity is easily lost with changes to its structure. Thus, the absence of the hydroxyl group (as in tryptamine) or substitution of the hydrogen atom in this group (as in 5-methoxytryptamine and 5-benzyoxytryptamine) results in loss of activity. Similarly, moving the hydroxyl group into either position 4 or 6 also abolishes activity (as in 4-hydroxytryptamine and 6-hydroxytryptamine), and substituting the side chain (as in 5-hydroxytryptophan ethyl ester and 5-hydroxytryptophan) also results in an inactive compound. Variation to the terminal amino group, however, can produce active compounds. Methylation results in equally active compounds (N-methyl-5-hydroxytryptamine and bufotenine), although acetylation (N-acetyl-5-hydroxytryptamine) reduces the activity by 300 times on R. prolixus tubules and abolishes it on C. morosus tubules. In contrast to the small number of tryptamine derivatives that can stimulate secretion by R. prolixus tubules, rather a larger number can interfere with the serotonin-induced secretion of the tubules. Interestingly, these derivatives also interfere with the activity induced by extracts of the M T G M (interpreted by Maddrell et al., 1971 as interfering with the action of the assumed, peptidergic,

INSECT DIURETIC AND ANTIDIURETIC HORMONES

343

._0 ~SE

z

.~J

=~

"I-

~.~

W

~

.~

..I

z

I-0

o

I.I.I

W

T

D_

b

t

o -z-

c o

~ ~

2==>

O

¢c

5= e0

o

B~ .=o= ._~

~ N

.~ o

,~

a_,_ o+ ~

oz

Oo~

F--~

~ o

344

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

DH). In addition, a number of the tryptamine derivatives interfere with the response of the MTs to externally applied cyclic AMP, thereby posing problems with interpreting the site of action of these inhibitors. In a general sense, then, serotonin and a few of its derivatives stimulate secretion by the isolated MT of R. prolixus and C. morosus. Substitutions of the terminal amino group of serotonin have little effect on activity, but changes to the indole ring and the hydroxyl group tend to abolish activity. Most of the derivatives of serotonin and other related amines that have little or no stimulatory activity on MTs, actually act as inhibitors of secretion on R. prolixus tubules. However, their site of action is unclear, since not only do they inhibit secretion induced by serotonin, but also secretion induced by the apparent peptidergic DH and by cyclic AMP itself. It would be interesting to test these inhibitors on serotonin and peptidergic DH-induced accumulation of cyclic AMP. to determine if the inhibition is at the level of the receptor or downstream from the hormone/receptor interaction. 5.5.2

C R F - r e l a t e d peptides

Structure activity studies with CRF-related diuretic peptides show that a region towards the N-terminus is important for receptor activation while the C-terminal two thirds of the peptide are required for receptor binding. Thus, the N-terminal truncated analogue Manse-DH(13~,l) binds Manse-DHR expressed in Sf9 cells with high affinity (ICs0 2.8 nM compared with 0.16 nM for the intact peptide) but has no effect on cyclic AMP production, whereas Manse-DH(3-41) has high affinity for the receptor and is a potent stimulant of adenylate cyclase activity (Reagan, 1995a). The binding affinities of ManseD H ( 2 1 ~ I ) and Manse-DH(26~[1) are reduced 100-fold and 1000-fold, respectively, and Manse-DH(31~41) has no binding activity at 1/,M (Reagan et al., 1993). Data obtained with N-terminal truncated analogues of Achdo-DP in a cricket tubule diuretic assay confirm the importance of the N-terminus for receptor activation (see Coast et al., 1994). The activities of Achdo-DP(6 46) and Achdo-DP(7 46) are virtually identical to Achdo-DP, whereas the activity of Achdo-DP(1146) is reduced by 60% and Achdo-DP(23-46) is inactive. Further research (P. Dey and D. A. Schooley, unpublished observations) has shown that deletion of the first four residues from Manse-DH decreases activity, while deletion of the next residue essentially abolishes it. Taken together with the results of Reagan (1995a), this defines the receptor activation domain of Manse-DH as lying within residues 4 1 2 . A Met residue within the corresponding region of Locmi-DH can become oxidised, causing loss of activity (I. Kay and G. M. Coast, unpublished observations), which could explain why [Ile 2"11]-Manse-DH (with lie replacing Met 2 and Met 1J) maximally stimulated cricket tubule secretion, whereas Manse-DH gave only a 60% response (Coast el al., 1992). Not surprisingly, Locmi-DH(l 23) and LocmiDH(24 46) are inactive, whether tested separately or together (Nittoli et al.,

INSECT DIURETIC AND ANTIDIURETIC HORMONES

345

1999; G. M. Coast. unpublished observations), showing the binding and activation domains must be linked to produce a biologically active peptide. Reagan et al. (1993) were unable to measure the affinity of Manse-DH-acid for the Manse-DHR because it was rapidly degraded in the receptor-binding assay. However, it stimulates cyclic AMP production by adult M. s e x t a MTs (Audsley et al., 1995) and post-eclosion diuresis in decapitated newly emerged P. rapae (Kataoka et al., 1989), but with a 1000-fold lower potency than Manse-DH. CRF is believed to assume a rod-like o~-helical conformation when interacting with receptors (Romier et al.. 1993: Miranda et al., 1997), whereas molecular modelling and circular dichroism (CD) studies of Manse-DH and Manse-DPI1 have led to a folded helix-loop-helix model, which brings the C-terminal amide close to the N-terminus (K. S. Copley, W. H. Welch Jr and D. A. Schooley, unpublished observations). To test this model, a disulphide bond was engineered between residues 3 and 40 of Manse-DH. Although the distance between the N- and C-termini of [Cys3"4°]Manse-DH is severely constrained, it stimulates cyclic AMP production by MTs from adult male moths with an efficacy and potency similar to that of the natural peptide, which is entirely consistent with the proposed model (K. S. Copley, W. H. Welch Jr and D. A. Schooley, unpublished observations). The loop region permits some variation in linear sequence length while still allowing conserved N-terminal and C-terminal residues to interact with the receptor, which explains why CRF-related peptides of differing length can bind and activate the Manse-DHR with high affinity (see section 5.2.3.1). Copley et al. (1998) have successfully expressed a synthetic gene encoding Manse-DH extended by a C-terminal Gly residue in a protease-deficient strain of the yeast, S a c c h a r o m v c e s cerevisiae. The purified peptide, Manse-DH-Gly was treated with peptidylglycine c~-amidating enzyme to convert it to ManseDH, which was shown to stimulate cyclic AMP production by MTs from adult male moths (Copley et al., 1998). The development of this expression system for CRF-related peptides will greatly facilitate future structure/activity studies that would otherwise be prohibitively expensive. 5.5.3

Kinins

The conserved C-terminal pentapeptide is the active core of the insect kinins, in that it is the minimum sequence required for diuretic activity (Coast et al., 1990a). However, while the C-terminal pentapeptides of Achdo-Ks and Leuma-Ks retain the activity and potency of the parent peptides in a cricket tubule assay, N-terminal truncation of Musdo-K results in a loss of potency on housefly tubules (Coast et ell., 2002). Thus, Musdo-K(10 15) is >1500 times less potent than Musdo-K (ECs0 0.13 nM), but retains full activity, whereas Musdo-K(ll -15) gives only a 20% response at 100#M. Residues outside of the pentapeptide core sequence are therefore required for high-affinity binding

346

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

to the Musdo-K receptor. Very similar results were obtained with the housefly hindgut, which suggests that the same receptor is present on both target organs (Coast el al., 2002). Within the active core, Phe I and Trp 4 are invariant (see Table 3) in insect kinins, and replacing either with Ala results in loss of activity, whereas other Ala substitutions have little effect on activity or potency in cricket tubules (Nachman et al., 1995). Synthesis of a biologically active, rigid cyclic hexapeptide analogue (cyclo[AFFPWG]) showed that the active core adopts a fi-turn when bound to receptors (Roberts et al., 1997). This analogue can adopt either a type-I fl-turn encompassing FPWG (Phe-Pro-Trp-Gly) with a trans proline or a type-Vl/?-turn encompassing FFPW (Phe-Phe-Pro-Trp) with a cis proline, To determine the active conformation, kinin analogues incorporating a tetrazole moiety that mimics a cis peptide bond were synthesised and tested on cricket tubules (Nachman et al., 2000; Nachman et al., 2002b). The tetrazole-containing analogue (Phe-Phe@[CN4]Ala-Trp-GIy-NH2) retains maximum diuretic activity, confirming the type-Vl /?-turn as the active conformation. This brings the aromatic side chains of Phe I and Trp a together on one face of the molecule where they can interact with the receptor, while the side chain of the variable second residue in the active core lies on the opposite face, away from the putative receptor-binding site. Significantly, a tetrazole analogue with D stereochemistry is a partial antagonist of Achdo-K-L and could prove to be an important lead compound in future studies (Nachman et al., 2002b). The pseudodipeptide analogue 4Pbm-eChd-Trp-Gy-NH2 (4Pbm 4-phenylbutylamine; eChd cis-cyclo-hexyldiacyl) retains diuretic activity, showing that the minimum requirement for receptor binding and activation is the C-terminal dipeptide (Trp-Gly) plus an N-terminal extension incorporating the equivalent of the aromatic ring of Phe appropriately distanced from the Trp residue (Nachman et al., 1998). Even the C-terminal amidation is not essential, because although A F F P W G - O H has low potency this is improved when the negative charge is neutralised, as in the ester analogue AFFPWGOCH3 (Nachman et al., 1995). 5.5.4

CA P21,

Manse-CAP2b is a potent stimulant of M. domestica tubules (ECs0 45nM; G. M. Coast and R. J. Nachman, unpublished observations) and it is likely that a similar peptide is present in the housefly, as suggested by the presence of CAP2b-like peptides in another dipteran insect, D. melanogaster (see Table 4). The minimum sequence required for full activity is the C-terminal hexapeptide (YAFPRV-NH2), within which the critical residues appear to be Arg and Val (G. M. Coast and R. J. Nachman, unpublished observations). Interestingly, many of the CAP2b analogues tested had reduced activity at high concentrations (> 10-times the ECs0), which was also noted with high concentrations of CAP2b on D. melanogaster tubules (Davies et al., 1995), and may be evidence

INSECT DIURETIC AND ANTIDIURETIC HORMONES

347

of receptor down-regulation or the activation of alternative second messenger systems that antagonise the diuretic response. Solution conformation studies on Manse-CAP2b and a restricted-conformation analogue suggest that it adopts either an ~-helical structure or a cis Pro type VI fi-turn, with the former being more likely, because an analogue incorporating the tetrazole mimic of a cis peptide bond is inactive (R. J. Nachman and G. M. Coast, unpublished observations).

5.5.5

Ion transport p e p t M e

1TP structure function studies were initiated to understand ITP interaction with its receptor (ITP-R), with novel pest control strategies in mind (i.e. antagonist development). Wang et al. (2000) and Zhao (2000) (reviewed in Phillips et al., 2001) used the plasmid-Kcl expression system and site-directed mutagenesis to identify key amino acids in Schgr-lTP required to activate its receptor. Of the six N-terminal residues (SFFDIQ), Ala substitutions at positions 1, 4, 5 and 6 had no efPect on activity, whereas both Phe residues were essential for stimulation of ileal lsc (Zhao, 2000; reviewed in Phillips et al., 2001). Transient expression yields a second minor (B) band (identified with ITP antibodies on western blots) that had the same initial amino acid sequence as the major ITP band (A) that co-migrates with SynlTP. This minor band disappeared when Ala substituted for Phe 2 or Phe 3, consistent with isomerisation of k t o D phenylalanine causing the minor band (see section 4.3.4.5). At the C-terminus, Wang el al. (2000) carried out progressive truncation of the final prepropeptide sequence ( M V E I L G K K ) , where G K K is the amidation signal plus dibasic cleavage site that is normally processed to M V E I L - N H > Removal of G K K reduces biological activity by 10-fol& and further truncation abolished all stimulation of ileal lsc even at 200-fold higher dosages. Single-site Ala substitutions confirmed that the amidation substrate Gly 7-~ is essential, but that a neutral substitution of Ala for Leu 72 is tolerated. In short, the Cterminal Leu-NH~ is essential for activity. Three disulphide bridges determine the distance between essential sites at the N- and C-termini. and breaking any one of these bridges by replacing a single Cys with Ala abolishes 90% of biological activity (King et al., 1999). Inactive analogues of ITP and ITP-L were tested for antagonistic actions on stimulation by the wild-type KcITP: two mutants (Ala substitution for Phe 3, and deletion of Leu-NH~) and KcITP-L all inhibited stimulation of ileal 1~ (Zhao, 2000). Ile TM was implicated in binding, because its deletion abolished the antagonist action of the truncated form, ending at residue 71. These structure function studies on locust ITP interaction with its receptor are the first for this large tamily of neuropeptides, of which 40 members have now been sequenced (Lacombe el al., 1999).

348

6 6.1

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

Distribution INTRODUCTION

Immunohistochemistry has become a pivotal tool for the neurobiologist and, together with in situ hybridisation, has revolutionised the detection and mapping of neuroactive chemicals in the nervous system (for a review, see N/issel, 1996a). In addition, the antisera used for immunohistochemistry can also be used for RIA, ELISA and western blotting, allowing for analysis of the distribution and quantification of antigen present. For the purposes of this review, we will focus our attention on the distribution of diuretic factors in the neuroendocrine system (since this would be the origin of the factors as neurohormones), and in possible peripheral target sites. Although concentrating on the presence of these diuretic factors in the neuroendocrine system, one should not forget the extensive distribution of processes from the NSC within the neuropil of the CNS, or the expression of the diuretic factors in interneurons within the CNS. Thus, these diuretic factors may also be involved in modifying the activity of other neurons within the CNS, some of which may be involved with behaviours associated with diuresis.

6.2

SEROTONIN

Serotonin is widely distributed throughout the insect nervous system, and its cellular distribution has been demonstrated in several species, using specific anti-serotonin antisera (reviews by N/issel, 1988; N~issel, 1996b). In addition to its presence in central interneurons, serotonin has also been detected in sensory neurons (Lutz and Tyrer, 1988), the stomatogastric system and gut (e.g. Davis, 1985; Klemm et al., 1986; Helle el al., 1995), in central neurons with efferent axons (Hustert and Topel, 1986; Davis, 1987; Ali, 1997), and, importantly for this review, in NSCs and their neurohaemal systems (e.g. see N/issel and Elekes, 1985; Orchard el al., 1989; Helle et al., 1995). A common feature in a number of insect species is the presence of a bilateral cluster of serotonin-like immunoreactive neurons in the SOG which have efferent axons. These neurons have been described in detail for P. americana, L. migraloHa, Leplinotarsa decemlineata, G. bimaculatus and M. sexta (Br/iunig, 1987; Davis, 1987; Griss, 1989; Haeften and Schooneveld, 1992; Hornet, 1999), but are also evident in other species (e.g. Lange et al., 1988; Homberg and Hildebrand, 1989). The peripheral projections and neurohaemal areas that they produce are extensive and circuitous, and are associated with the nerves of the mouthparts. An example is shown in Fig. 21 for P. americana. In this species, the serotonergic neurohaemal system in the head has been studied by cobalt backfilling, immunohistochemistry and electron microscopy (Davis, 1987). The neurohaemal system is supplied by two to three bilaterally paired neurons in

INSECT DIURETIC AND ANTIDIURETIC HORMONES

Frontal connective

349

Tritocerebrum

\

Jl

Labral nerve

Mandibular nerve 3 Suboesophageal ganglion

Serotonergic efferent neurons

, Maxillary nerve 2 Labial nerve

FIG. 21 Side view of the suboesophageal ganglion of P. americana illustrating the projections of the serotonergic efferent neurons to the nerves of the mouthparts. On these nerves the axons produce a meshwork of serotonergic neurohaemal areas. Redrawn from Davis (1987).

the SOG. Axons from these neurons project through the third branch of the mandibular nerve and then enter a link nerve leading to the second branch of the maxillary nerve. From the maxillary nerve, the serotonin-like immunoreactive axons extend back into the SOG, before projecting out again through the labial nerve and also through the circumoesophageal connectives into the labral nerve. The axons of these SOG neurons therefore enter all of the nerves of the mouthparts, where they branch into a fine meshwork of neurohaemal areas on the surfaces of these nerves. A similar situation has been described in L. mi~ratoria (Br'iiunig, 1987) and has been termed the ~satellite' nervous system in this species because the branching axons are actually present in a small nerve, which accompanies the major nerves to the periphery. Similarly, within adult C. erythrocephala, a neurohaemal plexus of serotonin-like immunoreactive processes is found in the neural sheath of the labrofrontal nerves, the maxillary labial nerves, the posterior part of the SOG, the cervical connective, the cervical nerves, the frontal prothoracic nerves and the dorsal

350

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

surface of the fused thoracico-abdominal ganglion (Nfissel and Elekes, 1985). The serotonin-like immunoreactive processes in the neural sheath appear to be derived from neurons within the SOG (Nfissel and Elekes, 1985). A most striking example of an extensive peripheral serotonergic system in insects is also shown in R. prolixus (Lange et al., 1988; Orchard et al., 1988, 1989). In this insect, five serotonin-like immunoreactive dorsal unpaired medial (DUM) neurons in the abdominal neuromeres of the M T G M project axons through their respective abdominal nerves. The axons result in neurohaemal terminals on the abdominal nerves, but also continue on to project over the entire surface of the epidermis on the dorsal cuticle. Neurohaemal terminals are also observed over several of the thoracic ganglia nerve roots. In addition, serotonin-like immunoreactivity is found in the CC, and over the surfaces of the salivary glands and digestive tract. Similarly, in many other species, serotonin-like immunoreactive processes have been shown to be present in neurohaemal areas (including the CC), salivary glands, stomatogastric nervous system and digestive tract (reviewed by Davis, 1985; N~issel and Elekes, 1985; Klemm et al., 1986; Konings et al., 1988; N~issel, 1988; Luffy and Dorn, 1991). Electron microscopy coupled with immunogold labelling or labelling with 5,7-dihydroxytryptamine (5,7-DHT) indicates that serotonin is present in typical neurohaemal terminals in the neural sheath of the ganglia and peripheral nerves, and is associated with a variety of granule types within these terminals (see Nfissel and Elekes, 1985; N',iissel, 1988; Haeften and Schooneveld, 1992; Miksys and Orchard, 1994). As discussed, serotonin has been identified in both central and peripheral nervous tissue of a variety of insects. In addition to its possible direct action on tissues supplied by serotonergic neurons, serotonin may also be considered a neurosecretory hormone in insects. Evidence for this is accumulating from a number of observations, including: the responsiveness to serotonin of tissues, in some insects, that do not appear to be innervated by serotonin-containing neurons (see Nfissel, 1988), including MTs (Maddrell et al., 1969, 1971), heart (Tublitz and Truman, 1985), visceral muscle (Brown, 1967) and salivary glands (Trimmer, 1985a, b); the presence of extensive serotonin-like immunoreactive neurohaemal areas in the neural sheath of ganglia and peripheral nerves of numerous insects (see above); the presence and quantification of serotonin in neurohaemal structures (Lange et al., 1988); the Ca2+-dependent release of serotonin from neurohaemal areas (Lange et al., 1988); and finally, the presence of serotonin within the haemolymph (Lange et al., 1989). Serotonin is considered a true DH in R. prolixus (see Mad&ell et al., 1991). Gorging in fifth-instar R. prolixus induces a reduction in staining intensity of serotoninlike immunoreactivity in the abdominal nerve neurohaemal areas, which coincides with a dramatic rise in the haemolymph concentration of serotonin (Lange et al., 1989; see section 7.1.1). In addition, interference with the serotonergic system, as with 5,7-DHT injection (Cook and Orchard, 1990, 1993a,b; see section 7.2.2), results in a delayed and reduced elimination of

INSECT DIURETIC AND ANTIDIURETIC HORMONES

351

urine following feeding. These findings, along with those of Maddrell et a/. (1991), lead to the conclusion that serotonin is indeed a neurohormone in R. prolixus released by the natural stimulus of feeding, with actions on at least the MTs. In addition, within A. aegypti larvae, the haemolymph titre of serotonin increases five-fold in response to elevated salinity (Clark and Bradley, 1997). Since serotonin has diuretic activity on larval tubules (Clark and Bradley, 1997), the authors consider that the MTs of these larval A. aegvpti arc hormonally stimulated by serotonin in response to increased salinity. Interestingly, they found no evidence for serotonin being a diuretic factor in response to feeding in larvae. 6.3

AVP-LIKE I N S E C J DH

In a detailed study of the localisation of AVP-like immunoreactive (VPLI) neurons in L. migratoHa, Thompson et al. (1991) describe a pair of cells in the SOG, each of which has a bifurcating axon. The posterior-directed branches of these axons give rise to aborisations in every ganglion of the VNC, and to plexuses of fine, blebbed processes in the proximal portion of most peripheral nerves arising from the VNC. Girardie and R~my (1980) suggested that these plexuses have a neurohaemal role, which is consistent with the presence of AVP-like immunoreactivity in haemolymph (Picquot and Proux, 1987), although, according to Thompson et al. (1991) they are not closely associated with the nerve sheath. The anterior-directed axons have relatively few branches in the tritocerebrum, but in the deuto- and protocerebrum they give rise to branches that run into the optic lobes. Extensive studies of 16 other acridoid species from a wide range of habitats (Tyrer el al., 1993) revealed the same pair of VPLI in the SOG, with extensive branching throughout the CNS. However, there appeared to be two distinct patterns of branching. In the first, exemplified by S. gregaria, most of the arborising processes are in the dorsal and lateral neuropil of all ganglia, whereas in the second, exemplified by L, mi,gratoria, the extensive arborisation is shifted peripherally to the optic lobes and peripheral nerves. Of the 17 acridoid species studied, the plexus of fine blebbed processes in the proximal parts of peripheral nerves, the putative neurohaemal release sites for AVPIDH, were only present in the subfamily Oedipodinae, of which k. migratoria is a member. Tyrer et al. (1993) conclude that the presence of these peripheral plexuses is determined by phylogeny rather than habitat. Species lacking the peripheral plexuses would not be able to release AVP-IDH into the circulation where it might function as a DH. Indeed, the conserved features of the branching pattern of VPLI neurons suggest that they have a central role. Perhaps one of the most surprising aspects of the extensive aborisations of the VPLI neurons in L. migrutoria is that there are no immunoreactive processes in the CC, which were identified by Mordue (1969) as the site of

352

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

DH storage and release, although Clynen et al. (2001) have detected a mass ion corresponding with the inactive F1 monomer (but not the F2 dimer) in dissected CC. Indeed, contrary to the findings of Mordue (1969), Proux et al. (1982) found that CC extracts had no effect on amaranth clearance from intact locusts or on urine production by isolated MTs, whereas both were stimulated by extracts of the SOG. Interestingly, when undamaged SOG were added to the medium bathing isolated MTs they had only a short-lived effect on fluid secretion, but when added together with an extract of the CC, the response was the same as that obtained with a ganglion homogenate, which suggests that the CC contain a factor that stimulated release of AVP-like material from the SOG (Proux el al., 1982). 6.4

CRF-RELATED PEPTIDES

The distribution of neurons expressing the CRF-related diuretic peptide phenotype has only been examined in a small number of insect species. Using antisera generated against Locmi-DH, Manse-DH and Manse-DPII. immunoreactive neurons have been shown to be widely distributed throughout the nervous system of L. migratoria, M. sexta, R. prolixus, and M, domestica. A consistent finding between these species is the presence of CRF-like immunoreactivity in the MNCs of the brain, posterio~lateral NSCs of the abdominal ganglia, and their associated neurohaemal organs. Within L. m~ratoria, Locmi-DH-like immunoreactivity is found in a group of approximately 120 MNCs in the PI of the protocerebrum with projections via the NCC 1 to the storage lobe of the CC (Fig. 22). The immunoreactive staining in the CC is pronounced, but is denser in the periphery of the storage lobe where it faces the haemocoel (Patel et al., 1994). Within the VNC, a pair of immunoreactive mid-line cells of the SOG may be the same cells that have previously been shown to project to the heart (see Patel et al., 1994). LocmiDH-like immunoreactive neurons are also found in the dorsal mid-line of abdominal ganglia, with immunoreactive projections to the anterior median nerve neurohaemal organs. Also prominent in the abdominal ganglia are a bilaterally paired group of posterior lateral NSCs, consisting of 2 3 cells in abdominal neuromeres 1-6, but only 1 2 cells in abdominal neuromere 7. These NSCs (Fig. 23) send their axons out through the sternal nerve of each abdominal ganglion, where they bifurcate at the junction with the transverse nerve. Axons continue medially along the transverse nerve to bifurcate again at the connection with the paramedial nerve. Some processes project over the ventral diaphragm, producing a varicose network, whereas others enter the perisympathetic neurohaemal organ and produce intensely stained neurohaemal terminals. The other branch, at the first junction with the transverse nerve, projects through the link nerve and segmental heart nerve to produce a varicose network in the lateral heart nerve (Patel et al., 1994). Immunoreactive peripheral neurons are also found on the surface of the abdominal tergal

INSECT DIURETIC AND ANTIDIURETIC HORMONES

353

\ VENTRAL~\\

Brain

Jt

\

i 00 microns NCC I

SL

Corpora cardiaca

B

FIG. 22 Locmi-DH-like immunoreactive neurons in the brain of insects. A Lateral view of the brain of L. migramria, illustrating the immunoreactive medial neurosecretory cells (MNCs) with axons projecting through the nervi corpori cardiaci 1 (NCC l) to the neurohaemal storage lobe (SL) of the corpora cardiaca. The glandular lobe (GL) has very few immunoreactive terminals. Redrawn from Patel el al. (1994). B Whole mount of fifth instar R. prolLvus brain (BR) viewed from the dorsal surface. Note the immunoreactive medial neurosecretory cells (MNC) in the protocerebrum and the highly immunoreactive neurohaemal corpora cardiaca (CC). Photomicrograph kindly provided by Victoria Te Brugge.

354

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

A ........

~';

~ f ' / " - "-"-.,-~,,._.j •

>v/o

midline~ l

Segmental boundary

,

,~ r

B

C

FIG. 23 Locmi-DH-like immunoreactivity in the abdominal segment of L. miL,ratoria. A The Locmi-DH-like immunoreactive posterior lateral cells of thc abdominal ganglia project out through the sternal nerve, resulting in neurohaemal areas of the perivisceral organ (PVO) of the transverse nerve, paramedial nerve and lateral heart nerve. Redrawn from Patel et al. (1994). B Whole mount preparation of abdominal ganglion, illustrating the immunoreactive posterior lateral cells (PLC) projecting through the sternal nerve (SN). Note the immunoreactive projections centrally to the neuropile. C Immunoreactive projections along the oviducal nerve (OVN) which branches from the sternal nerve of abdominal ganglion 7. Photomicrographs kindly provided by Rodney Kwok. The scale bar shown in (B) represents: A, 350/~m: B, 50Hm; C, 10#m.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

355

nerves, and at the junction of the transverse and paramedial nerve, with processes running along the nerves. In addition, immunoreactivity is also found in processes from the VIIth abdominal ganglion which project over the lateral oviducts (Fig. 23). In a similar fashion, CRF-like immunoreactivity is found in the MNCs and CC of M. sexta and R. prolixus (Veenstra and Hagedorn, 1991; Emery et al., 1994; Te Brugge et al., 1999). In R. prolixus, the immunoreactive processes also extend through the CC to the aorta (Te Brugge et al., 1999). Within M. chmTestica (laboni et al., 1998), the MNCs are also Locmi-DH-like immunoreactive, although in the larva they lie in a lateral position in the brain. Projections from these NSCs pass to the ventral sector (equivalent to the CC) of the ring gland (a complex structure incorporating the CC, CA and prothoracic gland) in the larva and to the CC in the adult. Also in M. s e x m and R. prolixus (Fig. 24) there is a prominent group of intensely immunoreactive posterior lateral NSCs associated with the abdominal ganglia or neuromeres. In M. sexta larvae, Manse-DPII-like immunoreactivity is found in these posterior-lateral NSCs with their axons projecting out of the ventral nerve into the next transverse nerve, where they branch to produce neurohaemal terminals (Emery et al., 1994). A prominent pair of neurons reactive against Manse-DH in the posterior portion of the terminal abdominal ganglion have axons which project out of the ipsilateral eighth ventral nerve to innervate the rectum and provide neurohaemal-like terminals in the cryptonephric complex (Chen et al., 1994b). In starved larvae, a pair of ventral, paramedial cells are also Manse-DH-like immunoreactive. These cells are believed to be the so-called M4 NSCs, which project anteriorly through the median procurrent nerve to the transverse nerve. These M4 cells also stain positively in adult M. se.vta, as do the posterior lateral NSCs with projections to the transverse nerve via the ventral nerves, where they produce neurohaemal-type terminals (Chen et al., 1994b; Emery et al., 1994). Within R. prolixus (Te Brugge et al., 1999), 10 12 posterior-lateral NSCs of the M T G M project through abdominal nerves 1 and 2 and produce positively stained neurohaemal areas lying on the surface of these nerves (Fig. 24). The immunoreactive processes can also be traced to the body wall, where staining is evident around the spiracles. Fine immunoreactive axons are also seen in abdominal nerves 3 5 and in the genital nerve. The situation in adult M. domestica is somewhat different, in that there is a fused thoracic abdominal ganglion with one pair of Locmi-DH-like immunoreactive cells located ventrally in each neuromere (laboni et al., 1998). lmmunoreactive processes from these NSCs project to the dorsal neural sheath overlying the entire ganglionic mass, where they form an extensive plexus of varicose Locmi-DH-like immunoreactive terminals. Midgut endocrine cells in R. prolixus stain weakly for Locmi-DH-like immunoreactivity (Te Brugge et al., 1999), although midgut endocrine cells, as well as endocrine cells in the ampullae of L. migratoria, stain positively for

356

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

A

MTGM It,

P L C / f " ~',':.:'~:~:

J

\NH

FIG. 24 Whole mount preparations of leucokinin-like and Locmi-DH-like immunoreactivity in two species of bugs, R. prolixus and Oncopellux fasciatus. A and D Leucokinin-like immunoreactivity in the mesothoracic ganglionic mass (MTGM) of R. prolixus. Note the immunoreactive posterior lateral neurosecretory cells (PLC) with central projections as well as projections out through abdominal nerves. Neurohaemal (NH) areas appear on the surface of abdominal nerves. The PLC are also immunoreactive for Locmi-DH-like immunoreactivity (not shown). B and C Higher magnification of posterior lateral neurosecretory cells (PLC) within the equivalent ganglion of O../ilscialus stained for leucokinin-like immunoreactivity (B), and axons from other lateral neurosecretory cells of O. jasciatus which possess apparent reservoirs of Locmi-DH-like immunoreactivity prior to their exit along abdominal nerves (C). These cells in O../asciants do not co-localise Locusfa-DH-like and leucokinin-like immunoreactivity. Photomicrographs kindly provided by Victoria Te Brugge. Scale bars: A. 100#m: B and C, 25#m; D, 15/zm.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

357

CRF-like peptides in A. aegypti (Veenstra et al., 1995) and L. migratoria (Montuenga et al., 1996). With regard to hindgut, Locmi-DH-like immunoreactive processes are present over the entire structure in R. prolixus (Te Brugge el al., 1999). At the ultrastructural level, Locmi-DH-like immunoreactivity is associated with dense-core neurosccretory granules (diameter 450 nm) of the endocrine cells in the ampullae of L. migratoria (Montuenga et al., 1996), and in a variety of morphologically distinct electron-dense granules in R. prolixus (Te Brugge et al., 1999). In R. prolixus, immunogold electron microscopy has been performed on the CC, aorta and abdominal nerves and reveals Locmi-DH-like immunoreactive material overlying the electron-dense neurosecretory granules located within the terminals (Te Brugge et al., 1999). In the CC, two types of granules are evident: round granules of approximately 73nm diameter: and oval granules, 121 x 74nm. In the aorta, only oval granules are immunoreactive, whereas the abdominal nerves again contain round granules approximately 110 nm in diameter. The presence of morphologically distinct immunoreactive granule types suggests the presence of multiple members of the CRF-related family. Structural evidence for this is available from other insects (see section 4.2.2.2) where different structural forms of the CRF-like diuretic peptides are present within an individual species. The distribution of CRF-like immunoreactivity (see above) and the presence of Locmi-DH in the PI and CC of locusts, as revealed by MALDIT O F MS (Matrix-Assisted Laser Desorption Ionisation Time-of-Flight Mass Spectrometer) (Clynen et al., 2001), is consistent with this family of peptides having a neurohormonal function in insects. In addition, Locmi-DH is released by high-K + saline (Audsley et al., 1997b: Clynen et al., 2001) from the CC of L. mi~,ratoria in a Ca:+-dependent manner (Audsley et al., 1997b). Moreover, Locmi-DH can be detected in the haemolymph of fed locusts (Patel et al., 1995), and anti-Locmi-DH antiserum specifically blocks Locmi-DH-induced diuretic activity in vivo, thereby preventing an increase in primary urine production in recently fed locusts (Patel et al., 1995: see section 7.1.3).

6.5

KINJNS

Detailed maps of the distribution of kinin-related peptides within the nervous system of a variety of insect species have been developed using antisera generated against insect kinins, most notably Leuma-K-l. Kinin-like immunoreactivity is distributed extensively throughout the nervous system of a wide range of insect species, and within a range of neuron types, including interneurons, NSCs and possibly sensory cells. Interestingly, although the distribution of kinin-like immunoreactive neurons in abdominal ganglia is very similar in all species examined, there are differences in the type and number of immunoreactive neurons in the brains of different species (see

358

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

N/issel, 1996c). Thus, there is apparently no leucokinin-like immunoreactivity in the brains of A. mell(/'era (Chen et al., 1994a); very few immunoreactive neurons in mosquitoes or flies such as D. melanogaster and Calliphora vomitoria (N~issel, 1993a; Kim, 1998), but large numbers in Phormia terraenova, L. m~ratoria, L. maderae, Spodoplera ]ittorina and R. prolixus (see Nfissel, 1993a; Te Brugge et aL, 2001). Consistent throughout the species examined is the presence of kinin-like immunoreactivity in NSCs and their concomitant neurohaemal areas. Interestingly, within this consistency, two variables have been noted by Chert et al. (1994a). The first is the degree to which the MNCs of the PI of the brain are kinin-like immunoreactive, and the second variable is the location and the extent of immunoreactive neurohaemal organs associated with the NSCs of the abdominal ganglia (see Chen et al., 1994a). This latter variability has been noted earlier for abdominal neurohaemal organs in general (see Grillot, 1983). In L. maderae, approximately 100 MNCs are found to stain intensely for leucokinin-like immunoreactivity in the PI of the protocerebrum (N'assel et al., 1992). These MNCs send axons to the CC via the paired NCC 1. In addition, there are approximately seven lateral neurosecretory cells (LNCs, dorsolateral in the protocerebrum) that are positive for leucokinin-like immunoreactivity, and which project axons via the NCC 2 to the CC. The MNCs and LNCs of L. maderae supply the storage lobes of the CC with large amounts of immunoreactive neurohaemal terminals, lmmunoreactive axons also continue on into the hypocerebral ganglion and then through the oesophageal nerve and recurrent nerve into the neuropil of the frontal ganglion. The immunoreactive axons then continue through the two frontal connectives into the tritocerebrum, where they produce swollen neuropilar terminals. Leucokinin-like immunoreactive MNCs and LNCs are also found in the PI of the cockroach Nauphoeta cinerea, the cricket A. domesticus and the mosquito, A. aegypti (Chen et al., 1994a). In contrast to the situation described above, an anti-Leuma-K-IV antiserum usually does not stain the MNCs of either larval or adult brains of M. sexta, but occasionally weak staining is observed in the larval group lla MNCs and adult IIb MNCs, with projections passing into the CC (Chen et al., 1994b). Within the CC, however, the intrinsic NSCs are intensely immunoreactive, as they are in the cockroach N. cinerea (Chert et al., 1994a,b). In a similar manner, the MNCs of R. prolixus typically do not stain for leucokinin-like immunoreactivity, but in some preparations from insects that have been starved for 10 weeks, there is staining in a subset of MNCs (Te Brugge et al., 2001). Under the conditions tested, there is no leucokinin-like immunoreactivity in the MNCs or NCC 1 of L. migratoria (N/issel, 1993a,b), S. americana, A. melllT"era (Chen el al., 1994a) or M. domestica (Iaboni el al., 1998). Another dipteran, however, P. terraenovcte, does express leucokinin-like immunoreactivity in the MNCs (sce N~ssel and Lundquist, 1991).

INSECT DIURETIC AND ANTIDIURETIC HORMONES

359

As mentioned earlier, the distribution of leucokinin-like immunoreactive neurons in abdominal ganglia is very similar in all species examined, although the distribution of their neurohaemal areas can differ. In all insects studied so far, sets of posterior lateral NSCs of the abdominal ganglia are leucokinin-like immunoreactive (see reviews by Chen et al., 1994a; Nfissel, 1996c). Typically, the abdominal ganglia 1 7 contain a set of leucokinin-like immunoreactive posterior lateral NSCs with axons projecting out through the posterior nerve. Again, typically, the last 1, 2 or 3 abdominal neuromeres lack these positively stained posterior lateral NSCs. The number of posterior lateral NSCs per ganglion varies between insects (from 1 to 7 pairs) and can also vary between ganglia of the same insect. In L. maderae, there are two pairs of leucokinin-like immunoreactive NSCs in each unfused abdominal ganglion, and, in R. prolixus, 6-7 pairs in the abdominal neuromeres of the M T G M . Similar bilaterally paired clusters of varying numbers of cells have been shown for N, cinerea (Chen el al., 1994a), P. americana (Agricota and Brfiunig, 1995), L. m&,raloria (Thompson el al., 1995), S. americana (Chen el al., 1994a), A. domesticus (Chcn et al., 1994a), Gryllus himaculatus (Helle el al., 1995), Agrotis segetum (Cantera eta/., 1992), M. se.vta (Chen et al., 1994a), A. mell(fera (Chen et al., 1994a), A. aeg)7~ti and Phalacrocera replicata (Cantera and N/_issel, 1992: Chen el al., 1994a), D. melanogaster, C, vomitoria, P. terraenovae, M. ~kmwslica (Cantera and Nfisscl, 1992; Iaboni el al., 1998) and R. prolixus (Te Brugge et a[., 2001). Processes from the leucokinin-like immunoreactive posterior-lateral NSC extend out of the posterior nerve in L. maderae and continue into the link nerve and then the dorsal segmental nerve of the next segment (N~ssel et al., 1992). The axons then project to the lateral cardiac nerve on the ipsilateral side. Prior to entering the dorsal segmental nerve, some axons branch and produce varicose processes to the main tracheal trunks near the spiracles. Also, there is a branch from each axon from the link nerve, which passes to the transverse nerve of that segment, thereby reaching the pcrisympathetic neurohaemal organ. Immunoreactive varicosities and terminals are found in the processes near the spiracles, in the lateral cardiac nerve, and in perisympathetic organs, indicative of neurohaemal release sites. A similar peripheral distribution of lcucokinin-like immunoreactive processes and neurohaemal sites has been described for N. citwrea (Chen et al., 1994aL L. mi~,ratoria (Thompson eta/.. 1995), A. domesticus (Chen el al., 1994a) and G. himaculatus (Helle el al., 1995). with additional neurohaemal sites in the sheath of the root of the dorsal nerve of the next posterior segment in the latter two species. Within larval 4. segettml, the abdominal posterior lateral NSCs send axons to the alary muscles, spiracles and transverse nerves; and, in the adult, the processes in the transverse nerve project to the neural sheath on the dorsal surface of the abdominal nerve cord and form an extensive immunorcactive plcxus (Cantera et al., 1992). The projections from abdominal NSCs in larwd and adult M. se.vta pass out of thc posterior nerve and form varicose processes in the neurohaemal region of the transverse nerve of the next posterior segment

360

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

(Chen et al., 1994a). The processes are more intensely stained and greater in number in the adult. In A. mell(/'era, the neurohaemal sites are located in the segmental nerve root near the ganglion, whereas in larval D. melanogaster, C. vomiloria and P. terraenovae the leucokinin-like immunoreactive processes supply terminals to the abdominal body wall muscles: and in the adult, they produce neurohaemal sites on the pericardial septum (Cantera and Nfissel, 1992), Within the M. domestica adult, there are seven to ten pairs of ventrolateral leucokinin-like immunoreactive neurons and a pair of dorsomedial neurons in the abdominal neuromeres of the thoracico-abdominal ganglionic mass (Iaboni et al., 1998). Seven or eight of the ventrolateral neurons have axons which project out of the two fused lateral abdominal nerves, where they produce immunoreactive neurohaemal sites. The dorsomedial neurons project out the median abdominal nerve and again result in immunoreactive varicosities on the nerve (laboni et al., 1998). In A. aegypti, the release sites appear to be on the surface of the segmental nerves (Chen et al., 1994a) and in A. mell(/i~ra there is a bulbous neurohaemal organ on the surface of the ventral nerve root near the ganglion (Chen et al., 1994a). Abdominal nerves 1 and 1I possess leucokinin-like immunoreactive neurohaemal sites originating from the posterior-lateral NSCs of the mesothoracic ganglionic mass in R. prolixus (Te Brugge eta/., 2001). Projections from these neurons also extend to the lateral margin of the abdomen, producing a complex branching pattern associated with the tergal sternal muscles (Te Brugge et al., 2001). In addition, there are paired multipolar peripheral neurons immunoreactive for leucokinins on branches of peripheral nerves in R. prolirus. The presence of kinins in apparently homologous posterior-lateral NSCs of abdominal ganglia of a wide variety of insect species has been suggested to be a highly conserved feature in insects (Chen et al., 1994a). The distribution of the abdominal neurohaemal areas associated with such cells reveals some variation, although the extensive neurohaemal areas produced are consistent with kinins being released into the haemolymph as neurohormones, Despite the presence of leucokinin-like immunoreactivity in NSCs and neurohaemal organs throughout the insect species studied (and by inference their potential as neurohormones), it is still worth exploring the possibility of innervation and release over target organs. Initial studies in L. maderae found no such innervation to the MTs, hindgut or fat body (N/issel et al., 1992), but leucokinin-like immunoreactive processes do project to the foregut (crop) of this species, and immunoreactive bipolar and tripolar neurons are also present in the crop. Using a different antiserum, one generated against Leuma-K-VIII, Meola et al. (1994) found projections of leucokinin-like immunoreactive processes from the posterior nerve of the terminal abdominal ganglion, which might indeed project to the hindgut. In R. prolixus, no kinin-like immunoreactive processes are seen projecting from the stomatogastric nervous system to the foregut, although the midgut does contain leucokinin-like immunoreactive endocrine cells, and leucokinin-like immunoreactive processes produce

INSECT DIURETIC AND ANTIDIURETIC HORMONES

361

a plexus of staining over the posterior midgut and hindgut (Te Brugge et al., 2001). Similarly, in A. aegypti and the crane fly, P. replicata, paired leucokininlike immunoreactive neurons in the terminal ganglion project to the hindgut (Cantera and N'assel, 1992). Within the terminal abdominal ganglion of M. s e x t a larvae, there are two leucokinin-like immunoreactive neurons (ventral unpaired median (VUM) neurons) with kinin-like immunoreactive axons projecting out of the terminal nerves, innervating the posterior hindgut and also forming an apparent neurohaemal area in the cryptonephric complex (Chert et al., 1994b). These two VUM neurons could not be distinguished in the adult. lmmunoreactive staining of NSCs and neurohaemal areas is certainly consistent with the antigen being a neurohormone, released from these structures into the haemolymph. This is further corroborated by the demonstration by HPLC and RIA or M A L D I - T O F of the presence of the kinins within neurohaemal areas and their release from such structures. Thus, all eight isoforms of the leucokinins have been shown to be associated with the CC and lateral heart nerves in L. maderae (Winther et al., 1996), and various P. americana kinins are associated with the different neurohaemal areas of that cockroach (Predel el al., 2001). In addition, in A. domesticus and L. maderae, kinin-like immunoreactivity has been found in the haemolymph, and in both these species Ca 2 ~dependent release of kinin-like material from the CC has been demonstrated in response to depolarisation with high K + saline (Muren et al., 1993: Chung et al., 1994: see section 7.1.4).

6.6

CAP2b/PERIVISCEROKININS

As mentioned earlier (see section 4.2.5), Manse-CAP2b, was originally sequenced from M . s e x t a (Huesmann el al., 1995) and belongs to a group of at least five peptides (CAPs) that collectively influence cardiac function and patterns of behaviour in M . se_vta (Tublitz et al., 1991). These peptides may be derived from midline NSCs that project to the abdominal transverse nerve perisympathetic organs (see Wegener et al., 2001). Manse-CAP2b belongs to a family of peptides which have been isolated and sequenced from cockroaches and locusts, and found within the D. m e h m o g a s t e r genome, and which share the common C-terminus PRVamide (see Table 4; section 4.2.5). It is of some interest that a sequence comparison of the cockroach PVKs (see Table 4: Wegener el al., 2002) illustrates similar N-terminal sequences, but variable C-terminal sequences. This is unusual for neuropeptide families, which tend to share the C-terminal sequence that typically contains the active core. Thus, while the PVKs (and Manse-CAP2b) that share the PRVamide C-terminus are biologically active on MTs, there is no evidence that other members of the PVK family are also active. For example, 0.1/IM Peram-PVK-1 has no effect on A. domesticus tubules, whereas 0.03#M Peram-PVK-2 has significant

362

G . M . COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

diuretic activity (G. M. Coast, unpublished observations). Since the immunohistochemical staining of insect nervous systems has used antibodies generated against the different PVKs, it is difficult to generalise on the distribution of those PVKs that may act as diuretic/antidiuretic peptides. With that caveat in mind, however, PVKs have been shown to be present in the anterior median nerve and abdominal perisympathetic organs of P. americana using MALDITOF MS (Predel, 2001). The ratio of peptides present was 2:1:0.2 for PeramPVK-I, Peram-PVK-2 and Leuma-PVK-2 respectively, although MALDI-MS is notoriously non-quantifiable. In contrast, the posterior median nerves do not have any detectable PVKs, and neither do the thoracic perisympathetic organs, nor the retrocerebral complex. Using antisera specifically generated against either Peram-PVK-1 or Peram-PVK-2, Eckert el al. (1999) and Predel el al. (1998) found these peptides to coexist in immunoreactive clusters of 6 10 NSCs in abdominal ganglia. In the five unfused abdominal ganglia, the cell bodies occur as three clusters (Ci C3) and as two cells ventrolateral to the C~ cluster. Axons from these cells project through the anterior median nerve to the perisympathetic neurohaemal organs, where they produce intensely stained neurohaemal areas. Processes continue on to innervate the hyperneural muscle, and, via the link nerve to the segmental nerve, project to the heart, alary muscles and segmental vessel (Eckert et al.. 1999). The fused terminal ganglion reflects a similar pattern of immunoreactive cells in only the seventh neuromere (Predel et al., 1998; Eckert el al., 1999). The more extensive study, using Peram-PVK-1 antisera (Eckert el al., 1999), reveals an intrinsic neuronal network of immunoreactive cell bodies and processes within the brain, SOG and metathoracic ganglion, and local interneurons in the proto- and tritocerebrum, again indicating that these neuropeptides play a central role, as well as acting as neurohormones. It should be remembered, however, that Peram-PVK-I is not known to have activity on MTs. A peptide with similar chromatographic properties and biological activity to M. sexta CAP2~, has been isolated from D. melanogaster (Davies et al., 1995), and two related peptides are predicted from the fruit-fly genome (Vanden Broeck, 2001; Wegener et al., 2002). CAP2h-like bioactivity in D. melanogaster has been found in a set of midline mesodermal cells that have axonal-like processes in the transverse nerve, suggesting a secretory function for the cells (see Huesmann et al., 1995). The PVKs, as defined by Wegener et al. (2002) are among the most abundant peptides stored in the abdominal perisympathetic organs of P. americana, as revealed by M A L D I - T O F MS (Wegener et al., 1999; see Wegener el al., 2001). Approximately 6 pmol Pea PVK-I is stored in these organs in P. americana. Similarly, the distribution and abundance of Manse-CAPeb in M. s e x t a abdominal perisympathetic organs is comparable with that of the cockroach (Wegener et al., 2001). Evidence for release of PVKs from these organs comes from K + depolarisation experiments. Thus, approximately 28 50fmol Peram-PVK-1 is releasable in a Ca2+-dependent manner, fl-om an abdominal nerve cord with intact perisympathetic organs

INSECT DIURETIC AND ANTIDIURETIC HORMONES

363

that has been exposed to a saline containing 100 mM K + (Wegener el al., 2001). To date, however, PVKs have not been shown to be present in the haemolymph. 6.7

( ' A L C I T O N I N - L I K E PEPTIDES

Antisera to the insect calcitonin-related diuretic peptides have only recently become available, and there are no published reports on their distribution. Preliminary results (V. A. Te Brugge, personal communication), using antisera against Dippu-DH31 (Furuya el al., 2000b; produced by V. C. Lombardi and D. A. Schooley), indicate neurohaemal sites on the abdominal nerve in D. punctata. Within R. prolixus, Dippu-DH31-1ike immunoreactivity is found throughout the CNS, with fine processes observed in some of the nerves of the M T G M . Weak staining of neurohaemal sites is also observed on abdominal nerve 2. Future work is needed to create a detailed map of the distribution throughout a variety of insect species. 6.~

ION TRANSPORT F'EFq-IDE

Neurosecretory cells immunoreactive to an antiserum raised against Orconectes limosus C H H originate in frontal mediolateral brain areas of L. migratoria and have branches within the superior median protocerebrum. Their axons run through the NCC I to the CC and then to the CA in the nervi corporis allati I (H. Dircksen, personal communication). The discrete localisation suggests this antisermn recognises ITP rather than ITP-L, and ITP m R N A is known to be expressed in the brain and CC (Meredith el al., 1996). Similar results (M. Fuse, personal communication) were obtained with an Schgr-ITP antiserum, which detected small numbers of immunoreactive NSCs in the P1 and CC of S. gregaria. As with the Orcli-CHH antiserum, it is not clear whether the ITP antiserum also recognised ITP-L. Additionally, there are three to ten C H H immunoreactive NSCs in each thoracic and abdominal PVO of L. migratoria and sometimes also in the link nerves (H. Dircksen, personal communication). Meredith et al. (1996) did not detect ITP m R N A in the VG of S. gregaria, but the preparation that they used may not have included the PVOs. The presence of C H H (ITP)-like immunoreactive material in the CC is consistent with it being released into the circulation although, to date, release of ITP from the CC has not been demonstrated and it has not been definitively identified in hacmolymph. 6.9

( O-LOCALISATION

The posterior lateral NSC which stain for kinin-like immunoreactivity within the abdominal ganglia of a variety of insects, have an identical cell body position and morphology to those that stain for CRF-like D H immunoreactivity.

364

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

Double-label immunohistochemistry confirms that in L. migratoria (Thompson et al., 1995), M. sexta (Chen et al., 1994b) and R. prolixus (Te Brugge el al., 2001), kinin-related and CRF-related diuretic peptides are co-localised to the same posterior lateral NSCs in the abdominal ganglia and their neurohaemal endings (in M. sexta only one (L3) in each of the two pairs of (L3 4) of leucokinin-like immunoreactive cells co-localise Manse-DH). This discovery led Thompson et al. (1995) to suggest that co-localisation within the posterior lateral NSCs of abdominal ganglia may be a conserved feature among insects. However, laboni et al. (1998) could not detect co-localisation of these two peptide families in either brain or abdominal neuromeres in M. domeslica. Moreover, despite co-localisation of Locmi-DH-like and Leuma-Kl-like peptides in the posterio>lateral NSC in the blood-feeding bug, R. prolixus (Fig. 24), co-localisation is not evident in the milkweed bug, Oncopehus Jitsciatus (V. A. Te Brugge, personal communication). Thus, as commented on by Te Brugge el al. (2001), co-localisation of CRF-related and kinin-like diuretic peptides in posterior lateral NSC of abdominal ganglia may well be a 'general phenomenon in insects' (Thompson el al.. 1995), but not necessarily a universal one. It is unclear if this co-localisation extends to the MNCs in the brain. Fewer insects have been examined for CRF-related DH immunoreactivity, and kininlike immunoreactivity has only been shown consistently in the MNCs of L. maderae, N. cbwrea, A. domesticus and A. aegypti. Unfortunately, corresponding double labelling experiments with the CRF-related diuretic peptide antisera have not been reported for these species. Within R. prolixus and M. sexta, although the MNCs in the brain stain intensely for the CRF-like DH immunoreactivity, the staining for Leuma-K-like immunoreactivity is seen in only a few preparations, which makes double-labelling immunohistochemistry hard to interpret (see Chen et al., 1994b: Te Brugge et al., 2001). In those R. prolixus preparations where there are Leuma-K-I-like immunopositive medial NSCs, these cells are single labelled for Leuma-K-I-like immunoreactivity (Te Brugge et al., 2001). However, other cell types (approximately 24-28 cells in the brain, two to four in the prothoracic ganglion, 12 18 in the M T G M ) are double-labelled for Leuma-K-l-like and Locmi-DH immunoreactivity (Te Brugge et al., 2001). Within L. mi,~ratoria, the posterio~lateral NSCs of abdominal ganglia which stain for Leuma-K-l-like immunoreactivity label with antisera to Locmi-DH-like immunoreactivity (see above), but also with lysine vasopressin (LVP, Thompson et al., 1995). Similar results are found for posterior lateral abdominal NSCs in L. maderae, N. cinerea and A. domesticus (Niissel el a[., 1992). N'assel et al. (1992) consider that the co-localisation with LVP-like immunorcactivity in L. maderae is due to cross-reaction bctwcen the LVP antiserum and the endogenous leucokinins, since preabsorption of the LVP antiserum with Leuma-K-1-BSA conjugate abolishes tissue staining. However, staining in L. migratoria is not blocked by such preabsorption

INSECT DIURETIC AND ANTIDIURETIC HORMONES

365

(Thompson et al., 1995), and these authors conclude, along with additional evidence from H P L C and immunoassay, that the abdominal ganglia in L. mi~raloria contain three distinct epitopes reacting with the three antisera. The medial NSCs in the PI of L. migratoria which are Locmi-DH-like immunoreactive are also immunoreactive to antisera generated against ovary maturing parsin, a putative neurohormonc which stimulates oogenesis in locusts (Tamarelle e l al., 2000). Other forms of co-localisation have been shown in P. americamt, where Peram-PVK-1, Peram-PVK-2 and Peram-pyrokinin-5 appear to coexist in the abdominal nerve cord and interneurons of the brain (see Wegener el al., 2001). The expression of these peptides starts at the same time in embryonic development, leading Wegener el al. (2001) to suggest that both PVKs and Peram-pyrokinin 5 are encoded together on a single gene in P. americmza as they have been shown to be in D. melam)¢asler (Vanden Broeck, 2001: see section 4.2.5). The M4 NSC in M. sexta appears to co-localise Manse-DHlike, PBAN-like and CCAP-Iike immunoreactivity (Chen el al., 1994b). Also, the SOG cells in L. migraloria which stain for Locmi-DH-like immunoreactivity, and which appear to be the same cells that have previously been shown to project to the heart, would be expected to co-localisc with FMRFamide-like immunoreactivity (Patel el al., 1994). Of considerable interest is the study by Zitnan el a/. (1995), which reports a map of NSCs in the brain of M. sexta that stain for prothoraciotropic hormone, bombyxin, allatotropin, allatostatin, DH, eclosion hormone and proctolin. Such a map allows inferences to be drawn about possible co-localisation with other neuropeptides when performing immunohistochemical studies in this species. Although there is no evidence for the co-localisation of diuretic peptides with serotonin within NSCs or neurohaemal terminals, these two groups of diuretic factors are co-localised in a number of interneurons within the various ganglia of R. prolixus (Te Brugge et al., 2001 ). It also seems reasonable to conclude that the wtrious diuretic factors might be expected to be co-localised within interneurons of other insects as well. To sum up, a common, although not universal feature in insects (see laboni et al., 1998; Te Brugge et al., 2001) is the co-localisation of kinin-like and C R F like diuretic peptides in the abdominal ganglia posterior lateral NSCs and their neurohaemal terminals. Such co-localisation does not seem to occur in the M N C s in the PI of insects studied, although a broader range of species needs to be examined. Other peptide phenotypes, for which there is no evidence of diuretic activity, may also be co-localised with the diuretic peptides in NSCs. Serotonin does not appear to be co-localised within NSCs that express diuretic peptides. It would appear that there is sufficient flexibility in NSC types for the kinins, CRF-like diuretic peptides and serotonin to be released independently into the haemolymph, or, at least in the case of kinins and CRF-like diuretic peptides, to be co-released from the same NSCs. In addition, there is the possibility that

366

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

these factors can be co-released or released independently within the neuropil of the central ganglia where they might modulate the activities of central neurons, which might well be involved in behaviours other than diuresis. The presence of immunoreactivity in processes lying over the digestive tract certainly implies physiological activities for these diuretic factors, which goes beyond diuresis itself, and suggests a broader role for them in feeding behaviour as a whole. It is interesting to note the antifeedant activities reported for Manse-DH in H. virescens larvae (Keeley et al., 1992), Manse-DPll in M. sexta neonates (Ma et al., 2000) and Locmi-DH in fifth-instar L. mi~gratoria nymphs (Coast and Goldsworthy, 1997; see section 9.2).

7

Physiological relevance

7.1

7.1.1

C I R C U L A T I N G LEVELS IN R E L A T I O N TO P H Y S I O L O G I C A L STATUS

hTtroduction

The identification of factors with diuretic or antidiuretic activity using in vitro bioassays is but a first step towards showing that they have a functional role as circulating neurohormones. Importantly, it is necessary to show that they are released from neurohaemal sites into the circulation and that the haemolymph concentrations achieved are appropriate for mediating a physiological response. This necessitates measuring haemolymph concentrations and investigating the effects of injecting putative hormones into intact insects (see section 7.2.3). Unfortunately, little information is available in either of these areas and, to date, there is no evidence to show that ITP, the only well characterised stimulant of hindgut reabsorption, has a hormonal function. 7.1.2

Serotonin

Feeding induces release of serotonin from an extensive network of serotoninlike immunoreactive neurohaemal areas and processes that arise primarily from the M T G M and supply the dorsal and ventral integument (Orchard et al., 1988; Lange el al., 1989). Within minutes of the onset of feeding in fifthinstars, serotonin is released from these abdominal nerve processes both at the integument, probably participating in cuticular plasticisation (Reynolds, 1974, Orchard et al., 1988), and into the haemolymph, resulting in a 15-fold increase in serotonin concentration (Lange et ell., 1989). The titre of serotonin within the haemolymph of fifth-instar R. prolixus is shown in Fig. 25. As can be seen, the concentration of serotonin in the haemolymph of unfed insects is low (6.8 nM), but rapidly increases following the onset of feeding, peaking to 115 nM within 5 min. Thereafter the titre declines, and remains low, but still elevated over unfed controls, for 24h. Adult male R. prolixus also have

INSECT DIURETIC AND ANTIDIURETIC HORMONES

367

120

100 c-

tO

8o e-

E

60

o

E m

-r

40

20

[

I

I

I

I

I

I

0

5

10

15

20

25

30

//"

I

60

Time after c o m m e n c e m e n t of feeding (min) FIG. 25 The titre of serotonin in the haemolymph of fifth instar R. proli.vu.~ at varying times after the commencement of Feeding on rabbit blood. Time zero represents unfed insects. Symbols represent means ±S.E. Redrawn from Lange ~'f al. (1989).

elevated serotonin levels within 3 min of the onset of feeding, with a peak concentration somewhat less than in fifth-instars, but with the same pattern of sustained elevation and decline towards baseline levels (Barrett et al., 1993). The serotonin that appears in the haemolymph is not derived from the blood meal, since similar titres are found when fifth-instars are fed on an artificial diet. Rather, as referred to above, the serotonin appears to be released from the peripheral nervous system, since there is a depletion of the serotonin-like immunoreactive staining of abdominal neurohaemal areas and processes over the body wall during feeding (Lange el al., 1988; Orchard el ol., 1988), and a depletion in content of serotonin as determined by HPLC with electrochemical detection. In addition, serotonin is released from these neurohaemal areas (Lange el al., 1988) in response to a depolarising stimulus (high K" saline). The peak titre of serotonin in the haemolymph of R. prolixus is sufficient to induce cyclic A M P elevation and fluid secretion from the MTs it7 vilro (Maddrell et al., 1971; Barrett ~,t al., 1993), and anterior midgut (crop)

368

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

(Farmer et al., 1981; Barrett et al., 1993), leading to the conclusion that serotonin is a neurohormone in R. prolixus, released by the natural stimulus of feeding, and controlling diuresis. It has been proposed that serotonin has a broad range of peripheral roles that result in it being a co-ordinator of feedingrelated activities in R. prolixus, including plasticisation, salivation and gut motility, as well as fluid secretion from crop and MTs (see Orchard et al., 1988; Barrett and Orchard, 1990; Barrett et al., 1993; Orchard and Brugge, 2002). 7.1.3

A VP-like insect D H

Changes in the haemolymph titre of AVP-IDH in adult L. migratoria over 24 h are shown in Fig. 26, along with an index of primary urine production based upon measurements of amaranth clearance (Picquot and Proux, 1987). The concentration of AVP-IDH fluctuates between 0.04 and 0.4 nM and, in insects that were fed daily, there is a peak at 12:00 h, which coincides with a peak in primary urine production. At other times, however, the two parameters vary independently of one another, which argues against any causal relationship. In contrast to the data of Picquot and Proux (1987), Baines et al. (1995) found very little (c.2 pM) AVP-like material in the circulation. Arguably, AVPIDH may be neither a diuretic nor a neurohormone: it has no effect on secretion by isolated tubules (see section 5.2.2); the haemolymph titre is low for a peptide hormone; and it is debatable whether there are neurohaemal sites for its release (see section 6.1.3). 7.1.4

C R F - r e l a t e d peptides

Using a highly sensitive RIA, Audsley et al. (1997b) showed that Locmi-DH is mainly stored in the brain and CC, with small amounts (a few 100 fmol) in thoracic and abdominal ganglia. In addition, a small amount of Locmi-DH is present in endocrine cells of the MT ampullae (Montuenga et al., 1996). Feeding causes an immediate release of Locmi-DH from the CC, which can be mimicked by depolarising the glands in high K + saline (Audsley et al., 1997b). The haemolymph titre of Locmi-DH increases throughout the meal (see Fig. 27: Audsley et al., 1997a), which normally lasts about 15 20rain. During this time, the amount of peptide stored in the CC returns to pre-feeding levels, reflecting its continued export from the brain, and Locmi-DH synthesis by MNC may be upregulated (Audsley et al., 1997b). Locmi-DH immunoreactive axons follow a spiral course within the brain (see Fig. 22) and contribute to the 'neuropilar reservoir' described by Highnam and West (1971). This reservoir empties of neurosecretory material within 5 rain of the onset of feeding (Highnam and West, 1971), which possibly reflects the export of Locmi-DH to the CC.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

6-

369

Fed

-0.6

eO

-0.5 ~> <

,-o

E

~4-

-0.4

-"I"

t.-. .:.-

-0.3

~:

-0.2 S"

~'2E

,--

-0.1

1

i

t

i

i

i

08.'00

12:00

16:00

20:00

24.'00

04:00

0.0

Time (h)

6-

Starved

tO

< E ~4-

2

e-

-0.3

~'2-

-O.2

~:

E 'r" 4~.1

i

i

i

t

i

i

08:00

12:00

16:00

20:00

24:00

04:00

0.0

~ m e (h)

FIG. 26 The haemolymph titre of AVP-IDH (dotted line) compared with an index of primary urine flow obtained from measurements oF amaranth clearance (solid line) over 24h in adult L. migratorkL Data are shown for two groups of locusts that had either been fed or starved for the preceding 48 h. Both groups were fed fi'esh grass at 08:00 h on the day of the experiment (vertical arrows), and there are coincident peaks in the titre of AVP-IDH and primary urine flow at c.12:00h in the fed group. At other times, however, there is no evidence of any correlation between the two parameters. Redrawn from Picquot and Proux (1987).

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

370

3-

e-

2-

:I:

13

Feeding

!

E

com

n

¢,,,) 0 1,_1

0

i O

i 10 Time

i 20

(min)

FIG. 27 The change in haemolymph titre of Locmi-DH during a 20rain meal. The horizontal arrow points to the haemolymph titre of Locmi-DH in locusts that were starved overnight and the vertical arrow shows when the animals were fed fresh grass. Redrawn from Audstey et al. (1997a).

About 1 pmol of authentic Locmi-DH is released from the CC within 5 rain of the onset of feeding (Audsley et al., 1997b), which is sufficient to stimulate M T secretion in vivo (Patel et al., 1995). Locmi-DH is therefore released at the right time and in sufficient amounts to function as a neurohormone controlling the postprandial diuresis of locusts.

7.1.5

Kinins

Kinins have been quantified by RIA in the brain, CC-CA, VNC and haemolymph of L. maderae and A. domesticus (Muren et al., 1993; Chung et al., 1994). The highest amounts are in the CC, from where they are releasable by K + depolarisation. Circulating levels range from 0.5 to 2.8 nM, which are typical of many neurohormones, but are several times higher than those needed to stimulate cricket tubule secretion (ECs0s ranging from 0.02 nM to 0.32 nM; Coast et al., 1990a) and cockroach hindgut contractions (threshold values ranging from 0.04 nM to 0.27 nM; Holman et al., 1990). However, in neither of these studies was haemolymph fractionated prior to RIA, and kinin levels could have been overestimated if the antisera recognised inactive fragments. Circulating levels increase 10-fold in crickets starved for 48 h without access to water, due partly to a 50% reduction in haemolymph volume, but then fall when the insects are fed (Chung et al., 1994). It is therefore unlikely that kinins

INSECT DIURETIC AND ANTIDIURETIC HORMONES

371

have a role in co-ordinating feeding-related processes such as diuresis in crickets (see Chung et al., 1994). 7.1.6

CAP_, peptides/periviscerokinins

In addition to their cardioacceleratory actions, CAP2 peptides are implicated in the control of gut emptying behaviour in fifth-instar M. sexta larvae, which takes place on day 4, just prior to the onset of wandering behaviour (Tublitz at al., 1992). By measuring hindgut contractions in vivo, Tublitz et al, (1992) showed that an increase in the frequency of contractions coincides with a peak in haemolymph CAP2 activity, while the amount stored in the VNC decreases. The haemolymph titre at the time of gut emptying is too low, however, to account for the observed increase in contractions, and CAP2 peptides appear to be released locally on to the hindgut from axons projecting in the proctodeal nerve (see Tublitz et al., 1992). It is not known which of the CAP2s regulate gut emptying, or whether they are involved in other aspects of the excretory process. Manse-CAP2b has no effect on secretion by adult tubules (N. Tublitz, personal communication), but it appears not to have been tested on larval tubules or on the cryptonephric complex. 7.2 7.2.1

I N T E R F E R I N G WITH THE N A T U R A l . TITRES OF C I R C U L A T I N G FACTORS

lntro&tction

Several groups have investigated the effects of injecting putative DHs into insects, but great care needs be taken in interpreting the results of these studies because of the difficulty in distinguishing between physiological and pharmacological effects. In some cases, very large doses of peptide factors need to be injected to see an effect. Although this may be necessary because of their rapid degradation in vivo (see section 7.3), it will result in haemolymph concentrations far higher than those achieved in life. A more robust approach to determining physiological function is to investigate the effect of selectively blocking hormone activity in vivo, and this has been used to show that serotonin, Locmi-DH and Musdo-K function as circulating hormones to control diuresis (see below). 7.2.2

R. prolixus

The neurotoxic analogue of serotonin, 5,7-DHT, has been used extensively to study the morphology and physiology of serotonergic neurons (for references, see Cook and Orchard, 1993b). Injection of 5,7-DHT into R. prolixus enhances serotonin-like immunoreactive staining of the cell bodies and axons of the five DUM neurons in the M T G M and severely depletes the serotonin-like immunoreactive neurohaemal complex on the abdominal nerves (Orchard et

372

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

al., 1989). Thus, 5,7-DHT appears to be a useful pharmacological tool for depleting peripheral serotonergic stores in R. prolixus. In R. prolixus tissue in vitro, 5,7-DHT competitively inhibits the uptake of tritiated serotonin by serotonergic neurohaemal areas, causes a significant release of previously accumulated tritiated serotonin in nominally Ca2+-free saline, and reduces the synthesis of serotonin (Cook and Orchard, 1993b). Injection of 5,7-DHT reduces, within 24h, the staining intensity of serotonin-like immunoreactive processes on the salivary gland ducts and abdominal nerves of R. prolixus and reduces the serotonin content (by 92% and 77% respectively) of these tissues as measured by HPLC with electrochemical detection, but has no significant effect on the serotonin content of the CNS (Cook and Orchard, 1990). Individuals of R. prolixus which have been injected with 5,7-DHT twentyfour hours previously, ingest a blood meal that is significantly smaller than that of control insects and they do not appear to undergo cuticle plasticisation. In a similar fashion, Maddrell et al. (1993a) injected fifth-instar R. prolixus with 5,7-DHT twenty-four hours before feeding, and found that the insects fed abnormally, but also that diuresis was either delayed or absent altogether, and that the pattern of plasticisation was abnormal. These results are consistent with serotonin being a DH in R. prolixus that is also involved in other feedingrelated activities such as salivation and cuticle plasticisation. The value of 5,7D H T as a pharmacological tool is further enhanced by the discovery that the effects of 5,7-DHT injection into R. prolixus are reversible with time (Cook and Orchard, 1993b). Thus, between one and two weeks following injection, the immunoreactive staining is similar to the control insects, the insects ingest a normal size blood meal, and plasticisation is apparent. Diuresis, however, is delayed 1 2h in these insects.

7.2.3

Lepidopteran insects

Injections of Manse-DH (Kataoka et al., 1989) and Manse-DPII (Blackburn and Ma, 1994) increase excretory water loss in decapitated P. rapae and M. sexta moths, respectively, and the estimated ECsos (Manse-DH, 20fmol; Manse-DPIl, 8pmol) suggest that these are physiological effects reflecting stimulation of MT secretion. A dose of 20 pmol Manse-DH was also shown to double excretory water loss in post-feeding, pre-wandering fifth-instar M. sexta larvae (Kataoka et al., 1989), and this too is probably within the physiological range, given the large haemolymph volume of these insects. Interestingly, these authors reported that Manse-DH also increased water loss across the larval body wall. This observation does not appear to have been followed up, but could be due to the opening of cuticular pores (Hadley et al., 1989). Manse-DH (20 100pmol) also increases faecal water loss in fifth-instar larvae of H. virescens (Keeley et al., 1992), as do 50 pmol injections of Helze-K-I, -II and -llI (Seinsche et al., 2000), which is consistent with the

INSECT DIURETIC AND ANTIDIURETIC HORMONES

373

stimulation of MT secretion. Interestingly, a much higher (pharmacological?) dose of Manse-DH (500 pmol) reduces the amount of faecal material produced and lowers its water content (Keeley et al., 1992), which could be evidence for the stimulation of fluid uptake from the cryptonephric complex (see section 5.4). 7.2.4

D i p t e r a , insects

Cady and Hagedorn (1999a) used THO loss as a measure of urine output from blood-fed, decapitated A. aegypti and showed this was increased by injections of diuretic peptides (Aedae-Ks, Culsa-Ks and Culsa-DP) and serotonin (EDsos ranging from 0.04 pmol for Aedae-K-3 to 42 pmol for serotonin). Interestingly, although Aedae-K-2 increases urine output in vivo (Cady and Hagedorn, 1999a), it has no effect on tubule secretion in vitro (Veenstra at al., 1997). Similarly, fraction I from a partially purified head (Petzel et al., 1986) stimulates urine output, but has no effect on tubule secretion (Wheelock et a/., 1988), and both may act by inhibiting fluid uptake from the hindgut, although this has not been investigated. In a similar study, (Coast, 2001a), showed that injections of Musdo-DP and Musdo-K increase urine output (measured by THO loss) in houseflies, with Musdo-K being the more effective, which is consistent with its greater activity in vitro (laboni et al., 1998). These peptides are known to act synergistically in vitro (Holman et al., 1999; see section 8.3), but this could not be demonstrated i , vivo (see Coast, 2001a), and it is conceivable that attempts to demonstrate this effect in vivo are confounded by the very rapid degradation observed for kinins in the circulation (see section 7.3.4). This might also explain the low potency of Musdo-K i , vivo (estimated ECs0 0.02 to 0.2/~M: Coast. 2001a) compared with in vitro (ECs0 0.65 nM; Holman et al., 1999). Hypervolemia stimulates urine output in houseflies and diuretic activity is detectable by bioassay in the haemolymph of insects injected with 1 txL saline (Coast, 2001a). Excretory water loss is correlated with the degree of abdominal distension (G. M. Coast, unpublished observations), which is consistent with a DH being released in response to stretching the tergal sternal muscles, as described in R. prolixus (see section 3.2). Importantly, the diuresis induced by hypervolemia is reduced by c.70% in flies previously injected with an antiserum that specifically recognises Musdo-K (Coast, 2001a), providing strong evidence for this being a t\mctional DH that is released in response to abdominal distension. 7.2.5 Orthopteran insects The rate of clearance of injected amaranth is increased three-fold in recently fed insects (Patel et al., 1995), reflecting the stimulation of MT secretion during the postprandial diuresis (Mordue, 1969). Injections of kocmi-DH mimic this

374

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

effect (Patel et al., 1995), and the estimated EDs0 (l pmol) is equivalent to a haemolymph concentration of 4nM, which is comparable with its potency in vitro (ECso 1.9nM; Coast, 1995). Prior injection of locusts with LocmiDH antiserum specifically blocks the activity Locmi-DH and prevents the postprandial increase in amaranth clearance, thereby providing unequivocal evidence for Locmi-DH being released into the circulation to stimulate diuresis (Patel et al., 1995). The haemolymph volume of locusts injected with 23 pmol Locmi-DH fails by c.20 # L within 5 rain (Patel et al., 1995). This is consistent with stimulation of MT secretion, although the required rate of urine production (4 # L min -1) is considerably higher than seen in vitro (c.0.5#Lmin l) or in vivo (0.3 0 . 7 g L m i n -l) (Mordue, 1972; G. M. Coast, unpublished data). Bernays and Chapman (1974) reported a 13% (18#L) fall in the haemolymph volume of fifth-instar L. migratoria nymphs that had taken a full meal, which they attributed to secretion of copious amounts of saliva. Since Locmi-DH is released on feeding (see section 7.1.3), it may have a role in co-ordinating other feeding-related activities such as salivation. This has not been investigated in locusts, but Zoone-DH has been shown to increase contractions of the salivary reservoir in R. prolixus (Orchard and Brugge, 2002). 7.3 7.3.1

DEGRADATION AND INACTIVATION

Introduction

The activity of circulating hormones is limited in time and space by their inactivation and/or removal from the haemolymph. This can involve their excretion, degradation by either circulating or membrane-bound enzymes, and internalisation of the hormone when bound to its receptor, After a blood meal, R. prolixus haemolymph contains about 0-50% more diuretic activity than is required to maximally stimulate tubule secretion, but after 3~4h the stimulus for release is removed and the hormone disappears from the circulation, which coincides with a rapid reduction in primary urine flow from 400nLmin -t to < 5 0 n L m i n I (Maddrell, 1964a). Although it is now clear that several factors contribute to the diuretic activity described by Maddrell (1964a), the proposal that diuresis be terminated by removing DH from the circulation remains valid. 7.3.2

Serotonbl

The 'peptidergic DH' in R. prolixus is not excreted by the tubules, but is rapidly inactivated by them (Maddrell, 1964a), whereas serotonin is excreted (Maddrell et al., 1971). Serotonin (detected by bioactivity) is present in the secreted fluid at about 10% of its concentration in the bathing fluid. Five tubules, set up in series, with each secreting into the fluid bathing the next,

INSECT DIURETIC AND ANTIDIURETIC HORMONES

375

can each be stimulated to secrete by adding a large dose of serotonin to only the first. The serotonin therefore appears to cross the walls of four tubules in series, although chemical characterisation is needed to confirm this in light of evidence that cyclic AMP is also secreted by R. prolixus MTs (Montoreano et al., 1990). The serotonergic neurohaemal sites on the abdominal nerves of R. prolLvus (one source of the serotonin released into the haemolymph during gorging) possess a high-affinity uptake mechanism for serotonin (Orchard, 1989). The high-affinity uptake system is Na+-dependent and, with the assumption that there is only one component, has an apparent KM of 0.22/~M and a Vm~,x of 585fmol 10min L. It is possible that other tissues in R. l)rolixus possess similar uptake systems, which may in part lead to the removal of serotonin from the haemolymph, thereby terminating the diuretic effect. Serotonin appears to act as a neurohormone in C. ~,icina, stimulating salivation (Hansen-Bay, 1978: Trimmer, 1985b). When C. ricimt is injected with tritiated serotonin, label is cleared from the haemolymph in two phases (Trimmer, 1985a). The first phase is rapid (<30 seconds), and may be the result of binding or cellular uptake. The second phase is more prolonged (about l h) and is mediated by organs in the abdomen. During this second phase, label appears in the faeces, suggesting that clearance may be mediated by the MTs. This was confirmed in ritro; isolated MTs, incubated in tritiated serotonin, metabolise serotonin and transport its products. Radiolabel is excreted at high concentrations (four to five times that of the bathing medium), and thin-layer chromatography reveals that the label is distributed among many products, one of which might be serotonin itself. These results show that the MTs are involved in the long-term elimination of serotonin from the haemolymph of ('. ricina. The study, however, also indicates that excretion is too slow to be the major inactivating mechanism for salivary gland stimulation, an inactivation which occurs within 5 min of injecting 20 #M serotonin. Thus, serotonin must be inactivated prior to being excreted. This was confirmed following injection of tritiated serotonin and running ethanolic extracts of the haemolymph on thin-layer chromatography. Three major peaks, possibly representing glucuronide and amino acid conjugates of a serotonin metabolite, appear in the haemolymph with a time-course sufficient to account for the termination of salivation. Thus, in C. ricina, inactivation of exogenous serotonin within the haemolymph is accomplished primarily by chemical modification to an inactive form, which provides a rapid means of terminating the stimulated salivation. The elimination of label from the haemolymph brought about by the MTs appears to be of secondary, longer-term importance. Interestingly, the eliminated products found in the faeces of injected flies are not the same as those produced by MTs in vitro, indicating that in vi~,o, MTs only excrete/process serotonin metabolites formed by, or in conjunction with, another tissue. Comparable studies have not yet been performed in R. proli.vus to see if a similar inactivating process is present.

376

7.3.3

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

CRF-related peptides

The peptidergic DH of R. prolixus is stable for up to 2 h in haemolymph, but is rapidly inactivated when incubated with MTs (Maddrell, 1964a). Likewise, Manse-DH is stable for 2 h in haemolymph from fifth-instar M. sexta larvae diluted 1 to 50 with saline, but is rapidly degraded when incubated with pieces of tubule (Li et al., 1997). The stability of the hormone in haemolymph is important, because it allows the intact peptide to reach target structures some distance from its site of release. The major degradation products formed after incubation of 4 0 # M Manse-DH with MTs were ManseDH(I-29), Manse-DH(30-41), Manse-DH(1-30) and Manse-DH(31-41), indicating scissile bonds between Leu 29 and Arg 3°, and between Arg 3° and Ala 31. These fragments are unlikely to have diuretic activity (see section 5.5.2), although this has not been confirmed. Interestingly, the bond between Leu 14 and Arg 15 is not attacked, possibly because it lies within a predicted helix whereas the Leu 29 Arg 3° bond lies in a loop region and is therefore accessible (see Li et al., 1997: K. S. Copley, W. H. Welch Jr and D. A. Schooley, unpublished observations). The degradation products produced with 40 # M Manse-DH were very complex and consequently the metabolism of Manse-DH was studied at 1 #M, at which level fewer non-specific proteolytic sites were observed. At this concentration, which is still much higher than the likely physiological concentration, 100pmol of peptide is degraded in 22rain when incubated with a 1 cm piece of tubule (Li et al., 1997). The first proteolytic cleavage between Arg 3° and Ala -~ occurs within 5rain. The amount of Manse-DH(1 30) increases over 60 rain, but then it disappears and a second product Manse-DH(1 29) appears, indicating cleavage between Leu 29 and Arg 3°. This is the major product remaining after 2 h. However, Manse-DH(1 29) is unlikely to derive from Manse-DH(l 30), because both C-terminal products, (30 41) and (31MI) are present in incubations with 4 0 # M Manse-DH. The enzyme responsible for these proteolytic cleavages appears to be a membrane-bound metalloprotease, since very little ManseDH(1 30) and no Manse-DH(l 29) are produced when EDTA is included in the incubation medium (Li et al., 1997). Additional research (W. Dahlke, E.-J. Gartreell and D. A. Schooley, unpublished observations) has confirmed the enzyme is a metallopeptidase, whose activity can be destroyed by adding EDTA, but more than restored on adding Co :+, a hallmark of a Zn :+ metallopeptidase. Its properties seem similar to Thimet oligopeptidase (Endopeptidase 24.15), one of the few well-understood neuropeptidase enzymes that will cleave at an Arg residue. This enzyme is believed to be largely soluble, although c.20% of the activity is membrane associated (O'Cuinn et al., 1995). It cleaves dynorphin 1-8, fl-neoendorphin and otneoendorphin at Leu Arg bonds (O'Cuinn et al., 1995), precisely one of the cleavage sites observed for Manse-DH. However, it also cleaves at a number of other sites with no obvious sequence specificity.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

377

In addition to inactivation by proteolytic cleavage, methionine residues (Met 2 and Met I ~) in Manse-DH are oxidised to the mono- and bis-sulfoxides during incubation with tubule pieces (Li et al., 1997). Very little oxidation occurs in control incubations, but the amount of oxidised peptide doubles between 1 and 5min after adding tubules. The oxidation is inhibited by catalase, which suggests the involvement of hydrogen peroxide, although the precise mechanism has not been determined. Malpighian tubules are high in two important enzymes of uric acid and allantoin biosynthesis, xanthine oxidase and urate oxidase, which have no reducing cofactor and consequently produce H202 as a by-product. It is quite possible therefore that these oxidative products are artefacts of in vitro metabolism. Oxidation of methionine residues in CRF-related peptides is believed to account for loss of activity during prolonged storage (G. M. Coast and I. Kay, unpublished observations), but this has not been rigorously tested. 7.3.4

Kinins

The cleavages of Manse-DH between Leu 29 and Arg 3° and between Arg 3° and Ala -~ are unusual compared with other insect neuropeptide degrading enzymes, which in general have angiotensin converting enzyme (ACE)dike and neprilysin-like activities. Purified ACE from M. domestica cleaves a C-terminal dipeptide from a wide range of neuropeptides, including Culsa-K-II and Leuma-K-I (see Isaac et al., 1998), which will be inactivated, because biological activity resides in the C-terminal pentapeptide (see section 5.5.3). ACE has been cloned from D. melanogaster and COS-7 cells transfected with the gene ( A n t e ) express ACE activity identical to that of fruit fly embryos (Cornell et al., 1995). Compared with mammalian ACE, the active site domain is highly conserved, but fruit-fly ACE does not have a hydrophobic C-terminal membrane domain. consistent with it being a soluble enzyme (see Isaac et al., 1998). It is unclear to what extent circulating ACE contributes to the degradation of neuropeptides in haemolymph, but the ACE inhibitor captopril potentiates the in vivo activity of Helze-Ks in H. virescens larvae (Seinsche el al., 2000). It should be noted that apparently the lowest concentration of peptide used in the metabolic studies of Lamango and Isaac (1993) was 50 btM. Given the lack of specificity observed in the metabolism of Manse-DH at 4 0 # M compared with 1 tzM (Li et al., 1997; see section 7.3.3), these data should be viewed with some caution. Malpighian tubules from H. _-ca larvae rapidly degrade Helze-K-lI (VRFSPWGa) by attacking bonds between Pro 5 and Trp (~, the primary cleavage site, and between Arg 2 and Phe 3 (Nachman et al., 2002a). These activities are consistent with a membrane-bound enzyme that has neprilysinlike activity. Neprilysin is a broad specificity metalloprotease that cleaves peptide bonds to the N-terminal side of hydrophobic amino acids such as phenylalanine and sometimes tryptophan. Incubation of 12.5nM Helze-K-II with pure pig kidney neprilysin results in almost complete loss of the intact

378

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

peptide over 1 h at an average rate of 132pmolh l ng-~ enzyme (Nachman et al., 2002a). This is consistent with the very rapid degradation of Locmi-K in vivo, which at physiological concentrations (10nM) has an estimated half-life of between 2 and 3 min (G. M. Coast, W. Lain and R. C. Rayne, unpublished observations).

8 8.1

Integrated activities of diuretic and antidiuretic hormones INTRODUCTION

A cocktail of biologically active signalling molecules regulate MT and hindgut function, and both may also respond autonomously to changes in haemolymph composition. In some instances, there is evidence of functional redundancy, with several factors apparently targeting the same process. For example, CRFrelated, calcitonin-like and CAP2b-related peptides are encoded in the fruit fly genome (Vanden Broeck, 2001), all of which stimulate MT secretion by activating the apical V-ATPase (see section 5.2}. Possibly they have subtly different roles and are released in response to different stimuli, but this has not been shown. Likewise, little is known of how the activities of DHs and ADHs are integrated to regulate haemolymph volume and composition. In part, this is because of the difficulty of measuring circulating hormone levels in small volumes of haemolymph. Additionally, there is currently no simple and reliable method to simultaneously measure MT secretion and urine output #l vivo, which is needed to quantify reabsorption/secretion in the hindgut. 8.2

M A I N T E N A N C E OF HAEMOLYMPH VOLUME AND COMPOSITION

A female mosquito imbibes c.3.5 #1 of blood in a single meal, of which 1.9/~1 is plasma that is hypo-osmotic to haemolymph and rich in NaCI (Williams et al., 1983). The blood is a source of protein, and >40% of the excess salt and water are excreted within 1 2 h (Williams et al., 1983). In the peak phase of diuresis, urine output can reach 8 0 n L m i n -1 (see Fig. 3) and diuretic activity is detectable in the haemolymph by bioassay (Wheelock et al., 1988). Significantly, the cyclic AMP content of the MTs is increased 10-fold (Petzel et al., 1987), which is consistent with the release of a diuretic peptide (MNP?) that activates this second messenger pathway. During the peak phase of diuresis, drops of NaCl-rich urine isosmotic to haemolymph are voided every 30s (see Fig. 3) and little or no modification occurs in the hindgut (Williams el al., 1983). Thereafter, urine output declines dramatically, but remains higher than in unfed insects, which void little or no urine. At the same time, the urine changes from being NaCl-rich to KCl-rich as excess K + from the imbibed blood cells is voided (Williams et al., 1983). Importantly, the urine is now hypo-osmotic to mammalian plasma, probably

INSECT DIURETIC AND ANTIDIURETIC HORMONES

379

due to NaCI uptake from the hindgut, which allows osmotically free water to be cleared and prevents dilution of the haemolymph. The switch from natriuresis to kaliuresis could be due in part to the disappearance of M N P from the circulation and the release of kinins, which do not have natriuretic activity (Pannabecker et al., 1993). Unlike mosquito tubules, which secrete NaCI-rich urine in the peak phase of diuresis, significant amounts of K + are secreted from serotonin-stimulated distal tubules of R. prolixus (see section 5.2.1.2), sufficient to deplete the haemolymph of K + within a minute at high urine flow rates (see Maddrell et ul., 1993)! This is prevented by K - reabsorption from the proximal tubule (Maddrell and Phillips, 1975), which is activated prior to distal tubule secretion (see Maddrell el ul., 1993). Since both the proximal and the distal tubule are maximally stimulated throughout diuresis, K + homeostasis depends upon their a u t o n o m o u s response to changes m haemolymph [K +] (Maddrell e! al., 1993). This is shown for A. domesticus tubules in Fig. 28. As can be seen, any change in bathing fluid [K +] is countered by a change in K + secretion, especially in Achdo-DP-stimulated tubules, and this is likely to be important for excreting excess K - after a meal. In the two to three hours after a blood meal, a volume of fluid equal to five to ten times the haemolymph volume is absorbed from the midgut of R. prolixus and voided via the excretory system (see Maddrell, 1980). To preserve

t00-

.=_o

80-

o E 60• E + ~° 40~ ~

20

--~

.........................

•

. . ~ .......

Of 0

',

10

,

,

20

30

40

50

[K*]bf (mM) FIG. 28 The effect of bathing fluid [K ~] on K + transport by A. domestictts tubules before (dotted line) and alter (solid line) stimulation with 50 nM Achdo-DP. Symbols represent means ±S.E. Cricket haemolymph contains 7.6 mM K ~ (shown by the vertical line), and any departure from this value is automatically countered by an increase or decrease in K ~ transport, most notably in the presence of Achdo-DP. G. M. Coast, unpublished data.

380

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

the haemolymph volume, it is necessary to balance fluid movement across the midgut and MTs, and Maddrell (1980) proposed that this be achieved with a DH that activates both processes, but with differing potency and efficacy. Assuming that the hormone has greater potency on MTs, but greater efficacy on the midgut, any imbalance is automatically rectified. For example, if MT secretion exceeds midgut absorption the haemolymph volume will fall, which will increase the concentration of D H in the circulation and further stimulate midgut transport. Serotonin might serve this purpose in that it stimulates fluid transport >six-fold across the isolated anterior midgut of fifth-instar nymphs (EC5o 50nM; Farmer el al., 1981), but is a more potent stimulant of distal tubule secretion (EC50 30~40nM; Maddrell el al., 1993). Serotonin has also been shown to stimulate ion transport by midgut and MTs of A. aegyt)ti larvae (Clark and Bradley, 1996; Clark el al., 1999), and may therefore be used to coordinate their activities when the insects are transferred to saline media (Clark and Bradley, 1997). Urine output declines dramatically from 4 0 0 n L m i n l to < 5 0 n L m i n t towards the end of the postprandial diuresis in R. prolixus (Maddrell, 1964a). Quinlan el al. (1997) suggest that release of a Manse-CAP2b-like peptide may contribute to this, since CAP2b has been shown to reduce secretion by serotonin-stimulated tubules by activating a cyclic GMP-dependent cyclic AMP phosphodiesterase (see section 5.2.5). In support of this, they show that the decline in urine output at c.3 h post-feeding correlates with a 56% increase in tubule cyclic G M P content. 8.3

SYNERGISM BETWEEN D I U R E T I C H O R M O N E S

The dose-response curve for Locmi-DH is shifted to the left and made steeper when tested in the presence of a low concentration of 0.05 nM Locmi-K (see Fig. 29; Coast, 1995), which is evidence of synergism. This is a predicted outcome of the peptides acting via different second messenger pathways to stimulate different transport pathways (see Clark el al., 1998a), and has also been demonstrated in housefly tubules with Musdo-DP and Musdo-K (Holman et al., 1999). Interestingly, there is no evidence of synergism between Achdo-DP and Achdo-K-I (Coast and Kay, 1994), which is consistent with the CRF-related peptide stimulating both cation and anion transport by A. domesticus tubules (see section 5.2.3.2). More surprising, however, Dippu-DH4{, and Dippu-DH31 act synergistically on D. punctata tubules (Furuya el al., 2000b), although they each use cyclic AMP as a second messenger and might therefore be expected to have an additive effect on tubule secretion. Similarly, the synergism between serotonin and either forskolin or a peptidergic D H from the M T G M in R. prolixus (Maddrell et al., 1993a) is unexpected, because both activate adenylate cyclase (Maddrell et al., 1993a). O'Donnell and Spring (2000) suggest that the synergism between diuretic factors that use cyclic AMP as a second messenger might result either from activation of different

INSECT DIURETICAND ANTIDIURETIC HORMONES 125-

381

Starved

Fed

e-

.o 100.= ¢n

75-

E •~

50-

/' /sis /

t"

E "~ o

25-

0

-3

//t

I

I

I

I

I

-2

-1

0

1

2

log c o n c e n t r a t i o n (nM) FIG. 29 Synergism between Locmi-DH and Locmi-K. The dose response curve for Locmi-DH (dotted line, triangles) is moved to the left and made steeper when tested in combination with 0.05 nM Locmi-K (solid line~ squares). Symbols represent means ±S.E. and the vertical lines show the haemolymph titre of Locmi-DH in fed and starved insects. Redrawn from Coast (1995).

isoforms of adenylate cyclase or through cross talk between second messenger pathways, and this clearly warrants further investigation. The peptidergic DH of R. pro/i_rus that synergises with serotonin has not been identified, but it is unlikely to be either a kinin or a CRF-related peptide although both are present in the M T G M (see sections 6.1.4 and 6.1.5). Kinins have no effect on secretion (Te Brugge eta/., 2002) and, in separate studies, neither Locmi-DH (Coast, 1996) nor Z o o n e - D H (Te Brugge eta/., 2002) has been shown to act synergistically with serotonin. There must therefore be some other peptidergic D H in the M T G M . Synergism produces an increase in potency (see Fig. 29), which means that less D H needs to be released. This permits diuresis to be switched off rapidly, because there is less D H to remove from the circulation. Using data from Coast (1995), it would require the release of about 90% of the Locmi-DH stored in the brain and CC (c.4nmol: Audsley et al., 1997b) to increase MT secretion from 10% to 90% of its maximum rate. The same effect can be achieved with the release of just 5% of stored peptide in the presence of 0.05nM Locmi-K. Synergism also makes the dose response curve steeper (see Fig. 29) allowing precise control of diuresis, because small changes in D H concentration have a large impact on tubule secretion. The independent and controllable release of the two hormones would further increase the precision with which tubule

382

G.M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

secretion is regulated, and it is noteworthy that although kinins and CRFrelated peptides co-localise in abdominal ganglia NSCs, this has not been reported for MNCs of the brain (see section 6.1.9). 8.4

CO-ORDINATING MALP1GHIAN TUBULE AND HINDGUT ACTIVITIES

Fluid recycling between MTs, hindgut and haemolymph increases several fold in recently fed locusts (Phillips and Audsley, 1995), allowing toxic waste to be eliminated rapidly from the circulation while minimising excretory water loss. Malpighian tubule secretion and hindgut reabsorption must therefore be coordinated, which might be achieved with one hormone controlling both processes, as proposed for the integration of midgut and MT function (see section 8.2). Coast et al. (1999) conducted reciprocal in vitro bioassays with pure peptides to show that Schgr-ITP does not influence locust MT secretion, and stimulants of tubule secretion (Locmi-K and Locmi-DH) do not stimulate either rectal or ileal ls~. or J,.. Thus, hormonal control of these two major segments of the locust excretory system is clearly separated. This may be needed for the precise regulation of haemolymph volume and composition, but requires the release of DHs and ADHs to be co-ordinated. Unfortunately, nothing is known about the circulating levels of ADHs, although the amount of CTSH-Iike activity in haemolymph increases after feeding (Spring and Phillips, 1980c). As mentioned previously, excretory water loss is the difference between the rates of fluid entry into the hindgut and reabsorption in the ileum and rectum. The capacity of the locust ileum and rectum to reabsorb fluid could make it difficult for the insect to deal with an hydration stress without reducing fluid uptake in the hindgut, and one way that this might be achieved is with the release of an A D H antagonist, such as Schgr-ITP-L (Phillips et al., 1998b). The natural function of ITP-L is unknown, but the identical N-terminal sequences of ITP and ITP-L to residue 40 suggest that it might bind to the ITP-R and act as an antagonist of ITP so as to curtail hindgut fluid reabsorption, or act to reduce synthesis and release of ITP at the brain-CC level. Phillips et al. (1998b) suggest that excess reabsorption of dilute fluid in the hindgut due to overstimulation by ITP might cause general tissue swelling. This might directly trigger widespread release of ITP-L fi'om various tissues to reduce ITPstimulated fluid recovery in the hindgut leading to diuresis. In this sense, ITP-L might be analogous to vertebrate atrial natriuretic peptide (ANP), which is found in several body tissues and which inhibits renal salt and hence fluid reabsorption. This working hypothesis might now be examined using expressed ITP-L and antibodies available to this peptide. Sf9 cells transfected with baculovirus do not correctly cleave expressed ITP and ITP-L, leading to an 11 amino acid extension of the N-terminus, which could cause the antagonistic action of ITP-L on the ileal I~c bioassay. This possibility was excluded by Pfeifer el al. (1999) and Wang et al. (2000), who

INSECT DIURETIC AND ANTIDIURETIC HORMONES

383

used Drosophila Kcl cells and a designed plasmid vector to express ITP (KcITP) and ITP-L (KclTP-L) with correct N-terminal cleavage. Wang el al. (2000) found that 5 nM KclTP-L caused 75% inhibition of ileal I~c stimulation by 0.08 nM KclTP. Thus the physiological function of ITP-L proposed above remains a viable one for future study.

9

The excretory system as a target for pest control strategies

Hormones control many aspects of the physiology, and the behaviour of insects and the endocrine system is seen as an important target in the development of novel, safe and specific insecticides (Masler e t a / . , 1993: Kelly et al., 1994). Of the numerous endocrine targets available, the control of water balance is particularly attractive because the adults of all major pest species are terrestrial and must carefully regulate excretory water loss. For example, the use of DH agonists and/or A D H antagonists could lead potentially to dehydration and death. The excretory system is also an important route for the clearance of toxic waste, and if this is compromised, t"o1"example with a DH antagonist, it is likely to adversely affect survival. Maeda (1989) inserted a synthetic gene encoding Manse-DH into the genome of a baculovirus. The gene contained an N-terminal signal sequence based upon a D. melanogaster cuticle protein and a sequence that encoded Manse-DH extended by a glycine residue for C-terminal amidation. Northern blot analysis of m R N A from fat body of silkworm (Boml)y.v mori) larvae infected with the recombinant virus (BmDH5) confirmed the synthetic gene was expressed. Moreover, haemolymph collected from these larvae contained several hundred times more diuretic activity, as determined by the stimulation of post-eclosion diuresis in decapitated newly emerged P. rapae butterflies, than haemolymph from an equivalent number of M. se.vta pupae (Maeda, 1989). Critically, however, diuretic activity was not measured in haemolymph from B. moll larvae infected with wild-type virus, which might have confirmed the expression and secretion of correctly processed Manse-DH. A detailed analysis of large (c.50 mE) volumes of haemolymph from B. mori infected with BmDH5 showed the presence of a number of peaks on RPLC with diuretic activity, but none had a retention time coincident with ManseDH (C. A. Miller, R. G. Troeschler, S. J. Kramer, and D. A. Schooley, unpublished data). It seems likely that the biologically active zones observed in haemolymph from infected animals could well represent endogenous B. mori kinins, or other factors. Nevertheless, larvae infected with BmDH5 suffered a 30% reduction in haemolymph volume compared with controls and with larvae infected with wild-type virus (Maeda, 1989). Interestingly, water lost from the haemolymph of BmDH5 injected larwm is retained in the midgut, which is consistent with Manse-DH acting on both the cryptonephric and free portions of the MTs to recycle fluid from the rectal complex to the midgut (see

384

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

section 2.3). Significantly, all BmDH5 infected larvae were dead within four days, whereas >95% of larvae infected with the wild-type virus survived (Maeda, 1989). The cause of death was unclear, but it could have resulted from a build up in the concentration of toxic waste in the haemolymph as its volume declined (see Maeda, 1989). Ma et al. (2000) investigated the effect of administering Manse-DPI1 to M. sexta neonates by applying it to tobacco leaf discs at doses of between 1.5 ng and 15 000 ng. The larvae gained less weight, developed more slowly and suffered higher mortality than controls, although it is unclear what caused the latter. Surprisingly, Manse-DPII had no effect when injected direct into the haemocoel, which suggests it may have an antifeedant effect. Thus, Keeley et al. (1992) showed that Manse-DH has antifeedant activity in H. virescens larvae, and Locmi-DH has a similar effect in L. migratoria (Coast and Goldsworthy, 1997), but in both of these studies the peptides were injected into the haemocoel. At present, it is unclear whether Manse-DPIl applied to leaf discs is assimilated in a biologically active form, although this would be surprising given that it is likely to be broken down by proteolytic enzymes in the gut. Seinsche et al. (2000) has shown that injections of 50 pmol Helze-K-I into fifth-instar H. virescens larvae along with 1 tzmol captopril (an ACE inhibitor) results in 83% larval mortality within 5 days, compared with 7% in salineinjected controls and 44% in animals injected with the kinin alone. The scissile bonds of insect kinins (see section 7.3.4) can be protected with a sterically hindered anainoisobutyryl (Aib) residue. The Aib residue is compatible with a type VI fl-turn and FF-Aih-WG-NH2 retains the potency and activity of the parent compound (FFSWG-NH2) in a cricket tubule bioassay, but is resistant to ACE (Nachman et al., 1997). The double Aib analogue Aib-FS-Aib-WGNHz is completely resistant to degradation by H. zea tubules, and is more potent (ECs0 1.2pM) than native Achdo-Ks (ECsos of between 22 and 324pM) in the cricket assay (Nachman et al., 2002a). This analogue is also active in a housefly tubule assay (Nachman et al., 2002a), but is less potent (ECs0 1.5#M) than Musdo-K (ECs0 0.13 nM), consistent with the role of the N-terminal region for high affinity receptor binding (Coast et al., 2002; see section 5.5.3). Nevertheless, injections of 50pmol Aib-FS-Aib-WG-NH2 and Musdo-K are equally effective in stimulating radiolabelled inulin excretion from intact houseflies at 1 and 2h post-injection, most likely because the peptidase-resistant analogue survives longer in the circulation. While these results demonstrate the feasibility of developing stable kinin analogues, a key to their use as pest control agents is that they should be topically active, i.e. able to penetrate the waxy epicuticle. Teal and Nachman (1997) have developed amphiphilic analogues of PBAN that have pheromonotropic activity when applied topically in water to H. virescens cuticle. These analogues have a hydrophobic moiety replacing the Phe residue in the PBAN active core (FTPRL-NH2) to counter the polar side-chain of Arg. The hydrophobic

INSECT DIURETIC AND ANTIDIURETIC HORMONES

385

moieties tested included 6-phenylhexanoic acid, 9-fluorenacetic acid and 1-pyrenebutyric acid. The same moieties have been attached to the N-terminus of a kinin analogue, A R F F P W G - N H 2 , the Arg residue being included to produce an amphiphilic molecule. All three analogues retained full diuretic activity in a cricket tubule assay (ECs0s 0.25 nM to 1.7nM; R. J. Nachman and G. M. Coast, unpublished observations), but the lack of a reliable in viro assay has prevented them from being tested after topical application. In a search for novel agonists/antagonists at DH receptors, a tripeptide combinatorial library constructed from D-amino acids has been screened in a cricket tubule assay (G. M. Coast and R. J. Nachman, unpublished observations). Several agonists with the general sequence Fmoc-dPro-dPro-dXxx-NH2 were identified, all of which form a nascent right-handed polyproline lI helix that can be superimposed on the kinin type VI fi-turn conformation (Moyna el al., 1999). Since these analogues are constructed from D-amino acids, they are resistant to peptidase attack and could therefore be important leads for future studies. To summarise, CRF-related peptides and kinins have been shown to disrupt insect growth and development and increase mortality. This may not result from their diuretic activity, because increased secretion of primary urine is probably countered (automatically?) by fluid reabsorption in the hindgut. It is unclear whether effects on feeding behaviour are mediated peripherally or centrally, although the latter would require that they cross the blood brain barrier. Increased visceral muscle activity in response to CRF-related peptides (Blake et al., 1996) and kinins (Holman et al., 1990) might disrupt the passage of food through the gut and reduce feeding, while the antifeedant activity of Locmi-DH is attributed to an effect on gustatory sensilla (Coast and Goldsworthy, 1997). The identification of ITP receptor antagonists (see section 5.5.5) also offers a promising area for future research by inserting their cDNAs into insect viruses, and trial studies have already been conducted (D. Theihnann, Agriculture Canada; personal communication) using locust ITP cDNAs provided by H. W. Brock, J. Meredith and J. E. Phillips (University of British Columbia).

10

Future directions

While MT and hindgut epithelial transport processes are reasonably well understood in some insects, it is very evident from this review that the field of endocrine control of MT and of hindgut function in particular is still at an early stage. It is equally clear that a comprehensive understanding of the excretory system's role in whole body osmotic and ionic homeostasis will depend firstly on establishing which factors normally control MT secretion and hindgut reabsorption and secretion in ~,ivo and secondly on elucidating

386

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

how their release is controlled and co-ordinated. This is a challenge for the next generation. To date, five different neuropeptide families have been shown to have diuretic or antidiuretic activity on isolated MTs, but the peptidergic hormone of R. prolixus that acts synergistically with serotonin has not been identified, and MNP and FopADF have still to be sequenced. Similarly, a number of distinct neuropeptide stimulants of locust hindgut (e.g. CTSH, Ventral Ganglia Factor (VGF), a second stimulant of ileal l~c in HPLC factions from CC) and of the M. se.vta cryptonephric complex (Manse-ADF A and B) still have not been sequenced because of inadequate biochemical separation methods for small amounts of acid-labile peptides. For such peptides, expression-cloning and identification of secreted stimulants in functional bioassays appears to be the most promising approach to obtain deduced amino acid sequences from their cDNAs. Given the size of some of these neuropeptides, their production for physiological and structure activity studies, and for generating specific antibodies, is more easily done using a good expression system (e.g. in the Kcl-plasmid vector system and in protease-deficient S. cerevMae) when a cDNA has been obtained. Development of new bioassay systems for insects other than locusts and the tobacco hornworm will be essential, and this requires that the search for stimulant cDNAs should be directed to large insects within any new taxonomic group. An exception is D. melanogaster, where the completed genome, extensive stocks of genetic strains, and genetic manipulations offer unique potential (see Dow and Davies, 2001). The identification of genes encoding DHs and ADHs and their receptors permits functional studies (functional genomics) by observing the resultant phenotype after the gene of interest is mutated (reverse genetics) or "knocked out' by double-stranded RNA (dsRNA)-mediated interference. For example, Macins et al. (1999) found that Schgr-lTP-like activity appeared abruptly in the head, thorax and abdomen of embryos about 80% of the way through development in locust eggs. In order to study the functions of this neuropeptide during development, the Drome-ITP gene is being inactiwtted using transposons (H. W. Brock, J. Meredith and J. E. Phillips, unpublished observations). Such treatment is lethal. Conceivably Drome-ITP controlled hindgut fluid reabsorption causes the body volume expansion necessary for emergence of the first instar. Chung et al. (1999) have provided the first evidence that an ITP holnologue, CHH, has such a role during ecdysis in crustaceans. An alternative genome-wide approach is the use of DNA micro- or oligonucleotide arrays to monitor the expression profiles of all identified genes. Quantitative comparisons of expression profiles for genes encoding neuropeptides, their receptors, signalling pathways and ion transport/channel proteins in flies reared under different environmental conditions will provide a unique insight into the multidimensional control of excretory processes. The complementary technique of peptidomic analysis uses a combination of RPLC

INSECT DIURETIC AND ANTIDIURETIC HORMONES

387

and MS to identify peptides present in biological samples from their unique M,-s (Clynen et al., 2002). By comparing the haemolymph complement of neuropeptides in insects held under hydrating and dehydrating conditions it should be possible to identify, those components that change and might therefore have a physiological role in the control of salt and water balance. The Drosophila genome has been scanned for peptide GPCRs, and the majority, possibly all, have been identified (Hewes and Taghert, 2001). There appear to be about 20 orphan receptors. It should therefore be relatively easy to identify the receptors for Drome-DH31, Drome-CAP2b and Drome-lTP from binding of the radiolabelled ligands to these orphan receptors expressed in a suitable cell system. In turn, the deduced amino acid sequences for the receptors would permit production of antibodies against unique regions that could be used to identify potential target organs in larval and adult D. melanogasler, or even to potentially block receptors ill vivo in physiological studies on a larger species of fly. While this review has concentrated on the structural identification of DHs and ADHs, their expression in NSCs, mode of action on MTs and hindgut, and physiological relevance for the control of excretion, it should not be forgotten that these NSCs are neurons and therefore can be expected to participate more generally in the complex cascade of events that results in an animal entering a new behavioural state. These behavioural states would be ones associated with events that might compromise salt and water balance such as feeding, high metabolic activities (flight), ecdysis, etc. Thus, the NSCs expressing DHs possess central projections, providing the necessary neural link to the nervous/endocrine systems. In addition, DHs are expressed in a variety of cell types including interneurons and possibly sensory neurons, and so it is feasible, and should be anticipated, that these neuropeptide families (and indeed serotonin) might represent functional units that interact to modulate behaviour. These functional units and the neuroactive chemicals that they release might bias, at many levels and sites, neuronal, hormonal and physiological events towards a new functional state of the animal. Behaviours such as feeding cannot occur in isolation, but must be part of a behavioural sequence with distinct phases. The families of DHs/ADHs are ideally suited to co-ordinate the quite disparate physiological events that are associated with a common behaviour. That said, it is also clear from this review that essentially nothing is known about the neurobiology of the DH/ADH-containing NSC (or other cell types); no concept of the uniquely identifiable D H / A D H cell', no information on the integration of these cells with the nervous/endocrine systems: and no information as to the neuroendocrine circuits leading to their activation, their feedback loops, regulation of synthesis and release, or integration with the peripheral (sensory) system. This is a vacuum that must be filled if we are to understand the true physiological relevance of the active factors and their participation in the varied behavioural states of the insect that ultimately leads to the success of insects. Research into insect DHs and

388

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

A D H s will require the full range o f m u l t i d i s c i p l i n a r y a p p r o a c h e s if we are to unravel their complexities in c o m m u n i c a t i o n a n d have a true u n d e r s t a n d i n g from genomics, t h r o u g h p h y s i o l o g y , to behaviour.

Acknowledgements The a u t h o r s t h a n k the researchers who have c o n t r i b u t e d to the w o r k presented in this review a n d have allowed us to r e p o r t u n p u b l i s h e d findings. W e are grateful to the N a t i o n a l Institutes o f Health and the N a t u r a l Sciences a n d Engineering R e s e a r c h Council o f C a n a d a for their s u p p o r t .

References Adams, M. D., et al. (2000). The genome sequence of Drosophila meNnogaster. Science 287, 2185 2195. Agricola, H. J. and Brfiunig, P. (1995). In "The Nervous Systems of Invertebrates: an Evolutionary and Comparative Approach." (O. Breidbach and W. Kutsch, eds), Comparative aspects of peptidergic signaling pathways in the nervous systems of arthropods, pp. 303 327, Birkhauser Verlag, Basel. Ali, D. W. (1997). The aminergic and peptidergic innervation of insect salivary glands. J. Exp. Biol. 200, 1941-1949. Audsley, N. (1991). Purification of a neuropeptide from the corpus cardiacum of the desert locust which influences ileal transport. Ph.D., University of British Columbia, Vancouver. Audsley, N. and Phillips, J. E. (1990). Stimulants of ileal salt transport in neuroendocrine system of the desert locust. Gen. Comp. Endocrhlol. 80, 127 137. Audsley, N., Mclntosh, C. and Phillips, J. E. (1992a). Actions of ion-transport peptide from locust corpus cardiacum on several hindgut transport processes. Y. Evp. Biol. 173, 275 288. Audsley, N., Mclntosh, C. and Phillips, J. E. (1992b). Isolation of a neuropeptide from locust corpus cardiacum which influences ileal transport. J. Exp. Biol. 173, 261 274. Audsley, N., Coast, G. M. and Schooley, D. A. (1993). The effects of Manduca sexta diuretic hormone on fluid transport by the Malpighian tubules and cryptonephric complex of Mamtuca sexta. J. Exp. Biol. 178, 231 243. Audsley, N., Mclntosh, C., Phillips, J. E., Schooley, D. A. and Coast, G. M. (1994). In 'Perspectives in Comparative Endocrinology' (K. G. Davey, R. E. Peter and S. S. Tobe, eds), Neuropeptide regulation of ion and fluid reabsorption in the insect excretory system, pp. 74 80, National Research Council of Canada, Ottawa. Audsley, N., Kay, l., Hayes, T. K. and Coast, G. M. (1995). Cross reactivity studies of CRF-related peptides on insect Malpighian tubules. Comp. Biochem. Physiol. A l l & 87-93. Audsley, N., Goldsworthy, G. J. and Coast, G. M. (1997a). Circulating levels of Locusta diuretic hormone: the effect of feeding. Peptides 18, 59 65. Audsley, N., Goldsworthy, G. J. and Coast, G. M. (1997b). Quantilication of Locusta diuretic hormone in the central nervous system and corpora cardiaca: influence of age and feeding status, and mechanism of release. Regul. Pept. 69, 25- 32.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

389

Baernholdt, D. and Andersen, S. O. (1998). Sequence studies on post-ecdysial cuticular proteins fi'om pupae of the yellow mealworm, Tenebrio molitor. Insect Biochem. Mol. Biol. 28, 517 526. Baines, R. A., Thompson, K. S. J., Rayne, R. C. and Bacon, J. P. (1995). Analysis of the peptide content of the locust vasopressin-like immunoreactive (VPLI) neurons. Peptides 16, 799 807. Baldwin. D. C., Schegg, K. M., Furuya, K., Lehmberg, E. and Schooley, D. A. (2001). Isolation and identification of a diuretic hormone from Zootermopsis nevadensis. Peptides 22, 147 152. Barrett, F. M. and Orchard, 1. (1990). Serotonin-induced elevation of cAMP levels in the epidermis of the blood-sucking bug, Rhodnius prolixus. J. Insect Physiol. 36, 625633. Barrett, F. M., Orchard, 1. and Te Brugge, V. (1993). Characteristics of serotonininduced cyclic AMP elevation in the integument and anterior midgut of the bloodfeeding bug, Rhodnius prolixus. J. Insect Physiol. 39, 581 587. Bernays. E. A. and Chapman, R. F. (1972). The control of changes in peripheral sensilla associated with feeding in Locusta migratoria (L). J. Ext~. Biol. 57, 755 763. Bernays, E. A. and Chapman, R. F. (1974). Changes in haemolymph osmotic pressure in Locusm mi~ratoria larvae in relation to feeding. J. Entom. A48, 149 155. Bernays, E. A. and Simpson, S. J. (1982). Control of food intake. Adv. Insect Physiol. 16, 59 118. Bertram, G., Schleithoff, L.. Zimmermann, P. and Wessing, A. (1991). Bafilomycin A I is a potent inhibitor of urine formation by Malpighian tubules of Drosophih~ hv&'i: is a vacuolar ATPase involved in ion and fluid secretion? J. Insect Physiol. 37, 201 2(19. Bertsch, A. (1984). Foraging in male bumblebees (Bomhus lucorum L.): maximizing energy or minimizing water load. Oecolo~ia 62, 325 336. Beyenbach, K. W. (1993). In "Structure and Function of Primary Messengers in Invertebrates: Insect Diuretic and Antidiuretic Peptides" (K. W. Beyenbach, ed.), Extracellular fluid homeostasis in insects, pp. 146 173. Karger, Basel. Beyenbach, K. W. (1995). Mechanism and regulation of electrolyte transport in Malpighian tubules. J. Insect Physiol. 41, 197 2(17. Beyenbach, K. W. and Masia, R. (2002). Membrane conductances of principal cells in Malpighian tubules of Aedes aegypti..I. Insect Physiol. 48, 375 386. Bilgen, T. (1994). Investigation of an ion transport peptide in desert locust ventral ganglia. M.Sc. University of British Columbia, Vancouver. Black, K., Meredith, J., Thomson, B., Phillips, J. and Dietz, T. (1987). Mechanisms and properties of sodium transport in locust rectum. Can..I. Zoo/. Rer. Can. Zool. 65, 3084 3092. Blackburn, M. B. and Ma, M. C. (1994). Diuretic activity of Mas-DP II, an identified neuropeptide from Mamluca se.vta: an in vivo and in vitro examination in the adult moth. Arch. Insect Biochem. Physiol. 27, 3 10. Blackburn, M. B., Kingan, T. G., Bodnar, W. T. K., Wagner, R. M., Raina, A. K., Schnee, M. E. and Ma, M. C. (1991). Isolation and identification of a new diuretic peptide from the hornwornl, Manduca se.vta. Biochem. Biophys. Res. Commmt. 181, 927 932. Blackburn, M. B., Wagner, R. M., Shabanowitz, J., Kochansky, J. P., Hunt, D. F. and Raina, A. K. (1995). The isolation and identification of three diuretic kinins from tbe abdominal nerve cord of adult ttelicoverpa zea. J. Insect Physiol. 41. 723 730. Blake. P. D., Kay', I. and Coast, G. M. (1996). Myotropic activity of Acheta diuretic peptide on the foregut of the house cricket, Achela domesticus(L). J. Insect Physiol. 42. 1053 1059.

390

G . M . COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

Br/iunig, P. (1987). The satellite nervous system: an extensive neurohemal network in the locust head. J. Comp. Physiol. AI60, 69 77. Brown, B. E. (1967). Neuromuscular transmitter substance in insect visceral muscle. Science, 595-597. Cady, C. and Hagedorn, H. H. (1999a). The effect of putative diuretic factors on in vivo urine production in the mosquito, Aedes aegypti. J. Insect Physiol. 45, 317 325. Cady, C. and Hagedorn, H. H. (1999b). Effects of putative diuretic factors on intracellular second messenger levels in the Malpighian tubules of Aedes aegypti. J. Insect Physiol. 45, 327 337. Cantera, R. and Nfissel, D. R. (1992). Segmental peptidergic innervation of abdominal segments in larval and adult dipteran insects revealed with an antiserum against leucokinin I. (2"11 Tissue Res. 269, 459 471. Cantera, R., Hansson, B. S., Hallberg, E. and N/issel, D. R. (1992). Postembryonic development of leucokinnin I-immunoreactive neurons innervating a neurohaemal organ in the turnip moth, Agrotis segetum. Cell Tissue Res. 269, 65-77. Chamberlin, M. E. and Phillips, J. E. (1983). Oxidative metabolism in the locust rectum. J. Comp. Biol. 151, 191 198. Chamberlin, M. E. and Phillips, J. E. (1988). Effects of stimulants of electrogenic ion transport on cyclic AMP and cyclic G M P levels in locust rectum. J. Exp. Zool. 245, 9 16. Chen, Y., Veenstra, J. A., Davis, N. T. and Hagedorn, H. H. (1994a). A comparative study of leucokinin-immunoreactive neurons in insects. Cell Tissue Res. 276, 69 83. Chen, Y., Veenstra, J. A., Hagedorn, H. and Davis, N. T. (1994b). Leucokinin and diuretic hormone immunoreactivity of neurons in the tobacco hornworm, Mamhlca sexta, and co-localization of this immunoreactivity in lateral neurosecretory cells of abdominal ganglia. Cell Tissue Res. 278, 493-507. Cheung, C. C., Loi, P. K., Sylwester, A. W., Lee, T. D. and Tublitz, N. J. (1992). Primary structure of a cardioactive neuropeptide from the tobacco hawkmoth, Manduea sexta. FEBS Lett. 313, 165-168. Chung, J. S., Goldsworthy, G. J. and Coast, G. M. (1994). Haemolymph and tissue titres of achetakinins in the house cricket Aeheta domesticus: effect of starvation and dehydration. J. Exp. Biol. 193, 307 319. Chung, J. S., Wheeler, C. H., Goldsworthy, G. J. and Coast, G. M. (1995). Properties of achetakinin binding sites on Malpighian tubule membranes from the house cricket, Aeheta domestieus. Peptides 16, 375 382. Chung, J. S., Dircksen, H. and Webster, S. G. (1999). A remarkable, precisely timed release of hyperglycemic hormone from endocrine cells in the gut is associated with ecdysis in the crab Carcinus maenas. Proe. Natl Aead. Sci. USA 96, 13 103-13 107. Clark, T. M. and Bradley, T, J. (1996). Stimulation of Malpighian tubules from larval Aedes ae,~ypti by secretagogues. J. lnseet Physiol. 42, 593 602. Clark, T. M. and Bradley, T. J. (1997). Malpighian tubules of larval Aedes aegvpti are hormonally stimulated by 5-hydroxytryptamine in response to increased salinity. Arch. h~sect Bioehem. Physiol. 34, 123-141. Clark, T. M. and Bradley, T. J. (1998). Additive effects of 5-HT and diuretic peptide on Aedes Malpighian tubule fluid secretion. Comp. Biochem. Physiol. A l l 9 , 599 605. Clark, T. M., Hayes, T. K. and Beyenbach, K. W. (1998a). Dose-dependent effects of CRF-like diuretic peptide on transcellular and paracellular transport pathways. Am. .l. Physiol. 274, F834-F840. Clark, T. M., Hayes, T. K., Holman, G. M. and Beyenbach, K. W. (1998b). The concentration-dependence of CRF-like diuretic peptide: mechanisms of action..1. Exp. Biol. 201, 1753 1762.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

391

Clark, T. M., Koch, A. and Moffett, D. F. (1999). The anterior and posterior "stomach" regions of larval Aedes aegypti midgut: regional specialization of ion transport and stimulation by 5-hydroxytryptamine, J. Exp. Biol. 202, 247-252. Clottens, F. L., Holman, G., Coast, G. M.. Totty, N. F., Hayes, T. K., Kay, I., Mallet, A. 1., Wright, M. S., Chung, J.-S. and Truong, O. (1994). Isolation and characterization of a diuretic peptide common to the house fly and stable fly. Peptides 15, 971 979. Clynen, E.. Baggerman, G., Veelaert, D., Cerstiaens, A., Van der Horst, D., Harthoorn, L., Derua, R., Waelkens. E.. De Loof, A. and Schoofs, L. (2001). Peptidomics of the pars intercerebralis corpus cardiacum complex of the migratory locust, Locusta migratoria. Eur. J. Biochem. 268, 1929 1939. Clynen, E.. Stubbe, D,, De Loof, A. and Schoofis, L. (2002). Peptide differential display: a novel approach for phase transition in locusts. Comp. Biochem. Physiol. B!32, 107 115. Coast, G. M. (1995). Synergism between diuretic peptides controlling ion and fluid transport in insect Malpighian tubules. Regul. P~7~t. 57, 283 296. Coast, G. M. (1996). Neuropeptides implicated in the control of diuresis in insects. P~7~tides 17, 327 336. Coast. G. M. (1998a). The influence of neuropeptides on Malpighian tubule writhing and its significance for excretion. Peptides 19, 469 480. Coast. G. M. (1998b). In 'Recent Advances in Arthropod Endocrinology" (G. M. Coast and S. G. Webster. eds), The regulation of primary urine production in insects. pp. 189 209. Cambridge University Press, Cambridge. Coast, G. M. (2001a). Diuresis in the housefly (Musca ~hmws'tica) and its control by ncuropeptides. Peptides 22, 153 160. Coast, G. M. (2001b). The neuroendocrine regulation of salt and water balance in insects. Zoology 103, 179 188, Coast, G. M. and Goldsworthy, G. J. (1997). In "Advances in Comparative Endocrinology" (S. Kawashima and S. Kikuyama, eds), New aspects of insect diuretic hormone function, pp. 107-113, Monduzzi Editore, Bologna. Coast, G. M. and Kay, I. (1994). The effects of Acheta diuretic peptide on isolated Malpighian tubules from the house cricket Acheta domesticus. J. E.xT). Biol. 187, 225 243. Coast. G. M., Holman, G. M. and Nachman, R. J. (1990a). The diuretic activity of a series o[" cephalomyotropic neuropeptides, the achetakinins, on isolated Malpighian tubules of the house cricket, Acheta domesticus. J. bisect Physiol. 36, 481M-88. Coast, G. M., Wheeler, C. H., Totty, N. F., Philp, R. J. and Goldsworthy, G. J. (1990b). In "Insect Neurochemistry and Neurophysiology 1989" (A. B. Borkovec and E. P. Masler, eds), Isolation, purification, and characterization of a diuretic peptide (AP:I) fiom the house cricket, Acheta dmm'sticus, pp. 227 230, Humana Press, Clifton, New Jersey. Coast, G. M., Hayes, T. K., Kay, 1. and Chung, J.-S. (1992). Effect of Mamh~ca scxta diuretic hormone and related peptides on isolated Malpighian tubules of the house cricket Acheta domesticus (L.). J. K\p. Biol. 162, 331 338. Coast, G. M., Rayne, R. C., Hayes, T. K., Mallet, A. 1.. Thompson, K. S. J. and Bacon, J. P. (1993). A comparison of the effects of two putative diuretic hormones from Locusta migratoria on isolated locust Malpighian tubules..I. Kvp. Biol. 175, 1 14. Coast, G. M., Chung, J.-S., Goldsworthy, G. J., Patel, M.. Hayes, T. K. and Kay, I. (1994). In "Perspectives in Comparative Endocrinology" (K. G. Davey, R. E. Peter and S. S. Tobe, eds), Corticotropin releasing factor related diuretic pcptides in insects, pp. 67 73, National Research Council of Canada, Ottawa.

392

G . M . COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

Coast, G. M., Meredith, J. and Phillips, J. E. (1999). Target organ specificity of major neuropeptide stimulants in locust excretory systems. J. Exp. Biol. 202, 3195 3203. Coast, G. M., Webster, S. G., Schegg, K. M., Tobe, S. S. and Schooley, D. A. (2001). The Drosophila melanogaster homologue of an insect calcitonin-like diuretic peptide stimulates V-ATPase activity in fruit fly Malpighian tubules. J. Exp. Biol. 204, 1795 1804. Coast, G. M., Zabrocki, J. and Nachman, R. J. (2002). Diuretic and myotropic activities of N-terminal truncated analogs of Musca domestica kinin neuropeptide. Peptides 23, 701 708. Colas, J.-F., Choi, D. S., Launay, J.-M. and Maroteaux, I_. (1997). Evolutionary conservation of the 5-HT2B receptors. Ann. N.Y. Acad. Sci. 812, 149 153. Cook, H. and Orchard, I. (1990). Effects of 5,7-DHT upon feeding behaviour and serotonin content of various tissues in Rhodnius prolixus. J. Insect Physiol. 36, 361-367. Cook, H. and Orchard, I. (1993a). Mode of action of 5,7-DHT in insects, hzsect Biochem. Mol. Biol. 23, 895 904. Cook, H. and Orchard, I. (1993b). The short term effects of 5,7-dihydroxytryptamine on peripheral serotonin stores in Rhodnius prolirus and their long term recovery. Insect Biochem. Mol. Biol. 23, 895-904. Copley, K. S., Alm, S. M., Schooley, D. A. and Courchesne, W. E. (1998). Expression, processing and secretion of a proteolytically-sensitive insect diuretic hormone by Saccharomyces cerevisiae requires the use of a yeast strain lacking genes encoding the Yap3 and Mkc7 endoproteases found in the secretory pathway. Biochem. J. 330, 1333-1340.

Cornell, M. J., Williams, T. A., Lamango, N. S., Coates, D., Corvol, P., Soubrier, F., Hoheisel, J., Lehrach, H. and Isaac, R.E. (1995). Cloning and expression of an evolutionary conserved single-domain angiotensin-converting enzyme from Drosophila melanogaster. J. Biol. Chem. 270, 13 613.-13 619. Cornette, J. L., Cease, K. B., Margalit, H., Spouge, J. L. Berzofsky, J. A. and DeLisi, C. (1987). Hydrophobicity scales and computational techniques for detecting amphipathic structures in proteins. J. Mol. Biol. 195, 659 685, Coutchi6, P. A. and Machin, J. (1984). Allometry of water-vapor absorption in 2 species of tenebrionid beetle larvae. Am. J. Physiol. 247, R230-R236. Cox, K. J. A., Tensen, C. P., Van der Schors, R. C., Li, K. W., Van Heerikhuizen, H., Vreugdenhil, E., Geraerts, W. P. M. and Burke, J. F. (1997). Cloning, characterization, and expression of a G-protein-coupled receptor from Lymnaea stagnalis and identification of a leucokinin-like peptide, PSFHSWSamide, as its endogenous ligand. J. Neurosci. 17, 1197 1205. Dautzenberg, F. M. and Hauger, R. L. (2002). The CRF peptide family and their receptors: yet more partners discovered. Trends Pharmacol. Sci. 23, 71 77. Dautzenberg, F. M., Kilpatrick, G. J., Hauger, R. L. and Moreau, J.-L. (2001). Molecular biology of the CRH receptors in the mood. Peptides 22, 753 760. Davies, S. A., Huesmann, G. R., Maddrell, S. H. P., O'Donnell, M. J., Skaer, N. J. V., Dow, J. A. T. and Tublitz, N. J. (1995). CAP2b, a cardioacceleratory peptide, is present in Drosophila and stimulates tubule fluid secretion via cGMP. Ant. J. Physiol. 269, R1321 R1326. Davies, S. A., Stewart, E. J., Huesmann, G. R., Skaer, N. J. V., Maddrell, S. H. P., Tublitz, N. J. and Dow, J. A. T. (1997). Neuropeptide stimulation of the nitric oxide signaling pathway in Drosophila mehmogaster Malpighian tubules. Ant. J. Physiol. 42, R823 R827. Davis, N. T. (1985). Serotonin-immunoreactive visceral nerves and neurohemal system in the cockroach Periplaneta americana (L.). Cell Tissue Res. 240, 593 600.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

393

Davis, N. T. (1987). Neurosecretory neurons and their projections to the serotonin nenrohemal system of the cockroach Periphmeta americana (L.), and identification of mandibular and maxillary motor neurons associated with this system. J. Comp. Neurol. 259, 604-621. De Decker, N. (1993). Regulation of fluid secretion in Malpighian tubules of Formica polyctena by exo- and endogenous li~ctors. Ph.D., Limburgs Universitair Centruna. Belgium. Digan, M. E., Roberts, D. N., Enderlin, F. E., Woodworth, A. R. and Kramer, S. J. (1992). Characterization of the precursor for Manduca sexta diuretic hormone M~IsDH. Proc. Natl Aead. Sei. USA 89, 11 074-11 078. Donly, B. C., Ding, Q., Tobe, S. S. and Bendena, W. G. (1993). Molecular cloning of the gene t\~r the allatostatin family of neuropeptides fi'om the cockroach Diploptera pmwtata. Proe. Nat/Acad. Sci. USA 90, 8807 8811. Dow, J. A. T. (1981). Countercurrent flows, water movements and nutrient absorption in the locust midgut. J. l , secl Physiol. 27, 579 585. Dow, J. A. T. and Davies, S. A. (2001). The Drosophiltt melanogaster Malpighian tubule. Adv. Insect Physiol. 28, 1 83. Eckert, M., Predel, R. and Gundel, M. (1999). Periviscerokinin-like immunoreactivity in the nervous system of the American cockroach. Cell Tissue Res. 295, 159 170. Eigenheer, R. A., Nicolson, S. W., Schegg, K. M., Hull, J. J. and Schooley, D. A. (2002). Identification of a potent antidiuretic factor acting on beetle Malpighian tubules. Proc. Natl Acad. Sei. USA 99, 84 89. Eigenheer, R. A., Wiehart, U. M., Nicolson, S. W., Schools. L.. Hull, J. J., Schegg, K. M. and Schooley, D. A. (2003). Isolation, identification and localization of a second beetle antidiuretic peptide. Peptides (in press). Emery, S. B., Ma, M. C., Wong, W. K. R.. Tips, A. and De Loof, A. (1994). Immunocytochemical localization of a diuretic peptide Mamluca diuresin (MasDPll) in the brain and suboesophageal ganglion of the tobacco hawkmoth M a , d , ea sexta (Lepidoptera: Sphingidae). Arch. htseet Bioehem. Physiol. 27, 137 152. Endo, H., Nagasawa, H. and Watanabe, T. (2000). Isolation of a eDNA encoding ~l CHH-f:lmily peptide from the silkworm Bombvx mori. hTsect Biochem. Mol. Biol. 30, 355 361. Esch, F. S., Ling, N. C. and B6hlen. P. (1983). Microisolation of neuropeptides. Methods Enzymol. 103, 72 89. Farmer, J.. Maddrell, S. H. P. and Spring, J. H. (1981). Absorption of fluid by the midgut of RhocbEus. J. Exp. Biol. 94, 301 316. Fournier, B. (1991). Neuroparsins stimulate inositol phosphate formation in locust rectal cells. Comp. Biochem. Ph)'siol. B99, 57 64. Fournier. B. and Girardie, J. (1988). A new function t\~r locust neuroparsins: stimulation of water reabsorption. J. h~seet Physiol. 34, 309 313. Fournier, B., Herault, J. P. and Proux, J. (1987). Study of an antidiuretic factor from the nervous lobes of the migratory locust corpora cardiaca. Improvement of an existing bioassay. GeH. Comp. E, docrimd. 68, 49 56. Furuya, K., Schegg, K. M., Wang, H., King, D. S. and Schooley, D. A. (1995). Isolation and identitication of a diuretic hormone fi'om the mealworm Te,ebrio molitor. Proc. Natl Acad. Sci. USA 92, 12323 12327. Furuya, K., Schegg, K. M. and Schooley, D. A. (1998). Isolation and identification of a second diuretic hormone from Tenehrio molitor. Peptides. 19, 619 626.

394

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D, A. SCHOOLEY

Furuya, K., Harper, M. A., Schegg, K. M. and Schooley, D. A. (2000a). Isolation and characterization of CRF-related diuretic hormones from the whitelined sphinx moth ttyles lineata. § Bioehem. Mol. Biol. 30, 127 133. Furuya, K., Milchak, R. J., Schegg, K. M., Zhang, J., Tobe, S. S., Coast, G. M. and Schooley, D. A. (2000b). Cockroach diuretic hormones: characterization of a calcitonin-like peptide in insects. Proc. Natl Acad. Sei. USA 97, 6469 6474. Giannakou, M. E. and Dow, J. A. T. (2001). Characterization of the Drosophila mehtnogaster alkali-metal/proton exchanger (NHE) gene family. J. Exp. Biol. 204, 3703 3716. Gillett, J. D. (1982). Diuresis in newly emerged, unfed mosquitoes. I. Fluid loss in normal females and males during the first 20 hours of adult life. Proc. R. So~. Lond. B216, 201 207. Gilman, A. G. ([972). Protein binding assays for cyclic nucleotides. Adv. (~vclic Nueleotide Res. 2, 9 23. Girardie, J. and R6my, C. (1980). Particularit6s histo-cytologiques des prolongements distaux des 2 cellules fi 'vasopressine-neurophysine-like' du cricquet migrateur. J. Physiol. Paris 76, 265 271. Girardie, J., Girardie, A., Huet, J.-C. and Pernollet, J.-C. (1989). Amino acid sequence of locust neuroparsins. FEBS Lett. 248, 4 8. Girardie, J., Huet, J. C. and Pernollet, J. C. (1990). The locust neuroparsin-A sequence and similarities with vertebrate and insect polypeptide hormones. § Biochem. 20, 659 666. Girardie, J., Chaabihi, H., Fournier, B., Lagueux, M. and Girardie, A. (2001). Expression of neuroparsin cDNA in insect cells using baculovirus vectors. Arch. bisect Bioehem. Physiol. 46, 26 35. Goldbard, G. A., Sauer, J. R. and Mills, R. R. (1970). Hormonal control of excretion in the American cockroach, II. Preliminary purification of a diuretic and antidiuretic hormone. Comp. Gen. Pharmaeol. 1, 82 86. Grieco, M. A. B. and Lopes, A. G. (1997). 5-Hydroxytryptamine regulates the (Na + + K +) ATPase activity in Malpighian tubules of Rho~&ius prolixus: evidence for involvement of G-protein and cAMP-dependent protein kinase. Arch. § Bioehem. Physiol. 36. 203 214. Grillot, J. P. (1983). In 'Neurohaemal Organs of Arthropods' (A. P. Gupta, ed.), Morphology and evolution of perisympathetic organs in insects, pp. 481 512, Thomas, Springfield IL. Griss, C. (1989). Serotonin-immunoreactive neurons in the suboesophageal ganglion of the caterpillar of the hawk moth Manduea sexta. Cell Tissue Res. 258, 101 109. Hadley, N. F., Toolson, E. C. and Quinlan, M. C. (1989). Regional differences in cuticular permeability in the desert cicada Diceroprocta apache: implications for evaporative cooling. J. Exp. Biol. 141,219 230. Haeften, T. V. and Schooneveld, H. (1992). Serotonin-like immunoreactivity in the ventral nerve cord of the Colorado potato beetle, Leptinotarsa deeeml&eata: identification of five different neuron classes. Cell Tissue Res. 270, 405~413. Halcy, C. A. and O'Donnell, M. J. (1997). K + reabsorption by the lower Malpighian tubule of Rhodnius prolixus: inhibition by Ba 2+ and blockers of H+/K+-ATPases. J. £\-p. Biol. 200, 139 147. Haley, C. A., Fletcher, M. and O'Donnell, M. J. (1997). KCI reabsorption by the lower Malpighian tubule of Rho~&ius prolixus: inhibition by C1- channel blockers and acetazolamidc. J. bisect Physiol. 43, 657 665.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

395

Hanrahan, J. F. and Phillips, J. E. (1984). KC1 transport across an insect epithelium. 11 Electrochemical potentials and electrophysiology. J. Memhr. Biol. 80, 2747. H ansen-Bay, C. M. (1978). Control of salivation in the blowfly Calliphora. J. Kvp. Biol. 75, 189 201. Harrison, J. F. and Kennedy, M. J. (1994). In vitro studies of the acid base physiology of grasshoppers: the effect of feeding on acid base and nitrogen excretion. Physiol. Zoo/. 67, 120 141. Hayes, T. K., Pannabecker, T. L., Hinkley, D. J., Hohnan, G. M., Nachman, R. J., Petzel, D. H. and Beyenbach, K. W. (1989). Leucokinins, a new family of ion transport stimulators and inhibitors in insect Malpighian tubules. Lifi" Sci. 44, 1259 1266. Hayes, T. K., Holman, G. M., Pannabecker, T. L., Wright, M. S., Strey, A. A., Nachman, R. J., Hoel, D. F., Olson, J. K. and Beyenbach, K. W. (1994). Culekinin depolarizing peptidc: a mosquito leucokinin-like peptide that influences insect Malpighian tubule ion transport. Regul. Pept. 52, 235 248. ttazelton, S. R., Parker, S. W. and Spring, J. H. (1988). Excretion in the house cricket (Acheta domesticus): fine structure of the Malpighian tubules. Tissue Cell 20, 443M60. Hegarty, J. L., Zhang, B., Carroll, M. C., Cragoe, ,1. E. J. and Beyenbach, K. W. (1992). Effects of amiloride on isolated Malpighian tubules of the yellow fever mosquito (Aedes aegypti). J. Insect Physiol. 38, 329 337. Hegarty, J. L., Zhang, B.. Petzel, D. H., Baustian, M. D., Pannabecker, T. L. and Beyenbach, K. W. (1991). Dibutyryl cAMP activates bumetanide-sensitive electrolyte transport in Malpighian tubules. Am. J. Physiol. 261, C521 C529. Helle, J., Dircken, H., Eckert, M., Niissel, D. R., Sporhase-Eichmann, U. and Schurmann, F.-W. (t995). Putative neurohemal areas in the peripheral nerwms system of an insect, Grvllus himaculatus, revealed by immunocytochemistry. ('ell Tissue Res. 281, 43 61. Herault, J. P. E. and Proux, J. P. (1987). Cyclic AMP, the second messenger of an antidiuretic hormone from glandular lobes of migratory locust CC. J. Insect Physiol. 33, 487 492. Herault, J. P. E., Girardie, J. and Proux, J. P. (1985). Separation and characteristics of antidiuretic factors from the glandular lobes of the migratory locust corpora cardiaca. Int. J. Invert. Reprod. Dev. 8, 325 335. Hewes. R. S. and Taghert, P. H. (2001). Neuropeptides and neuropeptide receptors in the Drosol~hiht mehmogaster genome. Genome Res. 11, 1126 1142. Hietter, H., Van Dorsselaer, A. and Luu, B. (1991). Characterization of 3 structurallyrelated 8 9 kDa monomeric peptides present in the corpora cardiaca of Locusta a revised structure for the neuroparsins, hTsect Biochem. 21, 259 264. Highnam, K. C. and West, M. W. (1971). The neuropilar neurosecretory reservoir of Locusta migratoria migratorioMes R & F. Gen. Comp. Emlocrinol. 16, 574-585. Holman, G. M. and Hayes, T. K. (1997). In +Neurotransmitter Methods' (R. C. Rayne, ed.), HPLC methods to isolate peptide neurotransmitters, pp. 205 218, Humana Press, Totowa, NJ. Hohnan, G. M., Cook, B. J. and Nachman, R. J. (1986a). Isolation, primary structure and synthesis of two neuropeptides from Leucophaea maderae: members of a new family of cephalomyotropins. Comp. Biochem. Physiol. C84, 205 211. Holman, G. M., Cook, B. J. and Nachman, R. J. (1986b). Primary structure and synthesis of a blocked myotropic neuropeptide isolated fiom the cockroach, Leucophaea maderae. ('omp. Biochem. Physiol. C85, 219 224.

396

G . M . COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

Holman, G. M., Cook, B. J. and Nachman, R. J. (1986c). Primary structure and synthesis of two additional neuropeptides from Leucophaea ma&,rae: members of a new family of cephalomyotropins. Comp. Biochem. Physiol. C84, 271-276. Holman, G. M., Cook, B. J. and Nachman, R. J. (1987a). Isolation, primary structure, and synthesis of leucokinins V and VI: myotropic peptides of Leueophaea maderae. Comp. Biochem. Physiol. C88, 27 30. Hohnan, G. M., Cook, B. J. and Nachman, R. J. (1987b). Isolation, primary structure and synthesis of leucokinins VII and VllI: the linal members of this new family of cephalomyotropic peptides isolated from head extracts of Leueophaea maderae. Comp. Bioehem. Physiol. C88, 31 34. Holman, G. M., Nachman, R. J. and Wright, M. S. (1990). In "Progress in Comparative Endocrinology" (A. Epple, C. G. Scanes and M. H. Stetson, eds), Comparative aspects of insect myotropic peptides, pp. 35 39, Wiley-Liss, New York. Holman, G. M., Nachman, R. J., Schoofs, L., Hayes, T. K., Wright, M. S. and De koof, A. (1991). The Leueophaea maderae hindgut preparation a rapid and sensitive bioassay tool for the isolation of insect myotropins of other insect species. Insect Biochem. 21, 107 112. Holman, G. M., Nachman, R. J. and Coast, G. M. (1999). Isolation, characterization and biological activity of a diuretic myokinin neuropeptide from the housefly, Musca domestiea. Peptides 20, 1 10. Holmes. S. P., He, H., Chen, A.C., lvie, G. W. and Pietrantonio, P. V. (2000). Cloning and transcriptional expression of a leucokinin-like peptide receptor from the Southern cattle tick, Boophilus mieroplus (Acari: Ixodidae). hlseet Mol. Biol. 9, 457M65. Homberg, U. and Hildebrand, J.G. (1989). Serotonin-immunoreactive neurons in the median protocerebrum and suboesophageal ganglion of the sphinx moth Mamhtca se.vta. Cell Tissue Res. 258, 1 24. Homer, M. (1999). Cytoarchitecture of histamine-, dopamine-, serotonin- and octopamine-containing neurons in the cricket ventral nerve cord. Microscopy Res. Teeh. 44, 137 165. Huesmann, G. R., Cheung, C. C., Loi, P. K., Lee, T. D., Swiderek, K. M. and Tublitz, N. J. (1995). Amino acid sequence of CAP,b, an insect cardioacceleratory peptide from the tobacco hawkmoth Manduea sexta, kZ'BS Lett. 371,311 314. Hustert, R. and Topel, U. (1986). Location and major postembryonic changes of identified 5-HT-immunoreactive neurones in the terminal ganglion of a cricket (Acheta domesticus). Cell Tissue Res. 245, 615 621. laboni, A., Hohnan, G. M., Nachman, R. J., Orchard, I. and Coast, G. M. (1998). lmmunocytochemical localisation and biological activity of diuretic peptides in the housefly, Musea domestiea. Cell Tissue Res. 294, 549 560. lanowski. J. P. and O'Donnell, M. J. (2001). Transepithelial potential in Malpighian tubules of Rhodnius prolixus: lumen-negative voltages and the triphasic response to serotonin. J. bisect Physiol. 47, 411 421. Ianowski, J. P., Christensen, R. J. and O'Donnell, M. J. (2002). Intracellular ion activities in Malpighian tubule cells of Rhodnius prolixus: evaluation of Na +K+-2C1 cotransport across the basolateral membrane. J. Exp. Biol. 205, 1645 1655. lrvine, B. N. A., Lechleitner, R., Meredith, J.. Thomson, B. and Phillips, J. (1988). Transport properties of locust ileum it~ vitro: effects of cAMP. J. Exp. Biol. 137, 361 385. Isaac, R. E., Coates, D., Williams, T. A. and Schoot~, k. (1998). In 'Recent Advances in Arthropod Endocrinology' (G. M. Coast and S. G. Webster, eds), Insect

INSECT DIURETIC AND ANTIDIURETIC HORMONES

397

angiotensin-converting enzyme: comparative biochemistry and evolution, pp. 357 378, Cambridge University Press, Cambridge. lto, M., Matsuo, Y. and Nishikawa, K. (1997), Prediction of protein secondary structure using the 3D I D compatibility algorithm. C A B I O S 13, 415423. Jefl~, L. (1993). A pharmacological study of signal transduction mechanisms controlling fluid reabsorption and ion transport in the locust rectum. M.Sc., University of British Columbia, Vancouver. Jefl~, L. B. and Phillips, J. E. (1996). A pharmacological study of the second messengers that control rectal ion and fluid transport in the desert locust (Schistocerca y,re~aria). Arch. Insect Biochem. Physiol. 31, 169 184. Kataoka, H., Troetschler, R. G., Li, J. P., Kramer, S. J., Carney, R. L. and Schooley, D. A. (1989). Isolation and identification of a diuretic hormone from the tobacco hornworm, Manduca se.vla. Proc. Nat/Acad. Sci. USA 86, 2976 2980. Kay, 1., Coast, G. M., Cusinato, O., Wheeler, C. H., Totty, N. F. and Goldsworthy, G. J. (1991a). Isolation and characterization of a diuretic peptide fiom dcheta domesticus: evidence fbr a family of insect diuretic peptides. Biol. Chem. tfoppeSevler 372, 505 512. Kay, 1., Wheeler, C. H., Coast, G. M., Totty, N. F., Cusinato, O., Patel, M. and Goldsworthy, G. J. (1991b). Characterization of a diuretic peptide from Locusta miL,raloria. Biol. Chem. Hoptw-Seyler 372, 929 934. Kay, I., Patel, M., Coast, G. M., Totty, N. F.. Mallet, A. 1. and Goldsworthy, G. J. (1992). Isolation, characterization and biological activity of a CRF-related diuretic peptide from Peril)laneta americana L. Re gul. Pt7)t. 42, 111 122. Keeley, L. L., Chung, J. S. and Hayes, T. K. (1992). Diuretic and antifeedant actions by Mamh~ca se.vta diuretic hormone in lepidopteran larwle. E_vperientia 48, 1145 1148. Kegel, G., Reichwein, B., Weese, S., Gaus, G., Peter-Katalinic, J. and Keller, R. (1989). Amino acid sequence of the crustacean hyperglycemic hormone (CHH) from the shore crab, Carcinus maeltas. FEBS Lett. 255, 10 14. Kelly, T. J., Maslcr, E. P. and Menn, J. J. (1994). In "Natural and Derived Pest Management Agents." (P. A. Hedin, J. J. Menn and R. M. Hollingworth, eds), Insect neuropeptides: current status and avenues l\)r pest control, pp. 292 318, Oxford University Press, Washington, D.C. Kim, M. Y. (1998). Leucokinin and callitachykinin immunoreactive neurons during postembryonic development of Calliphora vomitoria (L.) (Diptera: Calliphoridae). hll. J. h~sect Moq~hol. Emho,ol. 27, 193 204. King, D. S., Meredith, J., Wang. Y. J. and Phillips, J. E. (1999). Biological actions of synthetic locust ion transport peptide (ITP). bisect Biochem. Mol. Biol. 29, 11 18. Klelnm, N., Hustert, R., Cantera, R. and Nfissel, D. R. (1986). Neurons reactive to antibodies against serotonin in the stomatogastric nervous system and in the alimentary canal of locust and crickets (Orthoptera, lnsecta). Neuroscience 17, 247 261. Konings, P. N. M., Vullings, H. G. B., Siebinga, R., Diederen, J. H. B. and Jansem W. F. (I 988). Serotonin-immunoreactive neurones in the brain of Locusta m~g,ratoria innervating the corpus cardiacum. Cell Tissue Res. 254, 147 153. Lacombe, C., Gr&ve, P. and Martin, G. (1999). Overview of the sub-grouping of the crustacean hyperglycemic hormone family. Neuropeptides 33, 71 80. Laenen, B. (1999). Purification, characterization and mode of action of endogenous neuroendocrine factors in the forest ant. Formica polvctena. Ph.D., Limburgs Universitair Centrum, Belgium. Laenen, B., De Decker, N., Steels, P., Van Kerkhove, E. and Nicolson, S. (2001). An antidiuretic factor in the forest ant: purilication and physiological effects on the Malpighian tubules. J. Insect Physiol. 47, 185 193.

398

G . M . COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

Lamango, N. S. and lsaac, R. E. (1993). Metabolism of insect neuropeptides: properties of a membrane-bound endopeptidase from heads of Musca domestica, h~sect Biochem. Mol. Biol. 23, 801 808. Lange, A. B., Orchard, I. and Lloyd, R. J. (1988). Immunohistochemical and electrochemical detection of serotonin in the nervous system of the blood-feeding bug, Rhodnius prolixus. Archs. bisect Biochem. Physiol. 8, 187 201. Lange, A. B., Orchard, I. and Barrett, F. M. (19893. Changes in haemolymph serotonin levels associated with feeding in the blood-sucking bug, Rhodnius prolixus. J. hlsect Physiol. 35, 393 399. Lavigne, C., Embleton, J., Audy, P., King, R. R. and Pelletier, Y. (2001). Partial purification of a novel insect antidiuretic factor from the Colorado potato beetle, Leptinotarsa decemlineata (Say) (Coleoptera: Chrysomelidae), which acts on Malpighian tubules. Insect Biochem. Mol. Biol. 31, 339 347. Lechleitner, R. A. and Phillips, J. E. (19893. Effect of corpus cardiacum, ventral ganglia, and proline on absorbate composition and fluid transport by locust hindgut. C~m. J. Zool. Rev. Can. Zool. 67, 2669-2675. Lechleitner, R. A., Audsley, N. and Phillips, J. E. (1989a). Antidiuretic action of cyclic AMP, corpus cardiacum, and ventral ganglia on fluid absorption across locust ileum in vitro. Can. J. Zool. Rev. Can. Zool. 67, 2655 2661. Lechleitner, R. A., Audsley, N. and Phillips, J. E. (1989b). Composition of fluid transported by locust ileum: influence of natural stimulants and luminal ion ratios. Can. J. Zool. Rev. Can. Zool. 67, 2662-2668. Lehmberg, E., Ota, R. B., Furuya, K., King, D. S., Applebaum, S. W., Ferenz, H.-J. and Schooley, D. A. (t991). Identification of a diuretic hormone of Locusta migratoriu. Biochem. Biophys. Res. Commun. 179, 1036 1041. Li, H., Wang, H., Schegg, K. M. and Schooley, D. A. (1997). Metabolism of an insect diuretic hormone by Malpighian tubules studied by liquid chromatography coupled with electrospray ionization mass spectrometry. Proc. Natl Acad. Sci. USA 94, 13463 13468. Liao, S. (2000). Characterization and purification of antidiuretic factor from Manduca sexta. Ph.D. University of Nevada, Reno. Liao, S., Audsley, N. and Schooley, D. A. (2000). Antidiuretic effects of a factor in brain/corpora cardiaca/corpora allata extract on fluid reabsorption across the cryptonephric complex of Manduca sexta. J. Exp. Biol. 203, 605 615. Linton, S. M. and O'Donnell, M. J. (1999). Contributions of K+: CI- cotransport and Na+/K+-ATPase to basolateral ion transport in Malpighian tubules of Drosophila melanogaster. J. Exp. Biol. 202, t561 1570. Linton, S. M. and O'Donnell, M. J. (2000). Novel aspects of the transport of organic anions by the Malpighian tubules of Drosophila melanogaster. J. Exp. Biol. 203, 3575-3584. Luffy, D. and Dorn, A. (1991). Serotoninergic elements in the stomatogastric nervous system of the stick insect, Carausius morosus, demonstrated by immunohistochemistry. J. Insect Physiol. 37, 269 278. Lutz, E. M. and Tyrer, N. M. (1988). immunohistochemical localization of serotonin and choline acetyltransferase in sensory neurons of the locust. J. Comp. Neurol. 267, 335 342. Ma, M., Emery, S. B., Wong, W. K. R. and De Loof, A. (2000). Effects of Manduca diuresin on neonates of the tobacco hornworm, Manduca sexta. Gen. Comp. Emlocrinol. 118, 1 7. Macins, A., Meredith, J., Zhao, Y., Brock, H. W. and Phillips, J. E. (1999). Occurrence of ion transport peptide (ITP) and ion transport-like peptide (1TP-L) in orthopteroids. Arch. hlsect Biochem. Physiol. 40, 107 118.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

399

Maddrell, S. H. P. (1963). Excretion in the blood-sucking bug, Rhodnius prolixus St/i.l. 1. The control of diuresis. J. Exp. Biol. 40, 247 256. Maddrell, S. H. P. and O'Donnell, M. ,1. (1993). Gramicidin switches transport in insect epithelia from potassium to sodium. J. Exp. Biol. 177, 287 292. Maddrelh S. H. P., O'DonnelL M. J. and Caffrey, R. (1993b). The regulation of haemolymph potassium activity during initiation and maintenance of diuresis in fed Rhodnius prolixus. J. Exp. Biol. 177, 273 285. M addrell, S. H. P. (1964a). Excretion in the blood-sucking bug, Rhochfius prolixus Stiil. II. The normal course of diuresis and the effect of temperature. J.Exp. Biol. 41, 163 176. Maddrell, S. H. P. (1964b). Excretion in the blood-sucking bug, Rho(hfius proli.\us StM. Ill. The control of the release of the diuretic hormone. J. Exp. Biol. 41,459 472. Maddrell, S. H. P. (1969). Secretion by the Malpighian tubules of Rhothfius. The movements of ions and water. J. Exp. Biol. 51, 71 97. Maddrell, S. H. P. (1978). In "Membrane Transport in Biology' (G. Giebisch, D. C. Tosleson and H. H. Ussing, eds), Transport across insect excretory epithelia, pp. 239 27t, Springer-Verlag, Heidelberg. Maddrell, S. H. P. (1980). In 'Insect Biology in the Future" (M. Locke and D. S. Smith. cds). The control of water relations in insects, pp. 179 199, Academic Press, New York. Maddrell, S. H. P. (1981). The functional design of the insect excretory system. J. Exp. Biol. 90, 1 15. Maddrell, S. H. P. and Gardiner, B. O. C. (1974). The passive permeability of insect Malpighian tubules to organic solutes. J. Exp. Biol. 60, 641-652. Maddrell, S. H. P. and Gardiner, B. O. C. (1976a). Diuretic hormone in adult Rho~hfius prolixus: total store and speed of release. Physiol. Entomol. 1, 265 269. Maddrelh S. H. P. and Gardiner. B. O. C. (1976b). Excretion of alkaloids by Malpighian tubules. J. Exp. Biol. 64, 267 281. Maddrell, S. H. P. and Gardiner, B. O. C. (1980). The retention of amino acids in the haemolymph during diuresis in Rho~hfius. J. Exp. Biol. 87, 315 329. Maddrell, S. H. P. and O'Donnell, M. J. (1992). Insect Malpighian tubules: V-ATPase action in ion and fluid transport. J. Lvp. Biol. 172, 417 429. Maddrell, S. H. P. and Phillips, J. E. (1975). Secretion of hypo-osmotic fluid by the lower Malpighian tubules of Rhochfius prolixus. J. Exp. Biol. 62, 671 683. Maddrell, S. H. P., Pilcher. D. E, M. and Gardiner, B. O. C. (1969). Stimulatory effect of 5-hydroxytryptamine (serotonin) on the secretion by Malpighian tubules of insects. Natm'e 222, 784 785. Maddrell, S. H. P., Pilcher, D. E. M. and Gardiner, B. O. C. (1971). Pharmacology ot the Malpighian tubules of Rho~htius and Carausiu,< the structure activity relationship of tryptamine analogues and the role of cyclic AMP. J. £v/~. Biol. 54, 779 804. Maddrelh S. H. P., Gardiner, B. O. C., Pilcher, D. E. M. and Reynolds, S. E. (1974). Active transport by insect Malpighian tubules of acidic dyes and of acylamides. J. Exp. Biol. 61, 357 377. Maddrell, S. H. P., Herman, W. S., Mooney, R. L. and Overton, J. A. (1991). 5Hydroxytryptamine: a second diuretic hormone in Rhmbfius proli.vu.s'. J. Exp. Biol. 156, 557-566. Maddrell, S. H. P., Herman. W. S., Farndale. R. W. and Riegel, .I.A. (1993). Synergism of hormones controlling epithelial fluid transport in an insect. J. E.vp. Biol. 174. 65 80. MacPherson, M. R., Pollock, V. P., Broderick, K. E., Kean, L., O'Connell, F. C., Oow, J. A. T. and Davies, S. A. (2001). Model organisms: new insights into ion channel

400

G . M . COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

and transporter function. L-type calcium channels regulate epithelial fluid transport in Drosophila melanogaster. Am. J. Physiol. Cell Physiol. 280, C394-C407. Maeda, S. (1989). Increased insecticidal effect by a recombinant baculovirus carrying a synthetic diuretic hormone gene. Bioehem. Biophys. Res. Commun. 165, 1177 1183. Marshall, A. T. and Xu, W. (1999). Use of Rb + and Br- as tracers for investigating ion transport by X-ray microanalysis in the Malpighian tubules of the black field cricket Teleogryllus oceanicus. J. Insect Physiol. 45, 265-273. Masler, E. P., Kelly, T. J. and Menn, J. J. (1993). Insect neuropeptides discovery and application in insect management. Arch. In.wet Biochem. Physiol. 22, 87-q 11. Meola, S. M., Clottens, F. L., Coast, G. M. and Holman, G. M. (1994). Localization of leucokinin VIII in the CNS of the cockroach, Leucophaea maderae, using an antiserum directed against an achetakinin-I analog. Neurochem. Res. 19, 805 814. Meredith, J. and Phillips, J. E. (1988). Sodium-independent proline transport in the locust rectum. J. Exp. Biol. 137, 341-360. Meredith, J., Moore, L. and Scudder, G. G. E. (1984). Excretion of ouabain by Malpighian tubules of Oneopeltus'j~lsciatus. Am. J. Physiol. 246, R705 R715. Meredith, J., Ring, M., Macins, A., Marschall, J., Cheng, N. N., Theilmann, D., Brock, H. W. and Phillips, J. E. (1996). Locust ion transport peptide (ITP): primary structure, cDNA and expression in a baculovirus system. J. Exp. Biol. 199, 1053 1061. Miksys, S. and Orchard, I. (1994). Immunogold labelling of serotonin-like and FMRFamide-like immunoreactive material in neurohaemal areas on abdominal nerves of Rhodnius prolixus. Cell Tissue Res. 278, 145 151. Miranda, A., Lahrichi, S. L., Gulyas, J., Koerber, S. C., Craig, A. G., Corrigan, A., Rivier, C., Vale, W. and Rivier, J. (1997). Constrained corticotropin-releasing factor antagonists with i-(i + 3) Glu Lys bridges. J. Med. Chem. 40, 3651 3658. Moffett, D. F. (1994). Recycling of K +, acid-base equivalents, and fluid between gut and hemolymph in lepidopteran larvae. Physiol. Zool. 67, 68-81. Montoreano, R., Triana, F., Abate, T. and Rangel-Aldao, R. (1990). Cyclic AMP in the Malpighian tubule fluid and in the urine of Rhodnius prolixus. Gen. Comp. Endoerinol. 77, 136 142. Montuengm L. M., Zudaire, E., Prado, M. A., Audsley, N., Burrell, M. A. and Coast, G. M. (1996). Presence of Locusta diuretic hormone in endocrine cells of the ampullae of locust Malpighian tubules. Cell Tissue Res. 285, 331 339. Mordue, W. (1969). Hormonal control of Malpighian tubule and rectal function in the desert locust, Schistoeerca gregaria. J. Insect Physiol. 15, 273 285. Mordue, W. (1972). Hormones and excretion in locusts. Gen. Comp. Endocrinol. Suppl. 3, 289 298. Morgan, P. J. and Mordue, W. (1984). 5-hydroxytryptamine stimulates fluid secretion in locust Malpighian tubules independently of cAMP. Comp. Biochem. Physiol. C79, 305 310. Moyna, G., Williams, H. J., Nachman, R. J. and Scott, A. I. (1999). Detection of nascent polyproline 11 helices in solution by NMR in synthetic insect kinin neuropeptide mimics containing the X- Pro-Pro-X motif. J. P¢7)t. Res. 53, 294-301. Muren, J. E., Lundquist, C. T. and Niissel, D. R. (t993). Quantitative determination of myotropic neuropeptide in the nervous system of the cockroach Leueophaea maderae: distribution and release of leucokinin. J. Exp. Biol. 179, 289 300. Nachman, R. J., Coast, G. M., Hohnan, G. M. and Beier, R. C. (1995). Diuretic activity of C-terminal group analogs of the insect kinins in Aeheta domesticus. Peptides 16, 8(19 813.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

401

Nachman, R. J., Isaac, R. E., Coast, G. M. and Hohnan, G. M. (1997). Aib-containing analogues of the insect kinin neuropeptide family demonstrate resistance to an insect angiotensin-converting enzyme and potent diuretic activity. Peptides 18, 53 57. Nachman, R. J., Holman, G. M. and Coast, G. M. (1998). In 'Recent Advances in Arthropod Endocrinology' (G. M. Coast and S. G. Webster, eds). Mimetic analogues of the myotropic/diuretic insect kinin neuropeptide family, pp. 379 391. Cambridge University Press, Cambridge. Nachman, R. J., Zabrocki, J,, Roberts, V. A. and Coast, G. M. (2000). In ~Peptides: Biology and Chemistry" (X. Y. Hu, R. Wang and J. P. Tam, eds), Tetrazole e/s'-amide bond mimetics identify the beta-turn conformation of insect kinin neuropeptides, pp. 158 162. Kluwer Academic Publishers, Dordrecht, The Netherlands. Nachman, R. J., Strey, A., Isaac, E., Pryor, N., Lopez, J. D., Deng, J.-G. and Coast, G. M. (2002a). Enhanced in vivo activity of peptidase-resistant analogs of the insect kinin neuropeptide family. Peptides 23, 735 745. Nachman, R. J., Zabrocki, J., Olczak, J., Williams, H. J., Moyna, G., Scott, A. I. and Coast, G. M. (2002b). cLs-peptide bond mimetic tetrazole analogs of the insect kinins identify the active conformation. Peptides 23, 709 716. Nfissel, D. R. (1988). Serotonin and serotonin-immunoreactive neurons in the nervous system of insects. Progress in Neurohiol. 30, 1 85. N/,issel, D. R. (1993a). Insect myotropic peptides: differential distribution of locustatachykinin- and leucokinin-like immunoreactive neurons in the locust brain. Cell Tissue Rex. 274, 27 40. N/,issel, D. R. (1993b). Neuropeptides in the insect brain: a review. Cell Tissue Res. 273, l 29. N/issel, D. R. (1996a). Advances in the immunocytochemical localization of neuroactive substances in the insect nervous system. J. Neurosei. Meth. 69, 3 23. N'assel, D. R. (1996b). Neuropeptides, amines and amino acids in an elementary insect ganglion: functional and chemical anatomy of the unfused abdominal ganglion. Pro g. Neurohiol. 48, 325 331. N/isscl, D. R. (1996c). Peptidergic neurohormonal control systems in invertebrates. Curr. Opin. Neurobiol. 6, 842 850. Nfissel, D. R. and Elekes, K. (1985). Serotonergic terminals in the neural sheath of the blowfly nervous system: electronmicroscopical immunocytochemistry and 5.7dihydoxytryptamine labelling. Neuroseience 15, 293 307. N/issel, D. R. and Lundquist, C. T. (1991). Insect tachykinin-like peptide distribution of leucokinin immunoreactive neurons in the cockroach and blowfly brains. Nemosei. Lett. 130, 225 228. N'assel, D. R., Cantera, R. and Karlsson, A. (1992). Neurons in the cockroach nervous system reacting with the antisera to the neuropeptide leucokinin I. J. Comp. Nemol. 322, 45 67. Nicolson, S. W, (1976). Diuresis in the cabbage white butterfly, Pieris hrassicae: fluid secretion by the Malpighian tubules. J. Insect Physiol. 22, 1347 1356, Nicolson, S. W. (1991). Diuresis or clearance: is there a physiological role for the diuretic hormone of the desert beetle Onymaeris? J. bisect Physiol. 37, 447 452. Nicolson, S. W. (1993). The ionic basis of fluid secretion in insect Malpighian tubules: advances in the last ten years. J. hTsect Physiol. 39, 451M58. Nicolson, S. W. and Hanrahan, S. A. (1986). Diuresis in a desert beetle'? Hormonal control of the Malpighian tubules of Onymacris plana (Coleoptera: Tenebrionidae). J. Comp. Physiol. Blfi6, 407~413.

402

G. M. COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

Nielsen, H., Engelbretch, J., Brunak, S. and Von Heijne, G. (1997). Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Proteh7 Eng. 10, 1 -6. Nieto, J., Veelaert, D., Derua, R., Waelkens, E., Cerstiaens, A., Coast, G., Devreese, B., van Beeurnen, J., Calderon, J., de Loof\ A. and Schools, L. (1998). Identification of one tachykinin and two kinin-related peptides in the brain of the white shrimp, Penaeus vannamei. Biochem. Biophys. Res. Commun. 248, 406411. Nittoli, T., Coast, G. M. and Sieburth, S. M. (1999). Evidence for helicity in insect diuretic peptide hormones: computational analysis, spectroscopic studies, and biological assays. J. Pept. Res. 53, 99 108. Norris, M. J. (1961). Group effects on feeding in adult males of the desert locust, Schistocer~a gregaria (Forsk.), in relation to sexual maturation. Bull. Entomol. Res. ill, 731-753. O'Connor, K. R. and Beyenbach, K. W. (2001). Chloride channels in apical membrane patches of stellate cells of Malpighian tubules of Aedes aegypti. J. E.\p. Biol. 204, 367 378. O'Cuinn, G., O'Connor, B., Gilmartin, L. and Smyth, M. (1995). In "Metabolism of Brain Peptides' (G. O'Cuinn, ed.), Neuropeptide inactivation by peptidases, pp. 99 158, CRC Press, Boca Raton. O'Donnell, M. J. and Machin, J. (1991). Ion activities and electrochemical gradients in the mealworm rectal complex. J. Exp. Biol. 155, 375~402. O'Donnell, M. J. and Maddrell, S. H. P. (1984). Secretion by the Malpighian tubules of Rhodnius prolixus Still: electrical events. J. Exp. Biol. II0, 275 290. O'Donnell, M. J. and Spring, J. H. (2000). Modes of control of insect Malpighian tubules: synergism, antagonism, cooperation and autonomous regulation. J. h~sect Physiol. 46, 107 117. O'Donnell, M. J., Aldis, G. K. and Mad&ell, S. H. P. (1982). Measurements of osmotic permeability in the Malpighian tubules of an insect, Rhodnius prolixus Still. Proc. R. Soc. Lond. B216, 267-277. O'Donnell, M. J., Maddrell, S. H. P. and Gardiner, B. O. C. (1983). Transport of uric acid by the Malpighian tubules of Rhodnius prolixus and other insects. J. Exp. Biol. 103, 169 184, O'Donnell, M. J,, Dow, J. A. T., Huesmann, G. R., Tublitz, N. J. and Maddrell, S. H. P. (1996). Separate control of anion and cation transport in Malpighian tubules of Drosophila mehmogaster. J. Exp. Biol. 199, 1163 1175. O'Donnell, M. J., Rheault, M. R., Davies, S. A., Rosay, P., Harvey, B. J., Maddrell, S. H. P., Kaiser, K. and Dow, J. A. T. (1998). Hormonally controlled chloride movement across Drosophila tubules is via ion channels in stellate cells. Am. J. Physiol. 274, R1039 R1049. Orchard, 1. (1989). Serotonergic neurohaemal tissue in Rhodnius prolixus: synthesis, release and uptake of serotonin. J. Insect Physiol. 35, 943 947. Orchard, I. and Brugge, V. T. (2002). Contractions associated with the salivary glands of the blood-feeding bug, Rhodnius prolLvus: evidence for both a neural and neurohormonal coordination. Peptides 23, 693-700. Orchard, l., Lange, A. B. and Barrett, F. M. (1988). Serotonergic supply to the epidermis of Rhodnius prolixus: evidence for serotonin as the plasticising factor. J. Insect Physiol. 34, 873-879. Orchard, I., Lange, A. B., Cook, H. and Ramirez, J. M. (1989). A subpopulation of dorsal, unpaired, median neurons in the blood feeding insect Rhodnius prolixus displays serotonin-like immunoreactivity. J. Comp. Neurol. 289, l l8 128. Pannabecker, T. L., Hayes, T. K. and Beyenbach, K. W. (1993). Regulation of epithelial shunt conductance by the peptide leucokinin. J. Memh. Biol. 132, 63-76.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

403

Patel, M., Chung, J.-S., Kay, 1., Mallet, A. 1., Gibbon, C. R., Thompson, K. S. J., Bacon, J. P. and Coast, G. M. (1994). Localization of Locusta-DP in locust CNS and hemolymph satisfies initial hormonal criteria. Peptides 15, 591-602. Patel, M., Hayes, T. K. and Coast, G. M. (1995). Evidence for the hormonal function of a CRF-related diuretic peptide (Locusta-DP) in Locusta migratoria..1. Exp. Biol. 198, 793 804. Payne, J. A. and Forbush, B. (1995). Molecular characterization of the epithelial Na K CI cotransporter isoforms. Curt. Opin. Cell Biol. 7, 493 503. Petzel, D. H., Hagedorn, H. H. and Beyenbach, K. W. (1985). Preliminary isolation of mosquito natriuretic factor. Am. J. Physiol. 249, R379 R386. Petzel, D. H., Hagedorn, H. H. and Beyenbach, K. W. (1986). Peptide nature of two mosquito natriuretic factors. Am. J. Physiol. 250, R328 R332. Petzel, D. H., Berg, M. M. and Beyenbach, K. W. (1987). Hormone-controlled cAMPmediated fluid secretion in yellow-fever mosquito. Am. J. Physiol. 253, R701 R711. Pfeifer, T. A., Hegedus, D., Wang, Y. J., Zhao, Y., Meredith, J., Brock, H. W., Phillips, J. E., Grigliatti, T. A. and Theihnann, D. A. (1999). Analysis of an insect neuropeptide, Schistoeerca gregaria ion transport peptide (ITP), expressed in insect cell systems. Arch. bisect Biochem. Physiol. 42, 245 252. Phillips, J. E. (1981). Comparative physiology of insect renal function. Am. J. Physiol. 241, R241 R257. Phillips, J. E. (1983). In 'Endocrinology of Insects' (R. G. H. Downer and H. Laufer, eds), Endocrine control of salt and water balance: excretion, pp. 411 425, A. R. Liss, New York. Phillips, J. E. and Audsley, N. (1995). Neuropeptide control of ion and fluid transport across locust hindgut. Am. Zool. 35, 503 514. Phillips, J. E., Mordue, W., Meredith, J. and Spring, J. (1980). Purification and characteristics of the chloride transport stimulating factor from locust corpora cardiaca: a new peptide. Can. J. ZooL-Rev. Can. Zool, 58, 1851 1860. Phillips, J., Spring, J., Hanrahan, J., Mordue, W. and Meredith, J. (1982). In 'Neurosecretion: Molecules, cells, systems" (D. S. Farner and K. Lederis, eds), Hormonal control of salt reabsorption by the excretory system: isolation of a new protein, pp. 371 380, Plenum, New York. Phillips, J. E.. Hanrahan, J., Chamberlin, M. and Thomson, B. (1986). Mechanisms and control of reabsorption in insect hindgut. Adv. h~sect Physiol. 19, 330-422. Phillips, J. E., Audsley, N., Lechleitner, R., Thomson, B., Meredith, J. and Chamberlin, M. (1988). Some major transport mechanisms of insect absorptive epithelia. Comp. Biochem. Physiol. A90, 643 650. Phillips, J. E., Thomson, R. B., Audsley, N., Peach, J. L. and Stagg, A. P. (1994). Mechanisms ot" acid-base transport and control in locust excretory system. Physiol. Zool. 67, 95 119. Phillips, J. E., Wiens, C., Audsley, N., Jells, L., Bilgen, T. and Meredith, J. (1996). Nature and control of chloride transport in insect absorptive epithelia. J. Evp. Zoo/. 275, 292 299. Phillips, J. E., Meredith, J., Audsley, N., Richardson, N., Macins, A. and Ring, M. (1998a). Locust ion transport peptide (ITP): a putative hormone controlling water and ionic balance in terrestrial insects. Am. gool. 38, 461~70. Phillips, J. E., Meredith, J., Audsley, N., Ring, M., Macins, A., Brock, H., Theilmann, D. and Tittleford, D. (1998b). In 'Recent Advances in Arthropod Endocrinology" (G. M. Coast and S. G. Webster, eds), Locust ion transport peptide (ITP): function, structure, cDNA and expression, pp. 210 226, Cambridge University Press. Cambridge.

404

G. M. COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

Phillips, J., Meredith, J., Wang, Y., Zhao, Y. and Brock, H. (2001). In 'Perspective in Comparative Endocrinology: Unity and Diversity' (H. J. Goos, R. K. Rastogi, H. Vaudry and R. Pierantoni, eds), Ion Transport Peptide (ITP): structure, function, evolution, pp. 745-752. Medimond, Bologna. Picquot, M. and Proux, J. (1987). Relationship between excretion of primary urine and haemolymph level of diuretic hormone in the migratory locust. Physiol. Entomol. 12, 455460. Pietrantonio, P. V., Gibson, G. E., Strey, A. A., Petzel, D. and Hayes, T. K. (2000). Characterization of a leucokinin binding protein in Aedes aegypti (Diptera: Culicidae) Malpighian tubule. Insect Biochem. Mol. Biol. 30, 1147 1159. Pietrantonio, P. V., Jagge, C. and McDowell, C. (2001). Cloning and expression analysis of a 5HTT-like serotonin receptor cDNA from mosquito Aedes aeg:vpti female excretory and respiratory systems, bisect Mol. Biol. 10, 357 369. Predel, R. (2001). Peptidergic neurohemal system of an insect: mass spectrometric morphology. J. Comp. Neurol. 436, 363 375. Predel, R. and Giide, G, (2002). Identification of the abundant neuropeptide from abdominal perisympathetic organs of locusts. Peptides 23, 621 627. Predel, R., Linde, D., Rapus, J., Vettermann, S. and Penzlin, H. (1995). Periviscerokinin (Pea-PVK): a novel myotropic neuropeptide from the perisympathetic organs of the American cockroach. P~7~tides 16, 61 66. Predel, R., Kellner, R., Rapus, J., Penzlin, H. and G/ide, G. (1997). Isolation and structural elucidation of eight kinins from the retrocerebral complex of the American cockroach, Periphmeta americana. Re gul. Pepl. 71, 199 205. Predel, R., Rapus, J., Eckert, M., Holman, G. M., Nachman, R. J., Wang, Y. and Penzlin, H. (1998). Isolation of periviscerokinin-2 from the abdominal perisympathetic organs of the American cockroach, Periphmeta americana. Peptides 19, 801 809. Predel, R., Kellner, R., Baggerman, G., Steinmetzer, T. and Schools, L. (2000). Identification of novel periviscerokinins from single neurohaemal release sites in insects. Eur. J. Biochem. 267, 3869 3873. Predel, R., Nachman, R. J. and G/ide, G. (2001). Myostimulatory neuropeptides in cockroaches: structures, distribution, pharmacological activities, and mimetic analogs. J. Insect Physiol. 47, 311 324. Proux, B. H., Proux, J. P. and Phillips, J. E. (1984). Antidiuretic action of corpus cardiacum (CTSH) on long-term fluid absorption across locust recta #l vitro. J. Exp. Biol. 113. 409 421. Proux, J. P. and Herault, J.-P. (1988). Cyclic AMP: a second messenger of the newly characterized AVP-like insect diuretic hormone, the migratory locust diuretic hormone. Neuropeptides 12, 7 12. Proux, J., Rougon, G. and Cupo, A. (1982). Enhancement of excretion across locust Malpighian tubules by a diuretic vasopressin-like hormone. Gen. Comp. Endocrinol. 47, 449~457. Proux, J., Proux, B. and Phillips, J. E. (1985). Source and distribution of factors in locust nervous system which stimulates rectal CI transport. Can. J. Zool. Rev. Cart. Zool. 63, 3741. Proux, J. P., Miller, C. A., Li, J. P., Carney, R. L., Girardie, A., Delaage, M. and Schooley, D. A. (1987). Identification of an arginine vasopressin-like diuretic hormone from Lo~usta m~gratoria. Biochem. Biophys. Res. Commun. 149, 180 186. Proux, J. P., Picquot, M., Herault, J. P. and Fournier, B. (1988). Diuretic activity of a newly identified neuropeptide the arginine vasopressin-like insect diuretic hormone: use of an improved bioassay. J. hTsect Ph)siol. 34, 919-927.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

405

Quinlan, M. C. and O'Donnell, M. J. (1998). Anti-diuresis in the blood-feeding insect Rhoehlius prolixus St'~l: antagonistic actions of cAMP and cGMP and the role of organic acid transport. J. bisect Physiol. 44, 561 568. Quinlan, M. C., Tublitz, N. J. and O'Donnell, M, J. (1997). Anti-diuresis in the bloodfeeding insect Rhmhfius prolixus Still: the peptide CAP2~, and cyclic GMP inhibit Malpighian tubule fluid secretion. J. Ext,. Biol. 200, 2363 2367. Raina, A. K. and Gfide, G. (1988). Insect peptide nomenclature, htsect Biochem. 18, 785 787. Ramsay, J. A. (1954). Active transport of water by the Malpighian tubules of the stick insect, Dixippus morosus (Orthoptera, Phasmidae). J. Exp. Biol. 31, 104 113. Ramsay, J. A. (1964). The rectal complex of the mealworm Tenehrio molitor L. (Coleoptera, Tenebrionidae). Phil. Trans. R. Soc. Lond. B248, 279-314. Ramsay, J. A. (1976). The rectal complex in the larvae of Lepidoptera. Phil. Trans. R. Soc. Lond. B274, 203 226. Reagan, J. D. (1994). Expression cloning of an insect diuretic hormone receptor: a member of the calcitonin/secretin receptor family. J. Biol. Chem. 269, 9 12. Reagan, J. D. (1995a). Functional expression of a diuretic hormone receptor in baculovirus-infected insect cells: evidence suggesting that the N-terminal region of diuretic hormone is associated with receptor activation, h?secl Biochem. Mol. Biol. 25, 535 539. Reagan, J. D. (1995b). Molecular cloning of a putative Na + K + 2C1 cotransporter from the Matpighian tubules of the tobacco hornworm, Mctmhtca se.vta, bisect Biochem. Molec. Biol. 25, 875 880. Reagan, J. D. (1996). Molecular cloning and function expression of a diuretic hormone receptor from the house cricket, Acheta domesticus. Insect Biochem. Mol. Biol. 26, 1 6.

Reagan, J. D., Li, J. P., Carney, R. L. and Kramer, S. J. (1993). Characterization of a diuretic hormone receptor from the tobacco hornworm, Manduca sevta. Arch. hi.sect Biochem. Physiol. 23, 135 145. Reynolds, S. (1974). Pharmacological induction of plasticization in the abdominal cuticle of Rhochfius. J. Exp. Biol. 61, 705 718. Reynolds, S. E. and Bellward, K. (1989). Water balance in Man&tea se.\ta caterpillars: water recycling from the rectum. J. Exp. Biol. 141, 33 45. Richardson, N. (1993). Ion transport in the ileum of the desert locust, Schistocerca gregaria F6rskal. M.Sc. University of British Columbia, Vancouver. Ring, M., Meredith, J., Wiens, C., Macins, A., Brock, H. W., Phillips, J. E. and Theilmann, D. A. (1998). Expression of Schistocerca gregaria ion transport peptide (1TP) and its homologue (ITP-L) in a baculovirus/msect cell system. Insect Biochem. Mol. Biol. 28, 51-58. Roberts, V. A., Nachman. R. J., Coast, G. M., Hariharan, M., Chung, J. S., Holman, G. M., Williams, H. and Tainer, J. A. (1997). Consensus chemistry and fi-turn conformation of the active core of the insect kinin neuropeptide family. Chemistry amt Biology 4, 105 117. Romier, C., Bernassau, J. M., Cambillau. C. and Darbon, H. (1993). Solution structure of human corticotropin releasing-factor by H - I - N M R and distance geometry with restrained molecular-dynamics. Protein Eng. 6, 149 156. Rosay, P., Davies, S. A., Yu, Y., Sozen, A., Kaiser, K. and Dow, J. A. T. (1997). Celltype specific calcium signalling in a Drosophila epithelium. J. Cell Sci. !10, 1683 1692. Sawyer, D. B. and Beyenbach, K. W. (1985). Dibutyryl-cAMP increases basolateral sodium conductance of mosquito Malpighian tubules. Am, J. Physiol. 248, R339 R345.

406

G . M . COAST, I. ORCHARD, J. E. PHILLIPS AND D. A. SCHOOLEY

Schoofs, L., Holman, G. M., Proost, P., Van Damme, J., Hayes, T. K. and De Loof, A. (1992). Locustakinin, a novel myotropic peptide from Locusta migratoria, isolation, primary structure and synthesis. Regul. Pepr. 37, 49-57. Schooley, D. A., Miller, C. A. and Proux, J. P. (1987). Isolation of two arginine vasopressin-like factors from ganglia of Locusta mi~,ratoria. Arch. Insect Biochem. Physiol. 5, 157-166. Seinsche, A., Dyker, H., Losel, P., Backhaus, D. and Scherkenbeck, J. (2000). Effect of helicokinins and ACE inhibitors on water balance and development of Heliothis virescens larvae. J. Insect Physiol. 46, 1423 1431. Soyez, D., Van Herp, F., Rossier, J., Lecaer, J. P., Tensen, C. P. and La Font, R. (1994). Evidence for a conformational polymorphism of invertebrate neurohormones: Damino acid residue in crustacean hyperglycemic peptides. J. Biol. Chem. 269, 18295 18298. Spiess, J., Dautzenberg, F. M., Sydow, S., Hauger, R. L., Ruhmann, A., Blank, T. and Radulovic, J. (1998). Molecular properties of the C R F receptor. Tren&' Endocrinol. Metal). 9, 140 145. Spring, J. H. and Phillips, J. E. (1980a). Studies on locust rectum. I. Stimulants of electrogenic ion transport. J. Exp. Biol. 86, 211 223. Spring, J. H. and Phillips, J. E. (1980b). Studies on locust rectum. 11. Identification of specific ion transport processes regulated by corpora cardiaca and cyclic AMP. J. Exp. Biol. 86, 225 236. Spring, J. H. and Phillips, J. E. (1980c). Studies on locust rectum, lII. Stimulation of electrogenic chloride transport by hemolymph. Can, J. Zool. Rev. Can. Zool. 58, 1933-1939. Spring, J. H., Morgan, A. M. and Hazelton, S. R. (1988). A novel target for antidiuretic hormone in insects. Sciem'e 241, 1096 1098. Tamarelle, M., Coast, G. M. and Veenstra, J. A. (2000). Ovary maturing parsin and diuretic hormone are produced by the same neuroendocrine cells in the migratory locust, Locusta migratoria. Peptkles 21, 737 739. Teal, P. E. A. and Nachman, R.J. (1997). Prolonged pheromonotropic activity of pseudopeptide mimics of insect pyrokinin neuropeptides after topical application or injection into a moth. Re eul. Pept. 72, 161 167. Te Brugge, V. A., Miksys, S. M., Coast, G. M., Schooley, D. A. and Orchard, I. (1999). The distribution of a CRF-like diuretic peptide in the blood-feeding bug Rhodnius prolixus. J. Exp. Biol. 202, 2017 2027. Te Brugge, V. A., Niissel, D. R., Coast, G. M., Schooley, D. A. and Orchard, 1. (2001). The distribution of a kinin-like peptide and its co-localization with a CRF-like peptide in the blood-feeding bug, Rhodnius prolixus. Peptides 22, 161 173. Te Brugge, V. A., Schooley, D. A. and Orchard, I. (2002). The biological activity of diuretic factors in Rhodnius prolixus. Peptides 23, 671 681. Terhzaz, S., O'Connell, F. C., Pollock, V. P., Kean, L., Davies, S. A., Veenstra, J. A. and Dow, J. A. T. (1999). Isolation and characterization of a leucokinin-like peptide of Drosophila melanogaster. J. Exp. Biol. 202, 3667 3676. Thompson, K. S. J., Tyrer, N. M., May, S. T. and Bacon, J. P. (1991). The vasopressinlike immunoreactive (VPLI) neurons of the locust, Locusta migratoria. 1. Anatomy. J. Comp. Physiol. A168, 605 617. Thompson, K. S. J., Rayne, R. C., Gibbon, C. R., May, S. T., Patel, M., Coast, G. M. and Bacon, J. P. (1995). Cellular co-localization of diuretic peptides in locusts: a potent control mechanism. Peptides 16, 95 104. Thomson, R. B. and Phillips, J. E. (1992). Electrogenic proton secretion in the hindgut of the desert locust, Schistocerca gregaria. J. Membr. Biol. 125, 133 154.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

407

Thomson, R. B., Thomson, J. M. and Phillips, J. E. (1988). NH~ transport in an acidsecreting insect epithelium. Am. J. Physiol. 254, R348 R356. Trimmer, B. A. (1985a). The inactivation of exogenous serotonin in the blowfly, Calliphora. Insect Bioehem. 15, 435 442. Trimmer, B. A. (1985b). Serotonin and the control of salivation in the blowfly Calliphora. J. Exp. Biol. 114, 307 328. Tublitz, N. J. (1989). Insect cardioactive peptides: neurohormonal regulation of cardiac activity by two cardioacceleratory peptides (CAPs) during flight in the tobacco hawkmoth, Manduca sexla, d. Exp. Biol. 142. 31 48. Tublilz, N. J. and Evans, P. D, (1986). Insect cardioactive peptides: cardioacceleratory peptide (CAP) activity is blocked in vivo and i, vitro with a monoclonal antibody..I. Neurosci. 6, 245t 2456. Tublitz, N. J. and Truman, J. W. (1985). Insect cardioactive peptides. 1. Distribution and molecular characteristics of two cardioacceleratory peptides in the tobacco hawkmoth, Manduea se.vta, d. Exp. Biol. 114, 365-379. Tublitz, N. J., Brink, D., Broadie. K. S., Loi, P. K. and Sylwester, A. W. (1991). From behaviour to molecules: an integrated approach to the study of neuropeptides. Trends Neurosci. 14, 254-259. Tublitz, N. J., Allen, A. T., Cheung, C. C., Edwards, K. K., Kimble, D. P., Loi, P. K. and Sylwester, A. W. (1992). Insect cardioactive peptides: regulation of hindgut activity by cardioacceleratory peptide-2 (CAP2) during wandering behaviour in Mclnduea se,vta larvae. J. E,vp. Biol. 165, 241 264. Tyrer, N. M., Davis, N. T., Arbas, E. A., Thompson, K. S. J. and Bacon, J. P. (1993). Morphology of the vasopressin-like immunoreactive (VPLI) neurons in many species of grasshopper. J. Comp. Neurol. 329, 385-401. Vanden Broeck, J. (2001). Neuropeptides and their precursors in the fruitfly, Drosophihl melanogaster. Peptides 22, 241 254. Van Kerkhove, E., Weltens, R., Roinel, N. and Dedecker, N. (1989). Haemolymph composition in Formica (Hymenoptera) and urine formation by the short isolated Malpighian tubules: electrochemical gradients for ion transport. J. h~sect Physiol. 35. 991 1003. Veenstra. J. A. (1988). Effects of 5-hydroxytryptamine on the Malpighian tubules of Aedes aeg~7~ti. J. hTsect Physiol. 34, 299 304. Veenstra, J. A. (1994). Isolation and identification of three leucokinins from the mosquito Aedes aegypti. Biochem. Biophys. Res. Commun. 202, 715 719. Veenstra, J. A. and Hagedorn, H. H. (1991). Identification of neuroendocrine cells producing a diuretic hormone in the tobacco hornworm moth, Mamhtca sexta. Cell Tissue Res. 266, 359 364. Veenstra, J. A., Lau, G. W., Agricola, H. J. and Petzel, D. H. (1995). lmmunohistological localization of regulatory peptides in the midgut of the female mosquito Aedes aegypti. Histochem. Cell Biol. 104, 337 347. Veenstra, J. A., Pattillo, J. M. and Petzel, D. H. (1997). A single cDNA encodes all three Aedes leucokinins, which stimulate both fluid secretion by the Malpighian tubules and hindgut contractions. J. Biol. Chem. 272, 10402 10407. Wang, S., Rubenfeld, A. B., Hayes, T. K. and Beyenbach, K. W. (1996). Leucokinin increases paracellular permeability in insect Malpighian tubules. J. Exp. Biol. 199, 2537 2542. Wang, Y., Zhao, Y., Meredith, J., Phillips, J., Theihnann, D. and Brock, H. (2000). Mutational analysis of the C-terminus in Ion Transport Peptide (ITP) expressed in Drosophila Kcl cells. Arch, Insect Biochem. Physiol. 45, 129 138.

408

G . M . COAST, I. ORCHARD, J. E. PHILLIPS A N D D. A. SCHOOLEY

Wegener, C., Predel, R. and Eckert, M. (1999). Quantification of periviscerokinin-I in the nervous system of the American cockroach, Periplaneta americana: an insect neuropeptide with unusual distribution. Arch. htsect Biochem. Physiol. 40, 203 211. Wegener, C., Linde, D. and Eckert, M. (2001). Periviscerokinins in cockroaches: release, localization, and taxon-specific action on the hyperneural muscle. Gen. Comp. Endocrinol. 121, 1 12. Wegener, C., Herbert, Z., Eckert, M. and Predel, R. (2002). The periviscerokinin (PVK) peptide family in insects: evidence for the inclusion of CAP2b as a PVK family member. Peptides 23, 605 611. Weis-Fogh, T. (1967). Respiration and tracheal ventilation in locusts and other flying insects. J. Exp. Biol. 4% 561 587. Weltens, R., Leyssens, A., Zhang, S. L., Lohrmann, E., Steels, P. and van Kerkhove, E. (1992). Unmasking an apical electrogenic H ~ pump in isolated Malpighian tubules (Formica polyctena) by the use of barium. Cell. Physiol. Biochem. 2, 101-116. Wheelock, G. D., Petzel, D. H., Gillett, J. D., Beyenbach, K. W. and Hagedorn, H. H. (1988). Evidence for hormonal control of diuresis after a blood meal in the mosquito Aedes aegypti. Arch. hlsect Bio~hem. Physiol. 7, 75 89. Wieczorek, H., Putzenlechner, M., Zeiske, W. and Klein, U. (199l). A vacuolar-type proton pump energizes K+/H + antiport in an animal plasma membrane. J. Biol. Chem. 266, 15 340 15 347. Wieczorek, H., Gruber, G., Harvey, W. R., Huss, M., Merzendorfer, H. and Zeiske, W. (2000). Structure and regulation of insect plasma membrane H + V-ATPase. J. Exp. Biol. 203, 127 135. Wiehart, U. 1. M., Nicolson, S. W., Eigenheer, R. A. and Schooley, D. A. (2002). Antagonistic control of fluid secretion by the Malpighian tubules of TenebHo molitor: effects of diuretic and antidiuretic peptides and their second messengers. J. Exp. Biol. 205, 493 501. Williams, D., Phillips, J. E., Prince, W. T. and Meredith, J. (1977). The source of shortcircuit current across locust rectum. J. Exp. Biol. 77, 107 122. Williams, J. C. J. and Beyenbach, K. W. (1983). Differential effects of secretagogues on Na and K secretion in the Malpighian tubules of Aedes aegypti (L.). J. Comp. Physiol. B149, 511 517. Williams, J. C. J., Hagedorn, H. H. and Beyenbach, K. W. (1983). Dynamic changes in flow rate and composition of urine during the post-bloodmeal diuresis in Aedes aegypti (L.). J. Comp. Physiol. B153, 257-265. Winther, A. M. E., Lundquist, C. T. and Nfissel, D. R, (1996). Multiple members of the leucokinin neuropeptide family are present in cerebral and abdominal neurohemal organs in the cockroach Leucophaea maderae. J. Neuroendocrinol. 8. 785-792. Yu, M. J. and Beyenbach, K. W. (2000). Leucokinin-VlII increases epithelial CI shunt conductance via a receptor-mediated pathway involving calcium. FASEB J. 14, A579. Yu, M. J. and Beyenbach, K. W. (2001a). Intraceltular Ca 2+ mediates the leucokininVIII induced increase in paracellular C1- conductance of Malpighian tubules. FASEB J. 15, A139. Yu, M. J. and Beyenbach, K. W. (2001b). Leucokinin and the modulation of the shunt pathway in Malpighian tubules..I, hJsect Physiol. 47, 263-276. Yu, M. J. and Beyenbach, K. W. (2001c). Leucokinin-Vlll induces paracellular CIconductance in Malpighian tubules of the yellow fever mosquito, Aedes aegypti. FASEB J. 15, AI40.

INSECT DIURETIC AND ANTIDIURETIC HORMONES

409

Zhao, Y. (2000). The structure activity relationship of the N-terminal domain in desert locust ion transport peptide (ITP). M.Sc., University of British Columbia, Vancouver. Zitnan, D., Kingan, T. G. and Beckage, N. E. (1995). Parasitism-induced accumulation of FMRFamide-like peptides in the gut innervation and endocrine cells of Mamhtca se.wa. Insect Biochem. Mol. Biol. 25, 669 678.

ADDENDUM The genome sequence of the malarial mosquito, Anopheles gambiae, was published in October 2002 (Holt et al., 2002). Mosquito homologues of the neuropeptides and receptors described in this review are readily identifiable (Riehle et al., 2002; Hill et al., 2002) although, interestingly, to obtain the full sequence of the CRF-like DH another intron/exon excision is needed (J. Vanden Broeck, personal communication) as previously found for Drome-DH (see section 4.2.2,2).

References Hill, C. A., Fox, A. N., Pitts, R. J., Kent, L. B., Tan, P. L., Chrystal, M. A., Cravchik A., Collins, F. H., Robertson, H. M. and Zwiebel, L. J. (2002). G protein-coupled receptors in Anopheles gambiae. Science 298, 176 178. Holt, R. A., et al. (2002). The genome sequence of the malaria mosquito Am)pheles gambiae. Science 298, 129 149. Riehle, M. A., Garczynski, S. F., Crim, J. W., Hill, C. A. and Brown, M. R. (2002). Neuropeptides and peptide hormones in Anopheles gambiae. Science 298, 172 175.

Index Ah('Ls'(m 30 Acanthodis curvidens 239 Acanthogry//us Jbrtipes 248 A('anthoplus speiseri 25 I acetylcholine receptors (AchRs) 114

dcherontia styx dopamine in 99 octopamine in 106

Acheta domesticus Achdo-Dp in 297 CRF-like diuretic hormones in 304, 329 diuretic hormone in 302 dimeric peptide in 293 haemolymph in 379 kinins in 305, 331,361,370 MNCs and LNCs in 358, 364 NSCs in 359 sound signalling in 168, 200, 217, 219. 221,223, 225, 251 synergism between diuretic hormones 380 transport in Malpighian tubules 285 acridid ear 176 80 adipokinetic/red pigment concentrating hormone (AKH) family 293 Aeries aeg;vpti 285, 373 CRF-related peptide in 330, 357 diuretic/myotropic kinin neuropeptides in 305, 33l, 333. 335 haemolyph in 380 kinins in 360, 361,364 MNCs and LNCs in 358, 364 mosquito natriuretic peptide (MNP) in 311 NSCs in 359 post-eclosion diuresis in 290 serotonin in 324, 325, 351 A~amtcris iiTsectivora 239 Agrotis ,s~?gellt171359

A I/omenobius socius 214 A//onemohius ./ilsciatus 214

Ambl)'cor37~ha parvipenni,s" 219, 248 Amelrus 229 1 -aminocyclobutane-trans- 1,3dicarboxylate 73 v-aminobutyric acid see GABA Amphiacusta m(O'a 246 Anahaella 26 ,4nahrus simplex 176 Ancistrura ni~woviltata 167, 201 2, 203, 215 Anopheles gambiae 11, 12, 281 Anostostoma australasiae 156 ANP binding 6 Antheraea pernyi 108 Antheraea polyphemus 27, 40 antidiuretic factors (ADFs) 284 antidiuretic hormones see diuretic and antidiuretic hormones Arltrozous p, pullidus 225

Amtrogryllus 156 Amtrogwllus arboreus 162 Apis melfilera diuretic and antidiuretic hormones and 301, 305,358, 359, 360 dopamine in 98 EAAT (apmEAAT) 64, 67, 70 arginine vasopressin-like insect diuretic hormone (AVP-IDH) 294 6, 326, 351 2, 368

Armadillicfium vulgare 317 aromatic amino acid decarboxylase (AADC) 59 Astacus aslacus, serotonin in 92 atrial natriuretic peptide 5 atropine 121 auditory interneurons 194 206 ascending 201 3 in grasshoppers 205 6 in the mole cricket 203 5 omega neuron 197 201 T-cell 194 7

412

auditory receptor organs in the tibia 182 5

Balanus *mbilis 122 Balboa libialis 239 Barbitistes 212 Barbitistes serricauda 225 Berkeley Drosophila Genome Project 116, 307 Blaberus 96 dopamine in 101, 104 octopamine in 109 Blaberus discoMalis 27 BLAST analysis of Drosophila genome 3, 8, 15, 23, 30, 293, 303 Bombyx, protein kinases in 27 Bomhyx mori 321, 322, 383 guanylyl cyclase in 2, 44 Boophilus microplus 331 Bullacris memhracioMes 160, 181, 211 ce-bungatoxin 116 C. morostts 342, 344 C. salinarius 302, 330 C. vicina 375 Caenorhahditis eh',gans 126 choline transporters 116 dopamine in 102 MsGC-fi3 21 MsGC-I 18 octopamine in 107 receptor GCs in 3, 34 calcitonin-like peptides 304, 336, 363 (2dliphora ervthrocephala 92, 349 Calliphora vomitoria 358, 359, 360 calyculin A 39 Cancer borealis 88 Carcinus maenas 70, 92, 317 cardioacceleratory peptide 2b (CAP2b) 42 3,307 8, 336, 346-7, 362 3, 37l Ceratitis capitata, protein kinases in 27 cGMP-dependent protein kinase (PKG) 1 chloride transport stimulating hormone (CTSH) 314--15 Choeroparm~ps 239 choline acetyltransferase (CHAT) 58 choline transporters 114~21 background 114~15 distribution 119 kinetics and pharmacology 119 21 regulation 121 structure 116 19

INDEX

Chomh'oderella borneenses 240 Chortkippus 214 Chorthippus h~uttulus 165, 179, 180, 187, 188, 189, 191, 192, 193,215 Chorthippus mollis 164 choruses 247 51 alternating 248 synchronous 247 unison bout singing 247 unison singing 247 Ckymomyza costata, dopamine in 101 Ciona intesthutlis, MsGC-I 18 cis-3-aminocyclohexanecarboxylic acid 87 (2S,3S,4R)-cis-(carboxycyclopropyl)glycine (CCG III) 73 "clockwork cricket" 161 Owmidophyllum e.vimium 228 cocaine 109, 110 Conocephahls 246 Conocephalus brevipemfis 214 Conocwhalus conocepkalus 235 Comwephalus maculatus 235 Conocephalus n&ropleurum 214, 218, 245 Copiphora 239 Copiphora hrevirostris 238, 239 corticotropin releasing-factor see CRF CRF-related diuretic peptides 293, 327 mode of action 329 1 receptors 326 CRF-related neuropeptides 296 304 isolation and purification 296 302 structures of CRF-related DH 3 0 3 4 CRF-related peptides 344~5, 352 7 circulating levels 368 70 degradation and inactivation 375 6 crustacean cardioactive peptide (CCAP) 21, 37, 293 crustacean hyperglycemic hormone (CHH) 8 cyclic GMP l 44 ecdysis 37 41 f o o d - s e a r c h b e h a v i o u r 41 2 function 26 43 Malpighian tubule regulation 42 3 molecular targets 26-32 cyclic nucleotide-gated channels 30 2 protein kinases and substrates 27 30 neuronal development 34-7 physiological function 32~43 regulation 2 26 sensory physiology 33~4

INDEX

cyclic nucleotide-gated channels 30 2 cyclic nucleotide-gated ion channel protein (cng) 30 cny#like (cn~/) 30 CvcloptiloMes canariensis 160, 161 Cvphoderris 156 Qwhoderris monstrosa 181, 186, 243, 244, 246 CyphoderrLs strepitans 221. 244

D-cysteate 73

Dectk'us verrucivorus 183 Deinacrida 157, 228 Deinacrida rugosa 229 desipramine 104, 105 Dictostelium, atypical guanylyl cyclases 15 Diploptera punctata 293, 294, 301, 380 EAAT (dipEAATI) 64 diuretic and antidiuretic hormones 279 388 cellular actions 324- 47 co-localisation 363-6 distribution 348 66 diuretic/myotropic kinin neuropeptides 331 5 calcitonin-like peptides 336 CAP?b/PVK-2 336 mode of action 333 5 partially characterised factors acting on Malpighian tubules 337 8 receptors 331 3 Tenehrio ADFoe (Tenmo-ADFoe) 337 fluid uptake from the cryptonephric complex 341 2 functions 289 91 clearance of toxic wastes 291 excretion of excess metabolic water 290 post-eclosion diuresis 290 postprandial diuresis 289 90 restricting metabolite loss 291 integrated activities 378 83 co-ordinating Malpighian tubule/ hindgut activities 382 3 excretory system as target for pest control strategies 383 5 future directions 385 7 haemolymph volume/composition maintenance 378 80 synergism between diuretic hormones 38/t 2

413

isolation/structural characterisation of active factors 291 324 in neurosecretory cells and neurohaemal structures 348-66 physiological relevance 366 77 purification and chemical structure of neuropeptides that act on Malpighian tubules 293 5 that stimulate locust hindgut 312 24 regulation of hindgut activity 338 40 of Malpighian tubule secretion 324 31 structure/activity studies 342 7 diuretic/myotropic kinin neuropeptides 305 7, 331 5 DocMocercus 239 DocMocercus gigliotosi 239 DOPA decarboxylase (DDC) 58 dopamine (DA) 91, 110 dopamine transporter (DAT) 58, 99 106 background 99 102 distribution 103 functional domains 102 3 glycosylation sites 103 kinetics and pharmacology 103 5 phosphorylation sites 102 regulation 105 6 st.ucture 102 3 Drepanoxiphus an~ustekmfinaltts 239

Drosophih* aspartate transporter (DrmEAAT2) 68, 69, 77, 78 atypical GCs in 22 Blotgene 112, 113 CG17922 gene 311-1 CG3536 gene 30 1 cn~ 30, 33 DA-ergic neurons in 101 DAT in 103 dopamine in 101, 104 drmDAT 96. 105 dunce (~hlc) 23 ea~ family 31 2 eclosion in 39, 4(/ enahh, d (enh) gene 30 excitatory glutamate in 62 /braging (lbr) gene 41, 42 GABA transporters 80, 86 glial cells in 88 glutamate receptors in 61

414

INDEX

Drosophila (contimwd)

EGPs 29

glutamine cycle 76 histamine in 122 hyperpolarization activated (//h) channels 31, 33 hwbriated (ine) gene 82, 83, 112, 113 hw transporter 112 Malpighian tubule regulation 42 MsGC-I 18, 22 neurotransmitter transporters in 60, 61 noq)A gene 33 octopamine in 106, 109 orphan transporters in 112 phosphodiesterases in 23, 24, 26 photoreceptors 34, 35 protein kinases in 27 PKG in 27, 29, 36 receptor guanylyl cyclases in 3 11 roho (axon guidance receptor) 30 Rosa gene product 112 rosA mutant 113 serotonin in 92 serotonin transporter (drmSERT) 93, 96 soluble guanylyl cyclases in 11-15 Drosophila Genoma Project 2

Elephantodeta nobilis 249

Drosophila mehmogaster calcitonin-like peptides 304 CAP?~ in 308, 336, 346, 361, 362 CRF-related diuretic peptides in 327, 330 diuretic/myotropic kinin neuropeptides in 331,333, 335 dopamine in 102 Drome-DH31 291,387 EAAT (drmEEATs 1 and 2) 64, 67, 70 GABA transporters 79 genome 126, 281,296, 303,386 ITP sequence 321,322, 323 kinins in 357 358, 359, 360, 377 Manse-DH in 383 octopamine in 110 Peram-pyrokinin 5 in 365 serotonin in 92, 325 songs in 222 V-ATPase in 329 Drosophiht punctata 301, 302, 336, 363 D-threo-3-hydroxyaspartate 73 ecdysis 37 41 ecdysis-triggering hormone (ETH) 37 eclosium hormone (EH) 37

Ena/VASP-like protein (EVL) 30 Eneoptera gto,anensis 227 Ephippiger 182, 223 Ephipp~zer ephippiger 155, 165, 167, 169, 222 El)hipl?igeri&~ taeniata 67 Eunemobius carol#ms 235, 245-6

Euphasioptet3,x ochracea (Ormia ochracea) 224, 229, 230, 231,233,241,252 excitatory amino acid transporters 61 129 applications to insect control 125-9 future directions 127 8 neurotransmitter transporters as new targets for 126 7 postgenomic prospects for research 128 9 relevance of insect neurophysiology to 125 Na+-dependent transporters II 121 3 Na-/C1-dependent GABA and monoamine transporters I 78 114 Na +/Cl--dependent transporters II 114 21 N a t / K +-dependent aspartate transporter 77 8 Na + K • -dependent glutamate transporters 61 77 putamine neurotransmitter transporters 123 5 excretion, physiology of 282-8 fluid reabsorption across the cryptonephric complex 288 food-search behaviour 41 2 introduction 282-4 transport processes in hindgut 285 8 in Malpighian tubules 284-5

Formica polyctena 304, 337 Formica polyctena antidiuretic factor (FopADF) 311 12

Formica ru/~l, serotonin in 92 frequenin 10, 18 frog epinephrine transporter (lET) 111

G protein-coupled receptor kinases 10 G. bimaculatus 168 GABA 78

INDEX

GABA transporters (GATs) 78, 79 91 background 79 80 co-localization of EAAT and GAT in glial cells 88 91 distribution 86 EFWER sequence in EL2 83 functional domains 81 6 heptan leucine zipper motif 82 3 ion-permeation site 82 kinetics and pharmacology 86 7 N-linked glycosylation sites 83 6 PKA and PKC phosphorylation sites 83 regulation 87 8 structure 80 6 substrate binding site(s) 82 Gampsoch'is huergeri 194 Gampsocleis graliosa 163, 182, 184, 185, 189 GBR12909 104 GC-activating proteins (GCAPs) 10 11 glial transporter protein 1 (GET-l) 69 gliapse 88 GLUT 62 glutamate decarboxylase (GAD) 59, 79 glycine transporters 124 Grompha~h;rhina portentosa 62 gryllid ear 180 1 Gryllodes supplicans 156 Gry//otall)a /wxadaclyla 235 Grvllotalpa major 156, 167, 248 Gryllomlpa vineae 160 Grvlhts 247 Gryllus himaculatus 101, 162, 164, 189, 197, 210, 212, 219. 220, 252, 348, 359 Grvlhts campestris 159, 161, 180, 197, 200, 201,222, 252 Gryllus,firmus 168, 212, 230 Gryllus./idtoni 167, 230

Grvllus in:eger see Grvllus texensis Grvllus liHeaticeps 168, 169, 221, 223, 241 Grvlhcs pem~sylvanicus 166 Gryllus ruhetts 167, 229, 230, 231,232, 251 Grvllus te.vensis 220, 224, 227, 229, 230, 231, 232, 233, 241,242, 251, 252 guanosine 3'5' cyclic monophosphate see cyclic GMP guanylyl cyclases 2 22 atypical 15 19 biochemical properties 11 15 ligands and activators 8 11 receptor 3 I1

415

sensory receptor 5 sequence analysis 3 8, 11 15 soluble 11 15

lta&'ogryllacrLv 229 Huematobia irrilans 307 Haenschie/hl ecuadorica 176, 239 haglid ear 181 hearing organs, structure of 1711 81 acridid ear 176 80 age, changes with 181 gryllid ear 180 1 haglid ear 181 tettigoniid ear 171 6 Helicoveq)a zea 305, 377, 384 tleliothis virexcens 26. 31 2, 33, 366, 372. 377, 384 Hemiamh'us 228, 229 hemicholinium-3 116. 120 Hemideina 228 Hemideimt crassidens 156 7 Hemisaga denticulata 190 high-affinity glutamate transporters (EAATs) 59 Hirur& 98 histamine (HA) 9 l histamine transporter (HAT) 58, 121 3 background 121 2 distribution 123 kinetics and pharmacology 123 molecular biology 123 histidine decarboxylase (HDC) 58

Homarus americam~s 317 llomorocor37~hus 235 Homotrixa alleHi 166, 230, 232, 233 human nor-epinephrine transporter (hNET) 93 ttyalophora cecr(qfia 23, 27 6-hydroxytryptamine .s'ee serotonin Hvles litleata 294, 301 imipramine 104, 105 inositol trisphosphate receptor (1P3R) associated PKG substrate (1RAG) 29 hlsclra covilleae 235 hlsara elegcols 235 ion transport peptide (ITP) 315 24, 347. 363 amino acid sequence 316 18 expression of 319 2(/ ITP-like (ITP-L) cDNA in locusts 318

416

INDEX

ion transport peptide (ITP) (continued) purification, partial sequencing and actions 315 16 sequence evolution among insects 3 2 0 ~ synthetic 318 Ischnomela pulchripennis 238 juxtamembrane hinge 6

Kawanaphila mir& 175 Kawanaphila nartee 175-6, 194, 216, 219, 246

Kawanaphila yarraga 175 kinase-like domain 6 kinins 345 6 circulating levels 370 l in neurosecretory cells and neurohaemal structures 357 61 Laupala 214, 215 Laupahl cerasina 214 L-cysteate 73

Leptinotarsa decemlineata 308 10, 348 Leptophyes punctatisshna 155, 220, 223, 244, 248

Lerneca jilscipennis 227 8 Leucophaea, excitatory glutamate in 62 Leucophaea maderae 305, 308, 358, 359, 360, 361, 364, 370 L-glutamate 62 Libanasidus vittatus 157 L~gurotettix coquilletti 218, 244, 250 Ligurotettix phmum 217, 246, 250 Limulus 122 Limulus polyphemus 117

Locusta mi,~,raloria 8 arginine vasopressin-like DH in 295 AVP-like immunoreactive neurons in 351, 368 calcitonin-like peptides 304, 336 CAP2b in 308 choline transporters 116, 121 co-localisation 364, 365, 366 CRF-related diuretic hormone 301 CRF-related diuretic peptide 352, 355, 357 diuretic/myotropic kinin neuropeptides in 331 dopamine in 99, 101 GABA transporters 79 haemolymph in 374 histamine in 123

ion transport peptide 363 kinins 358, 359 Locmi-DH in 297, 384 neuroparsins 312, 313, 314 octopamine in 110 serotonin in 92, 93, 324, 348, 349 sound signalling 225 taurine in 124 L-threo-3-hydroxyaspartate 73

L-trans-pyrroliginre-2,4-dicarboxylate (LPDC) 73

Lymnaea sta~nalis 306, 331 Manduca 96 atypical guanylyl cyclases 15, 17, 18, 19 20, 21 CNS 23 ecdysis 37, 40 EGPs in 29 MsGC-I 17, 18 neuronal development 34~5 orphan transporters in 113 protein kinases in 27 receptor GCs 3 soluble guanylyl cyclases in 11 15 VNCs 38 Mamhlca sexta 70, 372, 386 antidiuretic factors 310 11 CAP~b in 361, 362 cardioacceleratory peptides (CAPs) in 307, 371 co-localisation in 364, 365 CRF-like DH receptors 304 CRF-related peptides 352, 355, 376 cyclic AMP production 345 diuretic/myotropic kinin neuropeptides in 305 fluid uptake from the cryptonephric complex 341 GABA transporters 79, 80, 86 glutamate uptake 62 guanylyl cyclase in 2 histamine in 121, 123 kinins in 358, 359 Manse-DH 295, 296, 300--1 Manse-DPll in 366 octopamine in 106, 108, 109 orphan transporters in 113 receptor GCs in 4 serotonin in 92, 348

Mehmoplus sanguinipes 217 Metaballus litus 190

INDEX

Micronycteris" hirsuta 234, 238 Miogryllus 156 mosquito natriuretic peptide ( M N P ) 311 Motttweta isolala 228 MsGC-fi3 19 22 MsGC-1 17-19 MsGC-II 3 MULTICOIL program 12, 22

Musca GABA transporters 80 histamine in 122 Musca domeslica 77, 297, 299, 336, 346, 352. 355, 358, 359, 360, 364, 377 muscarinic ACH receptors 115-16 Mygalopsis rearM 182, 185. 244 Mygal~q~sis paul~erculus 190 Myopophyllum speciosum 166, 238, 239 Myoti~ mvotis 237 Myrme/eotettix macu/atus 222 Na ' -dependent transporters !I 121 3 Na-/C1--dependent GABA and monoamine transporters 1 78 114 Na ~/C1--dependent transporters il 115 21 Na - K ~ -dependent glutamate transporters 61 77 chloride channel domain 69 dihydrokainate (DHK) binding site 69 distribution 69 72 functional domains 67 9 glutamine cycle 76-7 histidine "326" 69 kinetics and pharmacology 72 4 N-linked glycosylation sites 69 Na ~ binding sites 68 in permeation site 67 8 PKA and PKC phosphorylation sites 68 regulation 74 5 structure 63 9 substrate selectivity domains 67 zinc-binding site 68 Na +-dependent aspartate transporter 77 8 natriuretic peptide clearance receptor (NPR-C) 7 Nauphoeta cinerea 358, 359, 364 Neobellieria bullata 307 Neoconocephalus 169, 245 Neoconocephalus caudellianus 247 Neoconoc~7~halus ensiger 195 6, 207, 229, 235. 236

417

Neoconocephalus Neoconocephalus Neoconocephalus Neoconocephalus

exiliscanorus 247 nebrascensis 247, 249 rohusms 169 spiza 217, 250

neuroealcin 10 neurolls

714 205 6 molecular chemistry 56 9 neurotransmitter uptake and vesicular storage 59 61 neuroparsins 312 14 neurotransmitter receptors 56 neurotransmitter transporters (NTTs) 56, 60 nicotinic receptors (nACHRs) 114 nipecotic acid 87 nisoxetine 104, 105, 109 nitric oxide (NO) 11

nitric-oxide (NO)-insensitive soluble guanylyl cylases (GCs) 2 nitric oxide synthases 1I Nyclophilus geq~/i'o3i 240 Nyctophilus mq/or 240

octopamine (OA) 78, 91, I10 octopamine transporters 106 11 distribution 208 kinetics and pharmacology 108 9 regulation 109 10 structure 10Z 8 tyramine transport 110 11 O D Q (1 lt-[1,2,4]oxadiazolol[4,3-a]

quinoxalin- l-one) 14

Oecanthus 214 Oecanthus celerinictus 162 Oecanthusjidtoni 247, 249, 250 Oecanthus n~,,ricornis 219 Oecanthus quadriptmctatus 162 okadaic acid 42 omega neuron 197 201 in acridids 201 1 (ON1) 197-200 2 (ON2) 200 1 Omocesttts viri~hdus 164, 165, 245 Oncopeltu.s' jasciatus 365 Onymacris p[ana 291 Orchelhmcm 167 Orchelinnm7 gladiator 248 Orche/imum n~wipes 248, 251, 252 Orche/imum vulgare 246, 248 Orconectes ]imosus 363

418

Ormia ochracea ( Euphasiopteryx ochracea) 224, 229, 230, 231,233,241. 252 Orocharis luteolira 230 orphan transporters 78-9, 111 14 background 111 distribution 113 kinetics and pharmacology 113 14 structure 112 13 Otus stops 225 1 H-[1,2,4]oxadiazolo[4,5-a]quinoxalin- 1one (ODQ) 20 3-N-oxalyl- L-2,3-diaminopropionate 74 oxotremorine 121 Pachnoda simuata 322 Paramecium, atypical guanylyl cyclases in 15 Parascopioricus exarmatus 228 Penaeus vannamei 306 Periplaneta americana choline transporters 115 co-localisation in 365 CRF-like diuretic hormone in 302 dopamine in 99, 101 2 GABA transporters 79, 80 glutamate uptake 62 Manse-CAP2wlike 308 NSCs in 359 PerampDP in 297 serotonin in 92, 348 taurine in 124 periviscerokinins see cardioacceleratory peptide 2b Phalacrocera replicata 359, 361 Pllaneroptera./~dcata 237 Phaneroptera nana 219, 248 Pholidoplera griseoaptera 185, 186, 225, 227, 248, 250 Phormia terraenovae 358. 359, 360 phosphodiesterases (PDEs) 22-6 Phv//omimus inversus 240 Pieriv hrassicae 290 Pieris' rapae 296. 345. 372 piperidenecarboxylic acid 88 Plagiostira all)onotata 235 Phtsmodium, atypical guanylyl cyclases in 15 Plal3"cleis. a/l)opunctata 225 Phttvcleis intermedia 247, 251 Platvslolus ohvius 155 Poecilimon 174, 254 Poecilimon a[finis" 215, 232

INDEX

Poecilhnon artedentatus 232 Poecilhnon mariannae 224. 232 Poecilimon nobilis 232 Poecilimon ornatus 216, 223 Poeci/hmm proprinquus 232 Poecilimon schmidti 165 Poecilh~wn thessalicus 241 Poecilimon veluchianus 232 Polysarchus denticauda 171, 172, 173, 174, 182, 183 proctolin 293 Promeca perakana 240 Promeca sumatrana 240 PROSITE analysis 26 Protein Family (Pfam) databases 26 protein kinases and substrates 27 30 Psammodronlus a]?irus 224 Psorodonotus il[yricus 182 Pterophylla beltrani 228 Pterophylla camelE/blia 228, 248, 250 putamine neurotransmitter transporters 123 5 Ramsay assay 282 rat dopamine (rDAT) 93 rat serotonin transporters (rSERT) 93 recoverin I0 Requena verticalis 168, 169, 174, 216, 219, 221 Rhodniusprolirus 125. 371 2, 373, 386 AVP-like immunoreactive neurons in 351 calcitonin-like peptides 304, 363 cardioacceleratory peptide 2b (CAP2b) in 43, 336 co-localisation 364, 365 CRF-related neuropeptides 302. 352, 355, 357, 376 diuresis in 282 dimetic/myotropic kinin neuropeptides in 305. 335 haemolymph in 374, 375, 379. 380 kinins in 358 Manse-CAP2b in 337 Malpighian tubule transport 285 metabolite loss, restricting 291 NSCs in 359, 360 serolonin in 91. 324, 325, 342, 344. 350, 366, 367-8 synergism between diuretic hormones 380, 381 rhodopsin kinase 10

INDEX

419

Ruspolia nitffUk~ 163 4 Ruspolia 235 Ru,v~olia df[lbrens 171, 182, 183, 184 S100B 10

Saccharomyces cerevMae 345, 386 Scapteriscus abhreviatus 203,205 Sccq~teriscus acletus 158 Sc~q)teriscus borellii 203 5, 230, 237 SCal)leriscu.s' didact),lis 235 Scapteriscus vicim~s 158 Schedocentrus 239 Schistocerca americana 301,358,359, 361. 362

Sch&tocerca gregaria (SgITP) 8 antidiuretic factors in 310 arborisation in 351 directional hearing in 187, 188 dopamine in 99 excretion in 284 GABA transporters 79.80 guanylyl cyclases in 22 hindgut activity 338 histamine in 122 1TP sequencing in 323, 363 neuroparsins in 313, 314 postprandial diuresis in 290 serotonin in 92 Sciara,s'aga quadrata 166, 190, 211,230. 231. 232, 233,234 Scopiorinus jra~ilus 228

Scudderia curvicauda 219 "selfish herd' effect 158 serotonin (5-hydroxytryptamine: 5-HT) 59, 78, 91, 110 circulating levels 366 8 degradation and inactivation 374 5 mode of action 325 6 in neurosecretory cells and neurohaemal structures 348 51 receptors 324-5 secretion by Malpighian tubule 324 6, 342 4 serotonin transporter (SERT) 59, 91 9 background 92 3 cocaine binding site 96 distribution 97 functional domains 94 6 heptan leucine zipper 94 ion permeation site 94 kinetics and pharmacology 98 9

monoamine-binding site(s) 94 regulation 99 structure 93 6 tricyclic antidepressant interaction site 94 6 SKF-89976A 87, 88 sound signalling in Orthoptera 151 254 analysis 189 207 auditory intemeurons 194 206 ascending 201 3 in grasshoppers 205 6 in the mole cricket 203 5 omega neuron 197 201 T-cell 194 7 auditory receptor organs in the tibia 182 5 components 209 cooperation/competition between males 243 52 choruses 247 51 satellite males and silent searching 251 2 spacing, aggregating and fighting 244 6 defences against acoustically orienting predators 226 9 against bats 234 40 against parasitoids 229 34 directional hearing 187 9 environment, effects on 209 I1 hearing and ears 169 89 heterospecific sounds 224 43 information content 207 24 mate choice 217 24 mate location 215 17 mating systems 154 9 patterns in calling 157 9 variation in 154 7 new directions 252 3 predator avoidance mechanisms, evolution of 242 3 primary afferents in acridids 186 7 in the prothoracic ganglion 186 sex recognition 215 sexual vs natural selection 241 2 songs and signals 159 69 analysis 190 3 changes with age 165 6 energetic costs of calling 168 9 intensity, distance and size 160 I

420

sound signalling in Orthoptera (continued) songs and signals (continued) mechanisms of sound production 161- 3 pattern generation in crickets 163 5 sex differences 167 temperature effects 167 8 vibratory communication 166 7 species recognition 212 15 structure of hearing organs t70-81 acridid ear 176-80 age, changes 181 gryllid ear 180-1 haglid ear 18 I tettigoniid ear 171 6 symmetry and asymmetry 206 7 tonotopic organization of receptor projections 186 7 of sense cells 183 5 Spodoptera 32 Spodoptera littoralis 106, 108 Spodoptera littorina 358 Steirodo, careovirgulatum 228 Stenobothrus lineatus 247 Steropleurus nohrei 155 Steuropleurux stali 155, 224 Stomoxys calcin'ans 297, 299, 307 swiss cheese (sws) gene 26 7 synaptotagmin 35 Syrbula admirubilis 247 Syrbula./itscovittata 247 taurine transporters 124 5 Teleogt3'lhc~ commodus 207, 213 Te/eogryllus oceanicus 185, 199, 200 -1, 203, 213, 218, 223, 231, 233, 236, 239, 240. 251, 252, 335 Tenebrio ADFc~ (Tenmo-ADF~) 337 Tenebrio molitor 63, 288, 294, 299-300. 304, 308 10, 337, 341 Tetrahymena 15 Tettigonia cantans 166, 183, 213, 235, 245 Tett~onia viridissima 172, 173, 186, 196 8,200, 201,203, 206, 207, 211, 213, 225, 235, 236, 244

INDEX

tettigoniid ear 171 6 Therohia h,onidei 225, 232, 241 tiagibine 88 transmembrane o~-helical domains (TMDs) 63 Trichoplusia ni DAT in 103 dopamine in 102 EAAT (trnEEAT1) 64, 71 GABA transporter 86 glutamate in 91 octopamine in 107 orphan transporters in I13 serotonin transporter in 96 taurine transporter in 125 TRP (transient receptor potential) 32 TRl°-like channels 32 tryptophan hydroxylase (TPH) 59 Tympanophyllum arcu/blium 240 tyramine (TA) 78, 91 tyramine/~-hydroxylase (Trill) 58 tyrosine hydroxylase (TH) 58 vasodilator-stimulated phosphoprotein (VASP) 3O vesicular acetylcholine transporters (yAChT) 60 vesicular excitatory amino acid transporters (vEAATs) 60 vesicular inhibitory amino acid transporters (vIAATs) 60 vesicular monoamine transporters (vMATs) 60

X81IOI)IlS oocytes 72, 98 orphan transporters in 112, l 14 spinal neurons 36 Xenopus laevis 327 Xestoptera cornea 228 xylamine 105 zaprinast 42 Zoolerntopsis nevadensis 294, 300, 301