/RJLQ + 5HJLVWHU
%URXJKWWR\RXE\ 8QLYHUVLW\RI0HOERXUQH,QIRUPDWLRQ'LYQ
%URZVH
6HDUFK
0\6HWWLQJ...
63 downloads
1722 Views
4MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
/RJLQ + 5HJLVWHU
%URXJKWWR\RXE\ 8QLYHUVLW\RI0HOERXUQH,QIRUPDWLRQ'LYQ
%URZVH
6HDUFK
0\6HWWLQJV
$OHUWV
+HOS
+RPH
4XLFN6HDUFK
7LWOHDEVWUDFWNH\ZRUGV
-RXUQDOERRNWLWOH
$XWKRU
--This Journal/Book--
9ROXPH
HJMVVPLWK
,VVXH
3DJH
$GYDQFHVLQ,PPXQRORJ\
$OHUWPHDERXWQHZ%RRN9ROXPHV
&RS\ULJKW(OVHYLHU,QF$OOULJKWVUHVHUYHG
$GGWR)DYRULWHV
6KRUWFXW85/WRWKLVSDJHKWWSZZZVFLHQFHGLUHFWFRPH]SUR[\OLEXQLPHOEHGXDXVFLHQFHERRNVHULHV
$UWLFOH)HHG
$ERXWWKLV%RRN6HULHV
)XOOWH[WDYDLODEOH 1RQVXEVFULEHG :KDWGRHVWKLVPHDQ" 9ROXPH SS 9ROXPH SS 9ROXPH SS
$,'IRU,PPXQRJOREXOLQ'LYHUVLW\
9ROXPH SS 9ROXPH SS 9ROXPH SS 9ROXPH SS
&DQFHU,PPXQRWKHUDS\
9ROXPH SS 9ROXPH SS 9ROXPH SS 9ROXPH SS 9ROXPH SS
&XPXODWLYH6XEMHFW,QGH[9ROXPHV
9ROXPH SS 9ROXPH SS
7&HOO6XEVHWV&HOOXODU6HOHFWLRQ&RPPLWPHQWDQG,GHQWLW\
9ROXPH SS 9ROXPH SS 9ROXPH SS 9ROXPH SS 9ROXPH SS 9ROXPH SS
9ROXPH SS 9ROXPH SS 9ROXPH3DJHV
Ż3UHYLRXVYROLVV1H[WYROLVVŹ
(GLWHGE\)UHGHULFN:$OW ,6%1
$UWLFOH/LVW)XOO $EVWUDFWV
DUWLFOHV
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
%URZVH +RPH
6HDUFK
0\6HWWLQJV
$OHUWV
+HOS
$ERXW6FLHQFH'LUHFW_&RQWDFW8V_7HUPV &RQGLWLRQV_3ULYDF\3ROLF\ &RS\ULJKW(OVHYLHU%9$OOULJKWVUHVHUYHG6FLHQFH'LUHFWLVDUHJLVWHUHGWUDGHPDUNRI(OVHYLHU% 9
CONTRIBUTORS
Numbers in parenthesis indicated the pages on which the authors’ contributions begin.
Nehad M. Alajez (249), Department of Immunology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 T. Ball (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria Christophe Benoist (217), Section of Immunology and Immunogenetics, Joslin Diabetes Center, Boston, Massachusetts 02215, and Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts 02115 Shenghe Cai (331), Harvard Medical School, CBR Institute for Biomedical Research, Boston, Massachusetts 02115 Casey A. Carlos (249), Department of Immunology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 Alvin E. Davis, III (331), Harvard Medical School, CBR Institute for Biomedical Research, Boston, Massachusetts 02115 Anne Durandy (295), INSERM U429, Hoˆpital Necker-Enfants Malades, 75015 Paris, France Olivera J. Finn (249), Department of Immunology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 Alain Fischer (295), INSERM U429, Hoˆpital Necker-Enfants Malades, 75015 Paris, France M. Focke (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria Victor Ghetie (155), Cancer Immunobiology Center, University of Texas Southwestern Medical Center, Dallas, Texas 75390 Jessica C. Kettel (249), Department of Immunology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 D. Kraft (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria ix
x
CONTRIBUTORS
B. Linhart (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria Dongxu Liu (331), Harvard Medical School, CBR Institute for Biomedical Research, Boston, Massachusetts 02115 Diane Mathis (217), Section of Immunology and Immunogenetics, Joslin Diabetes Center, Boston, Massachusetts 02215, and Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts 02115 Patrick P. McDonald (1), Pulmonary Division, Faculty of Medicine, Universite´ de Sherbrooke, Sherbrooke, Que´ bec JIH 5N4, Canada Paul A. Monach (217), Section of Immunology and Immunogenetics, Joslin Diabetes Center, Boston, Massachusetts 02215, and Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts 02115 N. Mothes (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria Roald Nezlin (155), Department of Immunology, Weizmann Institute of Science, Rehovot 76100, Israel V. Niederberger (105), Department of Otorhinolaryngology, University of Vienna, Medical School, A-1090 Vienna, Austria Patrick Revy (295), INSERM U429, Hoˆ pital Necker-Enfants Malades, 75015 Paris, France Jason Rice (49), Molecular Immunology Group, Tenovus Laboratory, Cancer Sciences Division, Southampton University Hospitals Trust, Southampton SO16 6YD, United Kingdom S. Spitzauer (105), Clinical Institute for Medical and Chemical Laboratory Diagnostics, University of Vienna, Medical School, A-1090 Vienna, Austria Freda K. Stevenson (49), Molecular Immunology Group, Tenovus Laboratory, Cancer Sciences Division, Southampton University Hospitals Trust, Southampton SO16 6YD, United Kingdom I. Swoboda (105), Clinical Institute for Medical and Chemical Laboratory Diagnostics, University of Vienna, Medical School, A-1090 Vienna, Austria R. Valenta (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria Anda M. Vlad (249), Department of Immunology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 S. Vrtala (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria
CONTRIBUTORS
xi
K. Westritschnig (105), Division of Immunopathology, Department of Pathophysiology, University of Vienna, Medical School, A-1090 Vienna, Austria Delin Zhu (49), Molecular Immunology Group, Tenovus Laboratory, Cancer Sciences Division, Southampton University Hospitals Trust, Southampton SO16 6YD, United Kingdom
advances in immunology, vol. 82
Transcriptional Regulation in Neutrophils: Teaching Old Cells New Tricks PATRICK P. MCDONALD Pulmonary Division, Faculty of Medicine, Universite´ de Sherbrooke Sherbrooke, Que´bec JIH 5N4, Canada
I. Introduction
Polymorphonuclear neutrophils are by far the most abundant leukocyte population, representing about 60% of all circulating leukocytes. They are terminally differentiated cells in which the apoptotic program is constitutive—a characteristic that accounts for their limited lifespan, estimated to average between 16 and 24 hr in the bloodstream [1]. Neutrophils are best known for their traditional role as professional phagocytes. Because many neutrophils are present everywhere in the microvasculature at any point in time, they can rapidly extravasate and migrate toward an inflammatory focus in response to infection or injury. Under most circumstances, neutrophils are indeed the first leukocytes to infiltrate inflammatory sites, where they accumulate in large numbers, and deploy a battery of responses against invading microorganisms, foreign particles, or cellular debris. Paramount among these functional responses is the removal and destruction of non-self targets by way of phagocytosis and subsequent intracellular digestion. Neutrophils also mount a comprehensive microbicidal response against their targets by generating many oxygen-derived reactive intermediates, and through the massive release of prestored granular components ranging from bactericidal proteins to various lytic enzymes (including lipases, nucleases, phosphatases, various sugardegrading enzymes, and a wide array of proteases) [2]. Finally, activated neutrophils are known to synthesize large quantities of the important inflammatory lipid mediators, leukotriene B4 (a potent neutrophil chemoattractant) and platelet-activating factor (a powerful vasoactive lipid), which together facilitate the extravasation and recruitment of additional waves of granulocytes to inflamed tissues. As a result, neutrophils are rightfully viewed as a formidable first line of defense against pathogenic agents and other immunogenic material. Because of the indisputable aptitude of neutrophils to function as professional phagocytes, the various aspects of neutrophil function relating to this role have been extensively studied, and remain active fields of investigation to this day. An unforeseen consequence of this intense research activity, however, is that it effectively overshadowed another key facet of neutrophil biology, that is, the ability to express such important inflammatory mediators as cytokines, chemokines, adhesion molecules, and so forth. Up to roughly a decade ago, in 1 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00
2
PATRICK P. MCDONALD
fact, most immunologists considered neutrophils as cells that are capable of little, if any, gene transcription or protein synthesis. The notoriously short lifespan of the neutrophils probably contributed to this perception of the neutrophil as a one-trick pony. Even early indications of the ability of neutrophils to transcribe genes and synthesize proteins gave the impression that such a process was not quite developed in these cells. For instance, Tiku et al. showed early that neutrophils can release an interleukin (IL)-1-like activity [3], but it was not clear whether this reflected the processing and secretion of preformed cytokine precursor, or the actual de novo generation of IL-1. In another early study, Lindemann and colleagues had shown that in response to granulocyte-macrophage colony-stimulating factor (GM-CSF), neutrophils accumulate tumor necrosis factor (TNF) mRNA, but that no protein is released [4]. Similarly, early reports had demonstrated that neutrophils could inducibly express the mRNA encoding the protooncogene, c-fos, but whether or not this resulted in the formation of a functional protein remained unclear [5–7]. A host of studies performed since the early 1990s, however, were to forever alter this rather simplified view of the neutrophil, by demonstrating that these granulocytes can be induced to express a number of genes whose products lie at the core of inflammatory and immune responses. These include growth factors, cell surface receptors, adhesion molecules, cytokines such as TNF-a, IL-1b, IL-12, and tumor growth factor (TGF)-b, and a wide array of chemokines such as IL-8, Groa, Mip-1a/b, Mip-3a/b, IP-10, MIG, I-TAC, and others (reviewed in [8]). Among these mediators, chemokines are particularly relevant to inflammatory processes because of their ability to selectively recruit discrete cell populations into sites of injury, thereby effectively regulating leukocyte trafficking. As a result, the ability of neutrophils to produce various inflammatory mediators has far-reaching implications. In view of the fact that neutrophils are usually the first blood cells to infiltrate inflamed tissues, and that they vastly outnumber other leukocytes in a variety of chronic inflammatory disorders, their ability to generate various inflammatory cytokines is likely to have a significant impact on the initiation and evolution of inflammatory reactions. Moreover, their ability to produce a plethora of chemokines raises the possibility that neutrophils might actively contribute to the sequential recruitment of distinct leukocyte populations that is typically observed in many, if not most, inflammatory reactions. The production of inflammatory cytokines and chemokines by neutrophils is typically preceded by (and largely dependent upon) an accumulation of the corresponding mRNA transcripts (reviewed in [8]). In the particular case of IL-1b, IL-8, and Mip-1a, inducible gene expression was further shown by us and others to reflect an increased transcriptional activity, as determined in nuclear run-on assays [9–11]. In short, the ability of neutrophils to inducibly transcribe many genes, and to synthesize the corresponding proteins, is no
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
3
longer a matter of debate. In recent years, a growing number of investigators have begun to elucidate some of the regulated transcriptional processes in neutrophils, and though much remains to be learned, significant progress has already been made. The purpose of this review is to provide an overview of the current state of knowledge regarding the issue of transcription factor activation in the context of inducible gene expression in human neutrophils. II. The STAT Family of Transcription Factors
Originally described as interferon-dependent transcription factors, the signal transducers and activators of transcription (STATs) are a family of cytoplasmic proteins that represent a point of convergence for the signaling events triggered by such diverse biological agents as cytokines, growth factors, and hormones [12–14]. As a result, the inducible expression of a number of genes in response to various stimuli is under the partial or full control of the STAT proteins. Seven STAT genes have thus far been identified in mammals; they are termed STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B, and STAT6 [12]. All STAT transcripts have been shown to be regulated by differential splicing, resulting in the generation of various isoforms for each STAT family member [15–21]. Additional variants are generated by posttranslational proteolytic processing, at least in the case of STAT5; similar variants were also postulated in the case of STAT3 [22–24]. Because these spliced or truncated forms generally lack the transactivation domain, they are thought to act as negative regulators of transcription [18,25–28]. This being said, some investigators have proposed that under certain circumstances, spliced or truncated products could preferentially activate a subset of genes that is distinct from those activated by their full-length counterparts [17,25]. In addition, the shorter isoforms may associate with other transcription factors to either enhance or repress transcription in a manner distinct from the corresponding full-length forms [17,23,29]. It is therefore conceivable that cell-specific expression of truncated isoform could be associated with a different pattern of gene expression for those cells [30]. The STATs are latent cytoplasmic proteins, which upon cell stimulation are rapidly recruited to SH2 domains nested within the cytoplasmic tail of cell surface receptors. The STAT proteins subsequently become tyrosine phosphorylated under the action of Janus kinases (JAKs) in the case of cytokine receptors, whereas the phosphorylation step is usually carried out by an intrinsic receptor kinase activity in the case of growth factor receptors [31]. Noteworthy is that JAK activation has also been observed in growth factorstimulated cells, but does not seem to be a prerequisite for STAT activation in response to epidermal growth factor (EGF) or platelet-derived growth factor (PDGF) [12,32,33]. Once they have become tyrosine phosphorylated, STAT proteins can dimerize into various DNA-binding configurations, which differ
4
PATRICK P. MCDONALD
according to the identity of their constituting STAT protein components. For instance, STAT1, STAT3, and STAT5A/5B can all form homo- or heterodimers, whereas it appears that STAT4 and STAT6 can only homodimerize [19]. By comparison, STAT2 does not homodimerize, but rather forms a complex with STAT1 and p48/IRF-9 [34–37]. Following STAT protein phosphorylation and dimerization, the resulting complexes are swiftly translocated to the nucleus, where they can bind specific DNA sequences in the upstream regulatory region of target genes. Binding to these cognate sequences has been reported to occur with varying affinities, depending on each dimer’s composition. For instance, the STAT1–STAT2–p48 complex recognizes a motif termed interferon-stimulated response element (ISRE), whereas many other STAT dimers bind to variants of another motif, known as g-interferon-activated sequences (GAS) [14]. On a final note, it has been reported that many STAT proteins (namely, STAT1, STAT3, STAT4, and STAT5) can become phosphorylated on serine residues (in addition to tyrosine residues) prior to nuclear translocation, and it appears that this phosphorylation could contribute, at least in part, to STAT-driven gene transactivation [13,38]. Finally, signaling through the JAK /STAT cascade is under negative regulation by at least three distinct mechanisms. First, several protein tyrosine phosphatases, including SHP-1, SHP-2, PTP1b, and PTPeC, were shown to inhibit JAK activity, and likewise, STAT activity was reported to be regulated by cytosolic and nuclear phosphatases [39–46]. Another type of negative regulation is that exerted by a family of cytokine-induced immediate-early genes, which encompasses the cytokine-inducible SH2 protein (CIS), and the seven suppressors of cytokine signaling proteins, SOCS1–SOCS7. The genes encoding CIS, SOCS1, SOCS2, and SOCS3 are rapidly induced following cell stimulation (by cytokines, hormones, or growth factors) through the JAK / STAT pathway, and the resulting proteins inhibit the JAK/STAT in a negative feedback loop, thereby terminating signaling [47–50]. This can result from a direct interaction of SOCS with JAKs, in which case the kinase inhibitory domain (KIR) inhibits JAK kinase activity, or from an association with cytokine receptors through the SH2 domain of the SOCS proteins, in which case the access of STATs to the receptor docking sites is blocked [50,51]. A third class of negative regulators of the JAK /STAT pathway is the protein inhibitors of activated STATs (PIAS) family of proteins. The latter can directly interact with activated STAT dimers, and inhibit STAT-mediated gene activation by interfering with the DNA-binding activity of STATs [52–54]. A. Activation of STAT Proteins by G-CSF in Human Neutrophils In neutrophils, the first two studies to address the issue of STAT protein expression and activation were published 5 months apart in 1995–1996. In one of these studies [55], Tweardy and colleagues investigated the signaling events
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
5
triggered by G-CSF—a critical regulator of granulopoiesis and neutrophil maturation [56]. Whole-cell extracts were prepared from neutrophils disrupted by repeated freeze–thaw cycles, and analyzed by electrophoretic mobility shift assay (EMSA) using a human serum-inducible element (hSIE/m67) oligonucleotide probe. A specific complex was found to be induced in G-CSFtreated neutrophils, whose mobility on native gels was similar to that of a STAT3 homodimer present in extracts from Hep-G2 cells [55]. However, the authors were unable to supershift the complex using an antibody raised against the C-terminal (750–769) portion of STAT3, or using antibodies raised against other STAT proteins. This led them to conclude that a novel STAT-like protein must be expressed in granulocytes, which they tentatively designated STAT-G [55]. At about the same time, Bovolenta et al. similarly reported that cytoplasmic extracts prepared from neutrophils disrupted by nitrogen cavitation contain a G-CSF-inducible complex when analyzed in EMSA using an hSIE/m67 probe [57]. In contrast to the previous study, however, the latter group was able to identify the STAT3 as a major constituent of the inducible complex, with STAT1 as a minor participant. Because the authors were interested in elucidating the mechanisms whereby G-CSF induces the expression of FcgRI /CD64, an important player in phagocytosis and antibody-dependent cellular cytotoxicity (ADCC) mediated by activated polymorphonuclear (PMN) cells [58,59], they also performed EMSA analyses using a 39-bp oligonucleotide probe spanning the g-interferon response region (GRR) of the FcgRI/CD64 promoter, which is necessary and sufficient for inducibility [60]. Again, STAT3 and STAT1 were identified as components of the G-CSFinducible GRR-binding complex; in fact, the complex could be entirely supershifted using a combination of anti-STAT3 and anti-STAT1 antibodies. In agreement with these results, Bovolenta et al. further showed that G-CSF treatment of neutrophils resulted in the tyrosine phosphorylation of STAT3, as determined by immunoblot [57]. A subsequent study independently confirmed that STAT3 indeed becomes phosphorylated on both tyrosine and serine residues in response to G-CSF stimulation of neutrophils [61]. While it is now clear that G-CSF can activate STAT3 and STAT1 in neutrophils, it was initially somewhat unsettling that only one of the two groups who originally described the G-CSF-inducible complexes could identify its components, despite using similar antibodies in supershift assays. This discrepancy was briefly discussed by Bovolenta et al., who attributed the difference to the procedures used to prepare neutrophil extracts, namely, nitrogen cavitation versus freeze–thaw cycles. Indeed, nitrogen cavitation had already been shown to represent a gentle, yet efficient way to disrupt human neutrophils while preserving the integrity of intracellular components such as granules and nuclei [62]. Conversely, repeated freeze–thaw cycles was known as an efficient approach to break open the protease-rich neutrophil granules [62,63].
6
PATRICK P. MCDONALD
Consequently, it was likely that disrupting neutrophils in this manner would result in a large-scale release of granular proteases, which could potentially proteolyze STAT proteins to the point that they can no longer be recognized by antibodies. The issue would be resolved 2 years later by two independent studies. In one instance, the Tweardy laboratory published a follow-up to their original article, in which they reported that the G-CSF-inducible hSIEbinding activity could be partially supershifted using anti-STAT3 antibodies directed toward a different part of STAT3 than the one used in their prior study [24]. Accordingly, immunoblot analysis of nuclear extracts from G-CSFtreated neutrophils revealed that the STAT3a protein was now readily detectable; a weaker 72-kDa band was also detected, which they called STAT3g. Importantly, the abundance of the STAT3g band was dramatically increased when the cells were disrupted in the absence of the potent serine protease inhibitor diisopropyl fluorophosphonate (DFP) (as in their previous study). Together, these results strongly indicated that the STAT3 form(s) present in the neutrophil extracts was truncated. Indeed, the authors concluded that STAT3g was derived from STAT3a by limited proteolysis; they also proposed that STAT3g generation occurs as a result of G-CSF stimulation [24]. At the same time as the latter study was published, and in agreement with its findings, we showed that neutrophil disruption by classic means (such as detergent lysis or successive freeze–thaw cycles) often results in the partial degradation of various STAT proteins, even when an elaborate antiprotease cocktail is used, whereas nitrogen cavitation is the only cell disruption procedure that consistently yields intact (i.e., undergraded) STAT proteins in human neutrophils [64]. In the particular case of STAT3, classic cell disruption procedures affected the protein in such a way that it was no longer detected as a single 92-kDa species on SDS–PAGE, but rather as a two distinct bands, including a faster migrating species reminiscent of the one observed by the Tweardy group. Similarly, the G-CSF-inducible hSIE-binding complex migrated faster in EMSA and was partially reactive with some (but not all) anti-STAT3 antibodies. By contrast, every STAT3 antibody tested readily allowed the detection of STAT3 (both in immunoblot and in EMSA) when neutrophils were disrupted by nitrogen cavitation [64]. Thus, much of the early speculation about the identity of the G-CSF-inducible STAT complexes in neutrophils was related to the manner in which the cells were disrupted. In summary, it has become quite clear that G-CSF stimulation of neutrophils leads to the phosphorylation and nuclear recruitment of STAT3 and of STAT1. In addition, it has been suggested that STAT3g, a cleavage product of STAT3a, might also play a role in the context of G-CSF signaling [24]. In this regard, however, it must be pointed out that other investigators who have studied the expression and/or activation of STAT3 in neutrophils did not report the presence of a fast-migrating isoform of STAT3 [61,65,66]. Similarly,
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
7
we reported that only a single STAT3 band of expected molecular size is detected under conditions preventing the solubilization of neutrophil granule proteases [64]. Thus, further evidence will be needed to determine whether STAT3g is specifically generated following G-CSF stimulation of neutrophils, as opposed to being cleaved during cell disruption/lysis. On a final note, it can be envisaged that STAT activation by G-CSF has at least a few consequences in neutrophils. Among them, the inducible expression of CD64 is known to be largely STAT dependent in other leukocytes, at least in response to agents such as interferon-g (IFN-g) [60,67–69], and the rapid accumulation of CD64 transcripts that follows G-CSF-R engagement and STAT activation [57] in neutrophils certainly suggests a role for G-CSF-activated STATs in this process. Similarly, it was recently reported that in human neutrophils, both SOCS-1 and SOCS-3 are rapidly induced by G-CSF, and that at least SOCS-3 negatively regulates STAT protein activation by G-CSF [70]. This is perhaps one of the mechanisms whereby G-CSF-mediated signaling is terminated. Given that the expression of SOCS proteins is known to be largely STAT driven [71], and that in particular SOCS-1 expression by G-CSF is STAT3 dependent in murine myeloid leukemia cells [47], it is conceivable that SOCS induction by G-CSF in human neutrophils similarly involves STAT activation. B. Activation of STAT Proteins by GM-CSF in Human Neutrophils GM-CSF is another growth factor involved in the regulation of leukocytes of granulocytic and monocytic lineage, and is known to exert numerous and diverse actions in mature neutrophils. In this regard, it was first described to act upon neutrophils as a potent priming agent for most of their classic functions (i.e., chemotaxis, degranulation, reactive oxygen intermediate generation, phagocytosis, and lipid mediator synthesis) [72–79]. GM-CSF was later found to stimulate the gene expression of several inflammatory molecules in neutrophils, including chemotactic receptors, phagocytic receptors, cytokines, and related products [including interleukin (IL)-1b, IL-8, and IL-1ra], key components of the leukotriene biosynthetic pathway, and MHC class II molecules [74,80–86]. Another important action exerted by GM-CSF is its inhibition of neutrophil apoptosis, which also appears to involve the onset of gene expression [87–90]. These various effects of GM-CSF naturally prompted some investigators to start elucidating the signaling events involved in GM-CSF-elicited gene activation. First, Brizzi and colleagues reported that JAK2 becomes tyrosine phosphorylated in GM-CSF-stimulated neutrophils [65]. This was accompanied by the rapid phosphorylation of both STAT1 and STAT3 on tyrosine residues, and evidence for a physical interaction of the two STAT proteins with both JAK2 and the b-chain of the GM-CSF-R was also presented [65]. By
8
PATRICK P. MCDONALD
EMSA, Brizzi et al. further observed that GM-CSF promotes the binding of a complex to an hSIE probe, which was almost completely supershifted using antibodies to either STAT1 or STAT3, suggesting that it mostly represents a heterodimer [65]. This said, it is worthy of mention that this GM-CSFinducible neutrophil complex migrated conspicuously faster than an authentic STAT1/STAT3 heterodimer present in nuclear extracts from IL-6-stimulated HepG2 cells. The authors tentatively attributed this fact to a difference in the composition of the STAT1- and STAT3-containing complex between the two cell types. However, it is also possible that in that study, the STAT proteins were partially degraded as a result of disrupting neutrophils by detergent lysis in the presence of relatively few protease inhibitors. Indeed, we later showed that under such conditions, both STAT1 and STAT3 undergo partial proteolysis, but retain the ability to dimerize and bind an hSIE probe in response to neutrophil stimulation, yielding faster migrating complexes in EMSA [64]. This observation notwithstanding, much of the findings of Brizzi et al. would later be reproduced by other investigators in following years. Not unexpectedly, one of the most prolific laboratories investigating GMCSF/neutrophil interactions (and one of the first to document the effects of GM-CSF on neutrophil responses), that of P.H. Naccache, eventually became interested in the issue of GM-CSF nuclear signaling in these cells. In a 1998 paper, Al-Shami and colleagues confirmed that JAK2 activation does take place following GM-CSF stimulation of neutrophils, and further established that among STAT proteins, both STAT5B and STAT3 become tyrosine phosphorylated in response to GM-CSF [66]. Whereas STAT5 phosphorylation represented a strong and sustained response, that of STAT3 appeared to be a more transient phenomenon. By comparison, no increase in STAT1 phosphorylation was observed in response to GM-CSF, in apparent contradiction to the previous observation of Brizzi et al., who had found that GM-CSF induces STAT1 DNA binding in EMSA [65]. A possible explanation is that STAT1 was constitutively phosphorylated in unstimulated cells in the study of Al-Shami et al.—a rather unusual occurrence. This said, another important finding of the study by Al-Shami et al. was that following GM-CSF stimulation, DNA affinity precipitation of neutrophil cellular extracts using the GAS sequence of the FcgRI promoter led to the detection of STAT5B by immunoblot, indicating that the tyrosine-phosphorylated protein could interact with the GAS motif [66]. That STAT5 activation had not been detected in the earlier study of Brizzi et al. probably reflects the fact that the probe that they used in EMSA (hSIE) does not bind STAT5. We independently confirmed that GM-CSF stimulation of neutrophils results in the nuclear mobilization and DNA-binding activity of STAT5, as demonstrated by the finding that neutrophil nuclear extracts contain an inducible complex that binds the GRR of the FcgRI promoter in EMSA,
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
9
and that this complex is completely supershifted by STAT5 antibodies, but not by antibodies to other STAT proteins [64]. By comparison, EMSA analysis of the same extracts using an hSIE probe (which strongly binds STAT1 and STAT3) revealed a GM-CSF-inducible complex that could be efficiently supershifted by anti-STAT3 and (to a lesser extent) by anti-STAT1 antibodies [64] (our unpublished data). Thus, on the basis of the previous studies, it appears that at least three different STAT proteins can be activated by GM-CSF in human neutrophils. More recent studies mostly confirmed (and sometimes extended) our knowledge of the various aspects of STAT protein activation in GM-CSF-treated neutrophils. In one instance, Kuroki et al. confirmed that tyrosine phosphorylation of STAT3 occurs in GM-CSF-stimulated neutrophils, as initially observed by Al-Shami et al., and further showed that STAT3 also becomes phosphorylated on serine residues under the same conditions [61]. Although the significance of STAT3 serine phosphorylation remains poorly understood, it does appear to play a role in transcriptional activation, since mutations of the serine phosphorylation sites within STAT1 and STAT3 were found to impair their ability to drive gene transcription in the U3A cell line [91]. Another recent study described the inducible DNA binding of a shorter form of STAT5 in GM-CSF-treated neutrophils, which could be supershifted using an Nterminal antibody, but not with antibodies to the carboxy-terminus of STAT5; similar results were obtained in IL-5-treated eosinophils [92]. Importantly, incubating nuclear extracts from GM-CSF-treated neutrophils with nuclear extracts from activated HL-60 cells resulted in the degradation of the fulllength STAT5A and STAT5B present in the latter extracts into the truncated form observed in the neutrophil extracts. Moreover, conducting this experiment in the presence of phenylmethylsulfonyl fluoride (PMSF), a general serine protease inhibitor, partially prevented the generation of the truncated form. The authors proposed that granulocyte extracts contain a STAT5-specific serine protease that can process STAT5 into a novel DNA-binding isoform in GM-CSF-treated mature neutrophils, or in IL-5-treated mature eosinophils [92]. While this possibility is conceptually attractive, an alternative explanation is that neutrophil disruption by detergent lysis in the presence of few protease inhibitors is likely to promote the solubilization of granule proteases, which can then cleave STAT5, as demonstrated in our prior study [64]. Indeed, we showed that while full-length STAT5 was detected in nuclear extracts of GM-CSF-treated neutrophils disrupted by nitrogen cavitation, conventional cell disruption procedures yielded a truncated STAT5 form in immunoblot analyses, and a faster migrating native complex that was no longer recognized by C-terminal STAT5 antibodies in EMSA [64]. Thus, although the data of Caldenhoven et al. [92] confirmed that STAT5 is activated in response to
10
PATRICK P. MCDONALD
GM-CSF in neutrophils, it also reemphasizes the need to overcome the action of endogenous proteases when working with neutrophil extracts. On a final note, Epling-Burnette et al. [93] recently confirmed many previous findings relating to the activation of the JAK /STAT pathway by GM-CSF in neutrophils, but also offered new insights by establishing a link between these findings and the effect of GM-CSF on neutrophil survival. On the one hand, they confirmed that GM-CSF induces STAT-containing DNA-binding activities in neutrophils [93]. Using an hSIE probe, the inducible complex could be mostly supershifted by a STAT3 antibody whereas it was moderately affected by an anti-STAT1 antibody. A GM-CSF-inducible complex was also identified using an MGFe probe (which preferentially binds STAT5), and was mostly supershifted with an anti-STAT5B antibody. These authors similarly confirmed that GM-CSF promotes the tyrosine phosphorylation of STAT3, and to a lesser extent, of JAK2 as well. Accordingly, neutrophil pretreatment with AG-490, a selective inhibitor of JAK2 and JAK3 kinase activities, interfered with the GM-CSF-induced STAT3 tyrosine phosphorylation [93]. Importantly, the antiapoptotic effect of GM-CSF in neutrophils was also attenuated by AG-490, as well as by incubating neutrophils in the presence of STAT3 antisense oligonucleotides [93]. The latter findings therefore suggest a role for the JAK /STAT signaling cascade in GM-CSF-mediated neutrophil survival, even though it is clear that additional mechanisms are involved. In summary, the various studies conducted to date have established that in neutrophils, GM-CSF can activate several STAT proteins, as well as the upstream kinase, JAK2. Indeed, STAT1, STAT3, and STAT5 were all described to form nuclear DNA-binding complexes in response to GM-CSF, and accordingly, at least STAT3 and STAT5 were shown to become tyrosine phosphorylated under these conditions. This being said, several lines of evidence indicate that STAT5 might represent the predominant target of GM-CSF action in neutrophils, with STAT3 a close second. First, it was reported that although both STAT3 and STAT5 become tyrosine phosphorylated in response to GM-CSF, the phosphorylation of STAT5 represents a strong and sustained response, whereas that of STAT3 appears to be more transient [66]. By comparison, STAT1 tyrosine phosphorylation in response to GM-CSF has yet to be demonstrated. Second, only STAT5 binding was detected in EMSA using a probe consisting of the GRR of the FcgRI promoter [64], despite the fact that both STAT1 and STAT3 can efficiently bind this sequence [57,94]. This suggests that STAT1 and STAT3 are present in relatively weaker amounts than STAT5 in nuclear extracts from GM-CSF-treated neutrophils. In the case of STAT1, experiments conducted using hSIE probes have made it clear that it is less efficiently induced than STAT3 [64,93]. These considerations notwithstanding, much work remains to be done to determine the actual impact of individual STAT family members on the various
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
11
responses exerted by GM-CSF in neutrophils, and in particular on the induction of gene expression. In this respect, the identification of potential target genes remains far from obvious, as few of the genes induced by GM-CSF are known to be dependent on STAT proteins for activation. To further complicate matters, the up-regulation by GM-CSF of mRNA transcripts often reflects its known ability to dramatically enhance mRNA stability. On the basis of the previous studies, it could be expected that the gene encoding FcgRI /CD64 might represent a likely candidate, since its expression is known to be largely STAT dependent, and since GM-CSF promotes the binding of STAT5 dimers to a GAS sequence in its promoter [64,66]. Despite this, CD64 is not induced by GM-CSF in neutrophils. This is possibly because unlike G-CSF or IFN-g, which activate CD64 expression and which are potent STAT1 inducers, GM-CSF is a relatively poor STAT1 activator in human neutrophils [64,66,93] and monocytes [95]. This said, other potential candidates are intracellular proteins involved in the regulation of apoptosis, in view of the profound effect of GM-CSF on neutrophil survival, and of the apparent role of the JAK/STAT pathway in this phenomenon [93]. In particular, GM-CSF treatment of neutrophils is known to stimulate the mRNA and protein expression of the antiapoptotic protein Mcl-1 [93,96,97], and it was recently shown that a JAK inhibitor (AG-490) could partially block this process [93]. In keeping with this result, it was reported that a serum-inducible element-like motif present in the murine mcl-1 promoter was needed to confer inducibility [98]. Because this motif closely resembles the hSIE, it is likely that it can bind STAT1 and STAT3 with high afinity. This could therefore represent a way in which STAT protein activation by GM-CSF results in the onset of mcl-1 gene expression, should the human mcl-1 promoter share functional similarities with its murine counterpart. Similarly, GM-CSF has been described as rapidly inducing the expression of CIS1 and SOCS-3 in human neutrophils [99], and this process has been shown to require STAT protein activation in other systems [71], including in response to GM-CSF [100,101]. C. Activation of STAT proteins by IFN-g in human neutrophils The pleiotropic cytokine, IFN-g, is a well known STAT activator in many cell types. In neutrophils, IFN-g has been shown to modulate the expression of numerous genes, especially those encoding inflammatory cytokines and chemokines [8]. In addition, IFN-g can also directly induce several neutrophil genes. In particular, IFN-g was shown to rapidly increase FcgRI/CD64 gene expression in neutrophils [102,103], and to a greater extent than G-CSF [57]; this eventually results in a dramatic up-regulation of FcgRI surface expression [102–104]. A similar effect is exerted by IFN-g toward FcgRI in monocytes, and was described as involving the activation of JAK1 and JAK2, as well as the downstream activation of STAT1 [104–108]. These observations initially led us
12
PATRICK P. MCDONALD
to examine whether a similar phenomenon might take place in human neutrophils. Stimulation of neutrophils with IFN-g resulted in the rapid induction of two specific DNA-binding activities to the GRR, a regulatory element within the FcgRI promoter that is required for transcriptional activation. This response preceded the onset of FcgRI gene expression, and involved STAT1 activation, since the IFN-inducible complexes could both be completely supershifted by anti-STAT1 antibodies, but not by antibodies against other STAT family members [94]. In keeping with these results, the complexes could be substantially displaced using an antiphosphotyrosine antibody [94], and the induction by IFN-g of tyrosine-phosphorylated STAT1 was directly observed in a later study [99]. Because the IFN-g-inducible complexes observed in neutrophils comigrated with the ones induced in autologous peripheral blood mononuclear cells (PBMC), it was concluded that the more intense, faster migrating band represents a STAT1 homodimer, whereas the slower one presumably represents a STAT1 tetramer [94]—based on observations made in other experimental systems [109,110]. Together, these results demonstrated that in IFN-g-treated neutrophils, STAT1 is tyrosine phosphorylated and is able to bind the GRR of the FcgRI promoter, as well as the hSIE. Once again, it must be emphasized that the previous data were obtained in large part because neutrophils were disrupted under conditions that prevent the release of intracellular proteases, as we demonstrated in another study published at about the same time. Indeed, classic cell disruption procedures (namely, detergent lysis, sonication, and freeze–thaw cycles) invariably resulted in the loss of the slower migrating STAT1 complex believed to represent a tetramer, and in partially proteolyzed STAT1 homodimers that migrated faster than the authentic STAT1 homodimer observed in human PBMC [64]. Moreover, the truncated STAT1 complex no longer reacted with antibodies raised against the N-terminal region of STAT1 (amino acids 1–194), consistent with the requirement of an intact N-terminus for tetramer formation [109,110]. Similarly, antibodies recognizing the last 10 amino acids of STAT1 did not react with the truncated complex. By contrast, all antibodies efficiently supershifted the STAT1 homodimer in nitrogen-cavitated neutrophils [64]. Finally, the truncated STAT1 protein was also found to migrate faster on SDS–PAGE when neutrophils were disrupted by conventional means, as opposed to nitrogen cavitation [64]. In keeping with the importance of protecting STAT proteins from the action of endogenous neutrophil proteases, the first study to document that an hSIE-binding activity is induced in IFN-g-treated neutrophils reported a complex that migrated much faster than an authentic STAT1 homodimer present in nuclear extracts from IFN-g-treated HepG2 cells [55]. In that study, the neutrophils were disrupted by detergent lysis in the presence of a few protease inhibitors, and accordingly, the IFN-inducible complex failed to react with a C-terminal anti-STAT1 antibody (amino acids
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
13
716–739), and no band corresponding to a STAT1 tetramer was detected either. This effectively prevented the identification of the IFN-inducible complex, which would only become characterized as a STAT1 homodimer 3 years later [64,94]. A recent study has since independently confirmed that IFN-g indeed has the ability to activate STAT1 in neutrophils [93]. The impact of IFN-g-mediated STAT1 activation on inducible gene expression remains to be formally demonstrated in neutrophils. Nevertheless, it is to be expected that the rapid accumulation of FcgRI/CD64 mRNA following neutrophil exposure to IFN-g is due in large part to STAT1 activation by the latter, based on observations made in other inflammatory cells. Similarly, both CIS1 and SOCS-3 are rapidly induced by IFN-g in neutrophils [99], and the IFN-g-elicited induction of CIS/SOCS proteins has been shown to be STAT dependent in other cell types [111,112]. Finally, the induction of several inflammatory cytokines and chemokines (including IL-12, IP-10, MIG, and I-TAC) in neutrophils requires IFN-g as a costimulus [8], and on the basis of studies conducted in other leukocytes, a role for IFN-activated STAT1 (and perhaps STAT2 as well) can be envisaged. D. Activation of STAT Proteins by IL-10 in Human Neutrophils IL-10 is an important immunomodulatory cytokine whose effects are overwhelmingly antiinflammatory. In neutrophils, IL-10 has been shown to downregulate the inducible expression and release of numerous proinflammatory cytokines and chemokines [113–118]; conversely, it has also been demonstrated to up-regulate the production of antiinflammatory mediators such as IL-1ra [114,119]. Up to a few years ago, however, the issue of whether IL-10 could exert direct effects toward neutrophils (as opposed to modulating ongoing responses) was still largely a matter of speculation. In one of the first studies to address this question, we investigated whether IL-10 might activate STAT proteins and thereby induce FcgRI expression in human neutrophils. The underlying rationale was that in monocytes, IL-10 had been shown to stimulate both processes [120–122]. Somewhat disappointingly, we were unable to demonstrate any up-regulation by IL-10 of FcgRI gene and protein expression in neutrophils [94], a finding that was also reported by another group at about the same time [123]. Similarly, we found that IL-10 failed to activate STAT proteins in neutrophils, whereas it readily activated STAT1 and STAT3 in autologous PBMC [94]. The inability of IL-10 to activate STAT proteins in freshly isolated neutrophils was not likely to reflect the absence of IL-10 receptors, especially in view of the various immunomodulatory effects of the cytokine on neutrophil function. Binding experiments revealed that IL-10R was likely expressed on neutrophils [94]. We concluded by speculating that in neutrophils, IL-10 receptors were either insufficiently expressed, or that they lacked cytoplasmic domains required for the recruitment of JAK /STAT
14
PATRICK P. MCDONALD
proteins, to explain the absence of STAT protein activation in response to IL-10. The issue was eventually resolved as a result of the persistent efforts deployed by the laboratory of M.A. Cassatella, who recently demonstrated by FACS analysis that while IL-10R2 is expressed at similar levels on the surface of both neutrophils and monocytes, IL-10RI surface expression is very low in circulating human neutrophils [124]. By comparison, a 4-hr culture of neutrophils resulted in a significantly higher expression of IL-10R1 mRNA and surface protein, an effect that was greatly amplified if 100 ng/ml lipopolysaccharide (LPS) was also included in the culture medium [124]. More importantly, the authors showed that under conditions that were found to enhance surface IL-10RI expression, neutrophil exposure to IL-10 potently promoted the binding of a complex containing both STAT1 and STAT3 to either a GRR probe or an hSIE probe in EMSA [124]. This was also paralleled by the tyrosine phosphorylation of the STAT proteins by IL-10, again in marked contrast to freshly isolated cells. Thus, IL-10 can be an effective STAT inducer in human neutrophils, provided that both chains of its receptor are expressed. The identity of genes that might represent targets for the action of IL-10 through its ability to activate STAT proteins remains largely elusive, and this owes much to the fact that in neutrophils, IL-10 is known to induce only very few genes. For instance, IL-10 was shown to transiently induce SOCS3 gene expression (but not protein synthesis) in the absence of any detectable STAT activation [99], which indicates that this process can proceed independently of STAT proteins, at least in neutrophils. This being said, the ability of IL-10 to induce the expression of SOCS3 was greatly enhanced in cultured neutrophils already expressing the IL-10RI, and this also resulted in the production of the corresponding protein [124]. Thus, sustained SOCS3 expression in IL-10treated neutrophils correlates well with the ability of IL-10 to induce STAT activation in these cells. Although a direct link between these two processes has yet to be demonstrated, similar observations have also been made in other cellular systems [125]. In any instance, a likely consequence of SOCS3 expression is for the protein to participate in the termination of IL-10 signaling. E. Activation of STAT Proteins by Other Agents in Human Neutrophils Various other stimuli have been tested for their ability to promote STAT activation in neutrophils. For instance, the effect of prolactin was investigated in view of its known ability to activate STAT proteins in other leukocyte populations [126,127]. In this respect, Dogusan et al. reported that STAT1 becomes tyrosine phosphorylated upon prolactin stimulation of neutrophils, albeit to a lesser extent than when IFN-g is used [128]. This action of prolactin is paralleled by the tyrosine phosphorylation of JAK2 and STAT5 in human
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
15
PBMC, but not in neutrophils [127,128]. Prolactin treatment of neutrophils was also found to result in the binding of a complex to a GAS probe; while the complex was shown to be specific in competition experiments, its composition remains unknown, as it failed to react with STAT antibodies [128]. A probable reason for this outcome is that a detergent-based kit was used to prepare neutrophil extracts. Despite this, prolactin was shown to rapidly induce the expression of SOCS-2 and IRF-1 in neutrophils, whereas SOCS-3 was not induced [128]. In another study, pituitary growth hormone (GH) was shown to induce the tyrosine phosphorylation of both JAK2 and STAT3 in neutrophils; these effects of GH could be prevented by pretreating the cells with the general tyrosine kinase inhibitor, genistein [129]. Finally, Kuroki et al. have shown that more classic neutrophil agonists such as N-formylated peptides (fMLP), C5a, and phorbol esters (PMA) were all able to induce the phosphorylation of STAT3 on serine, but not on tyrosin residues [61]. The biological significance of STAT3 serine phosphorylation by these stimuli, however, remains poorly understood. F. Concluding Remarks and Future Directions A recurring theme in many of the studies that have investigated STAT activation in neutrophils has been the observation of truncated or unidentifiable STAT proteins and STAT-containing native complexes, which has repeatedly prompted the conclusion that new isoforms must be expressed in neutrophils. In all cases, however, neutrophils were disrupted by classic procedures, which results in the solubilization of neutrophil granule contents [64]. This in turn can be problematic, given that all STAT proteins are susceptible to cleavage by neutrophil proteases (ref. 64 and our unpublished data). For this reason, the question of whether novel STAT isoforms are expressed in neutrophils remains a matter of debate, and strongly emphasizes the paramount importance of shielding neutrophil proteins from proteolytic degradation during cell disruption. Since the adverse effects of neutrophil protease release can be successfully circumvented (and even prevented altogether), future studies can hardly afford to sidestep this issue. Although much has been accomplished since the first descriptions of STAT activation in neutrophils 6 years ago, further studies are needed to extend and deepen our understanding of this process in neutrophils (Table I). First and foremost, the activation of STAT family members other than STAT1/STAT3/ STAT5 needs to be further investigated, especially since all STAT proteins are expressed in these cells (our unpublished data). In this regard, indirect evidence for STAT4 activation by IL-12 in neutrophils was recently reported [130]. On a related issue, it will be interesting to determine which stimulatory conditions lead to the serine phosphorylation of STAT proteins in primary cells such as neutrophils. What we currently know is limited to the observations that
16
PATRICK P. MCDONALD
TABLE I Effect of Various Agonists Toward STAT Activation (Tyrosine Phosphorylation and/or Nuclear DNA Binding) in Human Neutrophils Agonist
STAT1
STAT3
STAT5
References
C5a fMLP G-CSF GM-CSF GH IL-2 IL-10 IFN-a IFN-g PMA Prolactin
nda nd þþ þ nd þþ þþþ nd þ
þb þb þþþ þþ þc þþ þc (þ)d þb nd
nd nd þþþ nd (þ)d nd nd
[61] [61] [24,57,61,99,124,249] [61,64–66,92,93,128] [129] [93] [124] [124] [64,93,94,99,124,128,249] [61,99] [128]
a
nd, not determined. Serine phosphorylation only. c Tyrosine phosphorylation only. d Activation of depicted STATs by these agents is controversial and would necessitate independent confirmation. b
STAT3 becomes serine phosphorylated in response to chemoattractants and PMA [61], whereas neither STAT1 nor STAT3 is serine phosphorylated following neutrophil exposure to IFN-g or IL-10 in freshly isolated neutrophils [99]. Other promising research areas include the activation of JAK / Tyk by neutrophil stimuli, and the negative regulation of the JAK / STAT pathway in neutrophils (by SOCS and PIAS proteins, as well as by phosphatases), which has only begun to be investigated. On a final note, it is likely that STAT activation is relevant to neutrophil gene expression. Indeed, several of the genes that are induced in neutrophils are known to depend, at least in part, on STAT proteins for inducibility. These include surface molecules such as the high-affinity IgG receptor, CD64, whose gene is induced by several STAT activators in neutrophils [57,94,99], and the antiapoptotic protein, Mcl-1, which is inducible by GM-CSF [93,96,97]. In addition, most of the stimuli that can activate STAT proteins in neutrophils were also found to induce the expression of CIS/SOCS family members in these cells [70,99,124]. Finally, a number of inflammatory cytokines/chemokines that are known to require STAT activation for their expression can be induced in neutrophils by IFN-g in concert with another stimulus [8], and there is little doubt that other potential STAT-driven genes expressed by activated neutrophils will be identified in future studies. This said, the implication of STAT activation in inducible gene expression still awaits a direct
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
17
demonstration in neutrophils. Because of the importance of STAT proteins in inflammation and immunity, it is likely that selective STAT inhibitors will become available in coming years that will allow researchers to shed more light on this aspect of neutrophil biology. III. The NF-kB Family of Transcription Factors
Since its discovery 15 years ago, the NF-kB transcription factor has emerged as a central regulator of inflammatory and immune processes, given its key involvement in the induction of a multitude of immediate-early genes, including those that encode cytokines and chemokines. This ubiquitously expressed transcription factor consists of a dimer of NF-kB/Rel proteins, a family of related macromolecules that can be subdivided into two groups. The first group includes the products of the Nfkb1 and Nfkb2 genes, which are large precursors (p105 and p100, respectively) of the mature p50 and p52 proteins. The second group encompasses the p65/RelA, c-Rel, and RelB proteins. All of these proteins have the ability to bind DNA through a conserved motif, the Rel homology domain, but only members of the second group feature a transactivation domain [131]. The Rel homology domain also makes it possible for these proteins to homo- and heterodimerize, resulting in multiple variants of the transcription factor, NF-kB [131]. In unstimulated cells, NF-kB dimers are usually complexed to cytoplasmic inhibitor proteins, collectively termed IkB proteins. Upon cell stimulation by a wide variety of agonists, the various intracellular signals that are engendered eventually converge on a multisubunit IkB kinase complex (IKK), which in turn phosphorylates IkB on discrete N-terminal serine residues [132,133]. This is soon followed by the ubiquitination of two nearby lysine residues, a process that is thought to act as a signal for proteolytic degradation by the proteasome pathway [134]. The degradation of IkB unmasks the nuclear localization signal of NF-kB constituent proteins, making it possible for the dimers to enter the nucleus, where they can bind DNA and induce the transcription of target genes. Additionally, the NF-kB constituent proteins themselves are subject to phosphorylation, a process that is required for the full induction of kB-driven genes [135,136]. Among those genes whose expression is most rapidly enhanced in a kB-dependent manner is the one that encodes IkB-a [137,138]. This, in turn, leads to the rapid resynthesis of the inhibitor protein [137,138], of which a portion accumulates in the nucleus where it is able to dissociate NF-kB dimers from DNA, and thereafter retarget NF-kB to the cytoplasm [139]. A few examples of kB-dependent genes are the ones encoding cytokines such as TNF-a, IL-1a/b, IL-1ra, and IL-12; chemokines such as IL-8, Groa, Mip-1b, Mip-3a/b, and IP-10; or surface receptors such as ICAM [140,141].
18
PATRICK P. MCDONALD
Because all of these genes are rapidly expressed by neutrophils in response to stimuli such as LPS or TNF-a [8], which are known to be potent activators of NF-kB in other systems [140], it was only a matter of time before researchers would start investigating the ability of neutrophils to activate NF-kB. A. Expression and Activation of NF-kB Pathway Components in Human Neutrophils The first study to document the presence of NF-kB/Rel proteins in neutrophils was the demonstration that G-CSF could induce the tyrosine phosphorylation of c-Rel in these cells [142]. The authors also showed that following G-CSF treatment, more c-Rel could be recovered from neutrophil cellular extracts by DNA affinity purification. The consequences of this intriguing phenomenon, which has not been further explored, remain unclear to this date. Nevertheless, this early work had the merit of showing that at least one key transcriptional activator was expressed in neutrophils, which were still viewed as little more than short-lived phagocytes specialized in antibacterial defense. Three years later, this view was further challenged by another article, in which we extensively characterized the expression and activation of various components of the NF-kB system in human neutrophils [143]. We first demonstrated that neutrophils express substantial amounts of NFkB1/p50, p65/ RelA, and c-Rel, whereas the p50 precursor, p105, was far less abundant. By contrast, RelB and NFkB2/p52 (or its precursor, p100) were undetectable in neutrophils. In contrast to most other cell types, in which the bulk of NF-kB/ Rel proteins is located in the cytoplasm of unstimulated cells, the NF-kB/Rel proteins expressed in neutrophils were found to be distributed in roughly equal amounts between the cytoplasmic and nuclear fractions [143]—an observation that was independently confirmed at a later date [144]. By immunoprecipitation, we also showed that the three main NF-kB/Rel proteins present in neutrophils form various heterodimers (i.e., p50/RelA, p50/c-Rel, or p65/c-Rel), and that in resting cells all of these combinations (and perhaps homodimers as well) are physically associated with IkB-a [143]. Additionally, the various NF-kB/Rel proteins were found to be fairly stable, as their cellular levels were not diminished after a 3-hr incubation in the presence of cycloheximide. By comparison, the inhibitor molecule IkB-a was found to turn over rapidly, with an estimated half-life of about 60 min [143]. Together, these observations clearly indicated that most of the prerequisites for a functional NF-kB cascade were fulfilled in human neutrophils. Indeed, activation of neutrophils with stimuli such as LPS, TNF-a, or IL-1b leads to the rapid loss (within minutes) of IkB-a and concurrent accumulation of NF-kB/Rel proteins in the nucleus; this is accompanied by a transient induction of NF-kB DNA-binding activity [143]. By supershift analysis, the major inducible complex was shown to contain p50, RelA, and (to a lesser
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
19
extent), c-Rel. Within 20–30 min of stimulation, an accumulation of IkB-a mRNA transcripts was already detectable, and the IkB-a protein was reexpressed shortly thereafter. The latter event correlated temporally with the termination of nuclear NF-kB DNA-binding activity [143]. Other neutrophil agonists were also found to activate NF-kB in these cells—in particular, chemoattractants such as platelet-activating factor (PAF), the formylated tripeptide, f MLP, and leukotriene B4. However, they proved to be weaker NF-kB activators than LPS or TNF-a, and they also elicited a more delayed response [143]. Because all the aforementioned stimuli can induce the expression of IkB-a and of many other kB-dependent genes in neutrophils (in particular, those encoding inflammatory cytokines and chemokines), and in view of their ability to activate NF-kB in these cells, it was proposed that NF-kB activation must underlie their action toward human neutrophil gene expression. Conversely, such neutrophil activators as GM-CSF, G-CSF, IFN-a, IFN-g, IL-8, and IL-10, which by themselves do not induce inflammatory cytokine expression in these cells, were found to lack the ability to activate NF-kB in neutrophils [143]. At about the same time as the latter study was published, Browning et al. reported that f MLP could activate NF-kB in human monocytes, but not in neutrophils, which led them to conclude that the activation of NF-kB by fMLP is cell specific [145]. In fact, these authors failed to detect any NF-kB DNA binding in neutrophils, even after stimulation with LPS or TNF. Because p50 and RelA were barely detected in neutrophil cellular extracts, the observed lack of NF-kB activation was attributed to insufficient amounts of p50 and RelA in these cells [145]. In that study, however, few precautions were taken to counter the action of endogenous neutrophil proteases, as all extracts were prepared by detergent lysis of the cells in a buffer that contained a very low concentration of PMSF as the sole protease inhibitor. In a paper that was published the following year, we showed that in neutrophils, NF-kB dimers as well as individual Rel proteins are indeed undetectable under these conditions, due to extensive proteolytic degradation [64]. This was in fact a general characteristic of neutrophil extracts prepared by conventional cell disruption procedures (i.e., detergent lysis, successive freeze–thaw cycles, or sonication), and strong evidence that the loss of NF-kB/Rel proteins was due to the solubilization of neutrophil proteases was presented. In particular, coincubation of these neutrophil extracts with nuclear extracts from activated Jurkat cells was shown to result in the degradation of authentic NF-kB dimers present in the latter [64]. Noteworthy is that significant degradation occurred even when an elaborate antiprotease cocktail was included in the various buffers. By contrast, disruption of neutrophils by nitrogen cavitation, a procedure that preserves granule integrity [62], allowed the recovery of intact NF-kB proteins and DNA-binding complexes, which comigrated with authentic Rel
20
PATRICK P. MCDONALD
proteins on SDS–PAGE and with genuine NF-kB dimers on native gels, and which were consistently recognized by antibodies directed at various parts of the proteins [64,143]. Thus, the apparent discrepancy between our initial study and that of Browning et al. stemmed from the different cell disruption procedures used, and the consequences thereof on neutrophil NF-kB components. This explanation also helps reconcile a very interesting observation made by Browning and colleagues with our demonstration of fMLP- and TNF-induced NF-kB in neutrophils. Indeed, these authors had observed that as promyeloid HL-60 cells differentiate into granulocytes, they acquire the ability to activate NF-kB in response to f MLP, while their ability to activate NF-kB in response to TNF remains unchanged [145]. This indeed suggests that f MLP can activate NF-kB in mature neutrophils. A number of subsequent studies, by us and other groups, confirmed the ability of neutrophils to activate NF-kB under various experimental conditions. In addition to the stimuli described above, several other agents were described as NF-kB activators in neutrophils. Because neutrophils had been shown to express both the b- and g-subunits of the IL-2/IL-15 receptor complex [146–150], and because IL-2 receptor engagement results in NF-kB activation in several cell types, we investigated the ability of IL-2 and IL-15 to induce NF-kB activation in neutrophils. We reported that the high-affinity IL-15Ra subunit is expressed in neutrophils, and that accordingly, IL-15 elicits several functional responses in these cells, including the production and release of IL-8 [151]. Moreover, IL-15 was shown to transiently enhance a nuclear NF-kB DNA-binding activity (albeit to a lesser extent than LPS), which could be supershifted by antibodies to p50 or RelA. By contrast, IL-2 exerted none of these effects [151], in keeping with the fact that neutrophils express little, if any, high-affinity IL-2Ra subunits on their surface [146,148– 150]. In another study, we investigated whether phagocytic stimuli might also prove to be NF-kB inducers, on the basis that phagocytosis represents an important physiological trigger for the inducible expression of several kBdependent genes in human neutrophils, including those that encode TNF-a, IL-1, and IL-8 [152–154]. In neutrophils undergoing the phagocytosis of IgGopsonized Saccharomyces cerevisiae, a DNA-binding activity primarily consisting of the classic NF-kB heterodimer, p50/RelA, was rapidly induced, and this was mirrored by a decline in immunoreactive IkB-a protein [155]. The cellular IkB-a pool was replenished by 30 min, and by that time, NF-kB binding had returned to near-basal levels. Similar results were obtained in neutrophils engulfing unopsonized yeast particles. We therefore proposed that NF-kB activation could constitute one of the mechanisms whereby the expression of kB-responsive genes is enhanced in phagocytosing neutrophils [155]. In agreement with the previous data, Wakshull et al. reported the activation of NF-kB in neutrophils ingesting PGG-glucan, a glucose homopolymer derived
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
21
from the cell wall of the yeast S. cerevisiae [156]. The inducible DNA-binding complex was shown to be specific, and to consist of p50 and RelA; an anti-p52 antibody did not affect the complex, in agreement with our earlier finding that the p52 protein is undetectable in neutrophils [143]. Activation of NF-kB by PGG was concentration dependent, and could be prevented by pretreating neutrophils with an anti-lactosylceramide antibody [156]. The ability of neutrophils to activate NF-kB in response to phagocytic activity was also observed during the uptake of opsonized Staphylococcus aureus [157]. Interestingly, these authors noted that disrupting neutrophils by detergent lysis resulted in a NF-kB DNA-binding activity that migrated noticeably faster than genuine p50/RelA dimers present in extracts from Wu¨ rzburg T cells, and that could not be supershifted. Mixing the neutrophil and T cell extracts led to the degradation of the complexes present in the latter extracts, indicating the presence of active proteases in the neutrophil extracts. By contrast, when the neutrophil extracts were prepared in the presence of the naturally occurring serine protease inhibitor, a1-antitrypsin, the NF-kB complexes had an almost normal mobility in EMSA, and were now partially reactive with an anti-RelA antibody [157]. These observations are in full agreement with our data showing that failure to counter the action of endogenous granule proteases results in the partial degradation of NF-kB constituents [64]. The issue of NF-kB activation by LPS in neutrophils was revisited by Sugita et al., who noted that this response could vary significantly depending on the origin of the LPS used, and (to a lesser extent) on the presence or absence of serum [158]. Under serum-free conditions, a time- and concentrationdependent activation of NF-kB was observed in response to LPS from Porphyromonas gingivalis 381, a maximal effect being obtained using 500 ng/ml LPS. By comparison, a similar signal was obtained using a 5-fold lesser concentration of LPS from Escherichia coli O111:B4 (which is the same LPS that we had used in our prior study). Thus, not all LPS preparations can activate NF-kB with the same potency. Activation of NF-kB by both types of LPS was enhanced when the cells were stimulated in the presence of autologous serum, even though it must be emphasized that the extent of that increase was of only about 50% [158]. Under these conditions, preincubation of neutrophils with an anti-CD14 antibody diminished by a half the ability of LPS to activate NF-kB; conversely, the antibody had no effect toward this response when the cells were incubated in the absence of serum. This is in keeping with the fact that serum contains a LPS-binding protein, whose interaction with LPS facilitates the ability of the latter to signal through CD14 engagement [159]. On a related topic, Page and co-workers [160] investigated the ability of bacterial LPS to activate transcriptional events compared to that of eosinophil-derived inflammatory proteins that are released in acute asthma settings where neutrophils are also present, and that may
22
PATRICK P. MCDONALD
therefore contribute to neutrophil activation. For this purpose, they used several eosinophil granule proteins such as eosinophil cationic protein, eosinophil-derived neurotoxin, eosinophil peroxidase, and major basic protein (MBP). Among them, none was found to induce NF-kB, despite the fact that MBP could induce IL-8 gene expression and release, to an extent comparable to the effect of LPS from E. coli O55:B5 [160]. MBP instead proved to be a very potent activator of chemokine mRNA stabilization. By contrast, LPS elicited a weak NF-kB DNA-binding activity within 20 min, and maximal induction took place at 60 min [160]. These characteristics of induction differ somewhat from those that we previously observed using LPS from another strain of E. coli (O111:B4), and may reflect differences in the ability of individual strains to activate transcription factors. Finally, two recent studies by Vancurova and colleagues added valuable new insight to our current understanding of the NF-kB system in neutrophils. In one instance, these authors investigated NF-kB activation in neutrophils from newborn and adult donors, and found that TNF induces NF-kB DNA binding more potently in newborn neutrophils [161]. This correlates well with the elevated production of proinflammatory cytokines such as IL-8 and IL-1b previously reported in neonatal neutrophils [162,163]. The TNFinduced NF-kB complexes consisted of a p50 homodimer and of the classic p50/RelA, and confirm our earlier findings made in TNF-treated neutrophils [143,151]. More importantly, Vancurova et al. also demonstrated that therapeutic concentrations of dexamethasone inhibit the induction of the NF-kB complexes by TNF, both in newborn and in adult neutrophils [161]. We observed a similar effect of dexamethasone in neutrophils from adult donors (our unpublished data). In a subsequent study, the same investigators demonstrated that in neutrophils, and in stark contrast to most other cell types, the IkB-a protein is distributed in approximately equal amounts between the cytoplasm and the nucleus [144]. Moreover, stimulation with TNF-a resulted not only in an accumulation of NF-kB/Rel proteins to the nucleus, as previously shown, but also in the degradation of IkB-a in both cellular compartments. This TNF-induced response was partially inhibited by MG-132, a proteasome inhibitor that is also known to inhibit NF-kB [164], as well as by the PKCd inhibitor, rottlerin [144]. B. Regulation of NF-kB Activity by Reactive Oxygen Derivatives in Neutrophils Endogenously generated reactive oxygen intermediates (ROI) have been implicated in the regulation of NF-kB activation in several cellular models [165,166]. Because neutrophils probably produce more ROI than any other cell type, they must be particularly well protected from the adverse effects of ROI, and indeed, neutrophils were shown to resist particularly well to high
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
23
concentrations of exogenous hydrogen peroxide [167]. As a result, it can be expected that the impact of endogenous ROI on transcription factor activation should be minimal in neutrophils. In the first study to characterize calcineurin activity in neutrophils, Carballo et al. observed that the inhibition of calcineurin using cyclosporin A also resulted in the partial inhibition of LPSinduced NF-kB DNA binding [168]. Because previous studies had shown a role for peroxides in the modulation of calcineurin activity in other cell types [169–171], Carballo et al. examined the effect of H2O2 on calcineurin and NF-kB activation in neutrophils. Following pretreatment of the cells with a catalase inhibitor, 3-amino-1,2,4-triazole (AMT), H2O2 dose-dependently inhibited neutrophil calcineurin activity, as well as the LPS-elicited NF-kB activation [168]. The authors concluded that H2O2 exerted its effect on NF-kB through its ability to inhibit calcineurin activity. However, given that calcineurin was unaffected by exogenous H2O2 if endogenous catalase was not deliberately inactivated [168], the same reasoning would imply that under these conditions, H2O2 does not affect NF-kB either. Unfortunately, this was not addressed experimentally. In this regard, however, Vollebregt and colleagues reported that the activation of NF-kB occurring in neutrophils ingesting opsonized S. aureus was unchanged in the presence of various oxidant scavengers (such as exogenous catalase, superoxide dismutase, or methionine), and that accordingly, exogenous H2O2 failed to activate NF-kB [157]. They concluded that in phagocytosing neutrophils, NF-kB activation is independent of endogenous oxidant generation. In full agreement with these results, we also observed that exogenous H2O2 (up to 1 mM) does not induce NF-kB in neutrophils (our unpublished data). While the previous results lend significant support to the idea that endogenous oxidants have little or no impact on NF-kB in neutrophils, a recent study reached the opposite conclusion. Zouki et al. used a FACS-based assay to detect individual transcription factor subunits in nuclei obtained from mixed leukocytes that had been purified following the stimulation of whole blood [172]. By gating on singlet nuclei (which were stated to represent PBMC nuclei) or on doublet nuclei (which were said to represent neutrophil nuclei), this assay reportedly allows us to discriminate between nuclei originating from these leukocyte populations. In this experimental setting, stimulation of whole blood with TNF-a and IL-1b resulted in the nuclear accumulation of RelA in both neutrophils and PBMC [172]. Pretreatment of whole blood with pyrrolidine dithiocarbamate (PDTC), an antioxidant that can function as an NF-kB inhibitor [173], suppressed the cytokine-induced nuclear mobilization of RelA in all leukocyte populations; PDTC similarly inhibited the release of IL-8 in whole blood under the same conditions [172]. Although these results indicate a potential role for endogenous ROI in the activation of NF-kB in neutrophils, some aspects of the FACS procedure make it difficult to conclude unambiguously.
24
PATRICK P. MCDONALD
First, the nature of each of the two populations of nuclei recovered following the lysis of mixed leukocytes was not ascertained. While it is quite obvious that they must represent neutrophils and PBMC, it must be hoped that they are free from cross-contamination. For instance, should some monocyte or lymphocyte nuclei somehow aggregate, then they would no longer behave as singlet nuclei, but rather as ‘‘doublets’’—a widely distributed population that contains the neutrophil nuclei. Any such contamination by PBMC nuclei could be problematic, in view of the fact that the ability of these cells to activate NF-kB dwarfs that of neutrophils [143]. At any rate, this issue would deserve to be further explored. Similarly, a correlation between results obtained using this FACS-based method, on the one hand, and either immunoblot detection of Rel protein nuclear mobilization or EMSA analysis of specific NF-kB binding, on the other hand, has never been established in neutrophils. A more important difficulty in interpretation comes from the fact that all stimulations (and inhibitor pretreaments) were carried out in whole blood, from which leukocytes are later isolated. Thus, it cannot be excluded that the modulation of a given neutrophil function might result from factors that are released by other cells during stimulation in whole blood. This might explain why, for instance, IL-1 was found to be a stronger activator of RelA nuclear movement than TNF following stimulation of whole blood and subsequent analysis of neutrophil nuclei [172], while the opposite is observed when isolated neutrophils are directly stimulated with TNF or IL-1 [143]. More to the point, it is impossible to rule out that the observed nuclear translocation of neutrophil RelA might be due in large part to soluble mediators generated by another redox-sensitive leukocyte population, which are no longer produced in the presence of antioxidants. Indeed, Niwa et al. reported that the TNF-induced NF-kB activation observed in isolated neutrophils was only slightly inhibited following PDTC pretreatment [174]. In conclusion, while it is quite clear that exogenously provided ROI such as H2O2 do not affect NF-kB activation in neutrophils, there remains a certain confusion as to whether endogenous ROI can affect the ability of neutrophils to activate NF-kB. While the available evidence definitely argues against this possibility in phagocytosing cells (which produce far more ROI than TNF- or LPS-treated neutrophils), one of the studies [172] presented previously still leaves room for reasonable doubt. Clearly, further work is required to definitively settle the issue. C. Regulation of NF-kB by Nitrogen-Derived Oxidants in Neutrophils Morphine is an effective immunomodulator, acting in part by stimulating nitric oxide (NO) production in neutrophils, monocytes, and endothelial cells [175,176]. Nitric oxide, in turn, inhibits gene expression in a variety of cells, and was shown to directly impair NF-kB binding in a cell-free system [177]. This is perhaps one of the mechanisms whereby both morphine and NO can
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
25
inhibit such leukocyte functions as adherence, phagocytosis, and motility [178– 180]. These considerations led Welters et al. to investigate the effect of morphine and NO on the ability of neutrophils to activate NF-kB [181]. Whole blood was pretreated with morphine sulfate or the NO donor, Snitroso-N-acetylpencillamine (SNAP), and stimulated with LPS for 30 min, prior to FACS detection of RelA in leukocyte nuclei as described in the previous section. Expectedly, LPS induced the accumulation of nuclear RelA; pretreatment of whole blood with morphine or SNAP inhibited this action of LPS in a dose-dependent manner, and maximal inhibitions of about 30 and 40% were achieved using 50 mM morphine and 1 mM SNAP, respectively [181]. Two NO synthase inhibitors totally blocked this modulatory effect of morphine, thereby confirming that it procedds via NO generation. Thus, it appears that NO can potentially affect NF-kB activation in neutrophils. By comparison, the LPS-induced nuclear mobilization of RelA was abrogated following whole blood pretreatment with the antioxidant N-acetylcysteine. This result suggests that most of the observed effect of LPS toward neutrophils involved the production of oxidants by some of the cells present in whole blood, perhaps even neutrophils themselves. Using the same FACS-based assay to detect nuclear RelA in neutrophils stimulated in whole blood, Zouki et al. reported that peroxynitrite (which results from the reaction of NO with H2O2) may underlie the action of cytokines toward NF-kB activation and IL-8 gene expression in human leukocytes [172]. Specifically, stimulation of whole blood with either TNF or IL-1 resulted in the production of peroxynitrite by all leukocyte populations; expectedly, NO synthase inhibitors blocked this response. The NO synthase inhibitors also suppressed the cytokine-induced IL-8 mRNA expression by mixed leukocytes, and the cytokine-induced accumulation of RelA in both PBMC and neutrophils nuclei. This indicated a role for endogenously generated peroxynitrite in NF-kB activation and in chemokine expression, and accordingly, addition of peroxynitrite to whole blood yielded the same results [172]. In summary, even though the whole blood/FACS approach used in these two studies makes it difficult to determine whether the various stimuli and inhibitors affect neutrophils directly or via other leukocytes (as discussed in the previous section), the previous results do raise the possibility that endogenously generated nitrogen-derived oxidants could participate in the inducible activation of NF-kB in neutrophils. D. Potential Role of NF-kB in Neutrophil Apoptosis A singular characteristic of neutrophils is their huge turnover rate, with an estimated 80 million being released in the bloodstream every minute. This is necessarily offset by an equivalent loss of neutrophils—a reflection of the fact
26
PATRICK P. MCDONALD
that in these terminally differentiated leukocytes, the apoptotic program is constitutively engaged, resulting in a an average lifespan of about 20 hr (though they can last somewhat longer in culture). As a result, spontaneous apoptosis and its modulation by various inflammatory agents have been quite extensively characterized in neutrophils [182]. By contrast, the molecular events underlying this process remain largely undefined. Among several potential pathways that could participate in the regulation of neutrophil apoptosis, NF-kB stands out in view of its key role in the induction of various antiapoptotic cellular proteins, as described in many experimental models [183]. In the first study to address the possible link between NF-kB activation and apoptosis in neutrophils, Ward et al. demonstrated that several distinct NF-kB inhibitors (i.e., curcumin, PDTC, and the proteasome inhibitor, MG-132) increased the constitutive apoptosis of neutrophils after an overnight incubation [184]. A similar effect was exerted by the fungal metabolite, gliotoxin, which had been shown to be a potent and specific inhibitor of NF-kB [185]. This prompted these authors to examine the effect of gliotoxin on a proapoptotic stimulus, TNF-a (which accelerates neutrophil apoptosis at early time points), and on an antiapoptotic stimulus, LPS. At the 2-hr time point, gliotoxin and TNF synergistically increased apoptosis in neutrophils, whereas the survival effect of LPS was completely suppressed by gliotoxin [184]. Because both LPS and TNF are potent NF-kB activators [143], these results raised the possibility that gliotoxin might interfere with the protective effect of NF-kB on apoptosis. Indeed, preincubation of neutrophils with gliotoxin was found to inhibit NF-kB activation by either LPS or TNF [184]. The authors concluded that NF-kB plays a crucial role in regulating the physiological cell death pathway in granulocytes. In subsequent years, other investigators would follow a similar approach, and reach similar conclusions for the most part. For instance, Niwa et al. also observed that TNF induces NF-kB activation, as well as the concomitant degradation of IkB-a, albeit without affecting spontaneous neutrophil apoptosis at early time points (i.e., within 2 hr of incubation) [174]. Nevertheless, by interfering with the ability of TNF to induce NF-kB, the proapoptotic effect of TNF became apparent. The authors therefore concluded that the NF-kB pathway is crucial for neutrophil survival against TNF-a cell toxicity. Similarly, Nolan et al. reported that the inhibition of neutrophil apoptosis that occurs after 12–18 hr of incubation with either TNF-a or LPS was no longer observed in cells pretreated with the proteasome inhibitor PSI-I [186]. Pretreatment of neutrophils with PSI-I also abolished the ability of TNF-a and LPS to induce the nuclear translocation of RelA [186]. By contrast, Dunican et al. reported that the inhibition of neutrophil apoptosis observed following an overnight incubation in the presence of TNF-a was only slightly affected (by
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
27
about 10%) by prior treatment of the neutrophils with SN50, a specific inhibitor of NF-kB [187]. While this result is in apparent contrast to the findings of other groups, it must be stressed that no data were presented to confirm that SN50 did in fact prevent the activation of NF-kB by TNF in their experiments. Finally, Ward et al. recently examined the effect of prostaglandin D2 and of its metabolites on neutrophil apoptosis. Among the latter, both D12 PGJ2 and 15dPGJ2 were found to be proapoptotic in neutrophils, and to suppress the LPS-elicited inhibition of neutrophil apoptosis [188]. Moreover, both lipids were able to inhibit the LPS-induced degradation of IkB-a in neutrophils. Collectively, these various studies indicate that NF-kB is likely to play a protective role against the onset of neutrophil apoptosis. E. Concluding Remarks and Future Directions Whereas NF-kB activation (and the immediately related events) is a process that has been relatively well characterized in neutrophils stimulated by a wide array of agonists (Table II), many questions remain and deserve to be further explored. For instance, a role for reactive oxygen intermediates in this process
TABLE II Effect of Various Agonists Toward NF-kB Activation in Human Neutrophils Agonist
Effect
Eosinophil granule proteins fMLP G-CSF (up to 103 U/ml) GM-CSF (up to 3 nM) H2O2 (up to 1 mM) IL-1b IL-2 (up to 103 U/ml) IL-8 (up to 100 nM) IL-10 (up to 300 U/ml) IL-15 IFN-a (up to 103 U/ml) IFN-g (up to 300 U/ml) LPS
þ þþ þþ þþþ
Leukotriene B4 Platelet-activating factor Phorbol ester TNF-a S. cerevisiae IgG-opsonized yeasts IgG-opsonized S. aureus
þ þ þ þþþ þ þþ þþ
References [160] [143] [143] [143] [157; McDonald, unpublished] [143] [151] [143] [143] [151] [143] [143] [64,143,156–158,160,168,172,181,184,186,188,250] [155] [143] [143] [143] [64,143,144,157,161,174,184,186] [155] [155] [157]
28
PATRICK P. MCDONALD
remains unclear, and the potential involvement of nitrogen-derived oxidants in NF-kB activation is equally uncertain. Another potentially important issue that has remained untackled to this day is that of whether NF-kB constituent proteins become phosphorylated in response to various neutrophil agonists. Yet another area that remains mostly uncharted consists in the elucidation of the signaling steps that take place upstream of IkB degradation and/or NF-kB subunit phosphorylation. In view of the particularities of the NF-kB system (and its activation) in human neutrophils, such investigations are likely to be of broad interest. Finally, the collective evidence gathered to date makes it very likely that NF-kB activation plays an important role in inducible gene expression in neutrophils. A particularly telling example is that many genes encoding inflammatory cytokines and chemokines, which have been shown to be kB dependent in other cells, are induced in neutrophils by stimuli that activate NF-kB in these cells. Despite this, the actual impact of NF-kB on neutrophil gene expression has yet to be directly demonstrated in neutrophils. Conversely, there is good evidence that NF-kB plays a protective role in the context of spontaneous as well as modulated neutrophil apoptosis, but the targets of this NF-kB action have not been identified yet. There is little doubt at this stage that the continuing study of NF-kB activation and function in neutrophils should soon yield substantial advances in these areas of research. IV. Other Transcription Factors Potentially Involved in Neutrophil Activation
In addition to the STAT and NF-kB families, other transcription factors have been reported to be expressed in neutrophils, and others still are likely to be present. They will be briefly covered in this section. A. The Ets Family of Transcription Factors Transcription factors of the Ets family share a DNA-binding domain spanning about 85 amino acids, termed the Ets domain. This allows them to specifically bind to a series of related purine-rich DNA sequences, which regulate the expression of numerous cellular (and even viral) genes. The importance of the Ets family is illustrated by the fact that many such genes are critical for biological processes ranging from cellular proliferation and differentiation to apoptosis [189]. Among Ets family members, PU.1 is expressed specifically in hematopoietic lineages. It binds DNA as a monomer via its C-terminal Ets domain, to the consensus site 50 -AAAG(A /C/G)GGAAG-30 [190]. Phosphorylation of serine 148 in the N-terminal transactivating domain of PU.1 also allows it to interact with transcription factors of the interferonregulated factors (IRF) family, including IRF-4 [191]. Such an interaction makes it possible for the resulting complex to bind composite elements on
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
29
DNA, thereby adding an element of diversity to potential PU.1 targets. The interaction of IRF-4 or IRF-8/ICSBP with PU.1 in monocytic cells was indeed shown to enable transactivation via hybrid DNA sequences [192,193]. An essential role for PU.1 in neutrophil differentiation was demonstrated in mice with a targeted disruption of the PU.1 gene [194]. Such animals are born with no detectable leukocytes, and die of systemic bacterial infection within 2 days of birth. Nevertheless, their lifespan could be prolonged for up to 2 weeks with antibiotic therapy; under these conditions, T lymphocytes and neutrophillike cells appeared in the first week. Subsequent studies by the same group showed that neutrophils from PU.1-null mice do not express the genes encoding receptors for G-CSF and GM-CSF [195], or those encoding numerous secondary granule components, including CD11b, lactoferrin, gelatinase, lysozyme, or the NADPH oxidase component, gp91phox [196]. This is consistent with the observed inability of these neutrophils to mount a respiratory burst, and with their poor efficiency at bacterial killing, relative to normal cells [196]. Thus, while PU.1 is not required for early myeloid lineage or neutrophil commitment, it is essential for the full differentiation and maturation of neutrophils. The apparent role of PU.1 for the expression of numerous granule proteins is consistent with its reported requirement for the inducibility of most of the corresponding genes. High levels of PU.1 mRNA have been shown to be expressed in mature human neutrophils, whereas another highly related Ets family protein, Spi-B, is not expressed [197]. Accordingly, constitutive binding of PU.1 to its cognate sequence within the CD11b or gp91phox promoters has been shown in EMSA using neutrophil nuclear extracts [197,198]. Although the resulting complex was shown to be specific, it was observed to migrate much faster than the PU.1 complexes present in monocyte or B-lymphocyte nuclear extracts [197,198], and did not react with an antibody raised against the PU.1 N-terminal domain [197]. That another antibody directed to the DNA-binding domain of PU.1 did supershift the complex [197,198], indicated that the PU.1 recovered from the neutrophil extracts was partially degraded. This was confirmed in mixing experiments, in which incubation of in vitro-translated PU.1 with neutrophil extracts resulted in its truncation into the same fast-migrating complex that did not react with an N-terminal antibody [197]. This is consistent with our demonstration that several distinct transcription factors become degraded when neutrophils are disrupted by detergent lysis or other conventional means [64], as in the above studies [197,198]. Thus, PU.1 is expressed in mature neutrophils, and it constitutively binds DNA. Whether this binding can be further induced, or whether it can be modulated at all, remains to be demonstrated. On a final note, it is possible that Ets proteins other than PU.1 are expressed in neutrophils. Indeed, Tsutsumi-Ishii et al. have reported that in nuclear extracts from resting neutrophils, a DNA-binding activity could be
30
PATRICK P. MCDONALD
detected in EMSA using an oligonucleotide containing the Ets-like motif of the a-defensin (or HNP) promoter [199]. This DNA-binding activity was unaffected by an unlabeled PU.1 probe, but partially abrogated by a cold Ets oligonucleotide; the nature of the complex was not further explored. A role for PU.1 in inducible gene expression by mature neutrophils remains highly speculative at this stage. Nevertheless, there exist cases in which such a role is at least conceivable. For instance, PU.1 was shown to play a role in the induction of the gp91phox gene by IFN-g in monocytic cells, by a mechanism that does not increase PU.1 constitutive DNA binding [200,201]. The possibility therefore exists that a similar mechanism might be involved in the upregulation of gp91phox mRNA by IFN-g in neutrophils. Likewise, many others genes containing sites for PU.1 (or for other Ets family members) in their promoter region, and whose expression can be modulated in mature neutrophils, could represent possible targets for PU.1/Ets regulation. Examples include FcgRIb, MHCII, ICAM-1, IL-1b, IL-1ra, TNF-a, MIP-1a, CXCR1 (the IL-8 receptor), and the antiapoptotic protein, Mcl-1 [189]. Finally, several genes encoding proteins formed during granulocytic differentiation have been shown to be under the control of PU.1 These include myeloperoxidase, neutrophil elastase, lysozyme, defensins, lactoferrin [202], and G-CSF and GMCSF receptor subunits [203,204]. However, the expression of most of the latter genes has not been reported to be modulated in mature neutrophils. B. The C/EBP Family of Transcription Factors The CCAAT/enhancer-binding proteins (C/EBP) are transactivators known for their involvement in the regulation of acute phase protein expression [205], as well as that of numerous inflammatory chemokines, including IL-8, ENA78, Groa, and Mip-1a [206–210]. The family consists of several isoforms, including C/EBPa, C/EBPb, C/EBPd, C/EBPe, and CHOP, which can homo- and heterodimerize through their basic leucine zipper region [211], and even heterodimerize with other transcription factors, such as NF-kB [206,212]. Activation of the C/EBP proteins involves their phosphorylation in a negative regulatory domain, which is thought to relieve the inhibition of their DNA-binding and transactivation domains [213,214]. While various kinases have been proposed to phosphorylate the C/EBP proteins, ERK1/2 stand out among the most likely candidates [215]. In a manner similar to PU.1, certain C/EBP family members have been described as essential for granulopoiesis and terminal neutrophil differentiation [202]. More specifically, C/EBPa null mice lack neutrophils and eosinophils, but retain monocytes and lymphocytes [216], while C/EBPe null animals show important defects in neutrophil function, including the lack of secondary granules [217,218]. By comparison, mature neutrophils from mice with a targeted disruption of the C/EBPb or C/EBPd genes did not appear to suffer
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
31
from any severe defect [219–221]. In addition to their involvement in neutrophil development and maturation, C/EBP proteins can also be expected to participate in the induction of several genes in mature neutrophils, based on their known ability to regulate the expression of many proinflammatory mediators such as IL-1b, IL-12, IL-8, MIP-1a, MIP-1b, and TNF-a [207,222–226], which can all be induced in neutrophils [8]. This, in turn, begs the question of whether C/EBP proteins are expressed in mature human neutrophils, and various pertinent observations suggest that they probably are. Indeed, it has been reported that in murine granulocytecommitted progenitor cells, C/EBPa, b, and d exist as nuclear proteins [227,228]; nuclear C/EBPa has also been detected in human cells assuming a neutrophilic phenotype (i.e., in dimethylsulfoxide [DMSO]-differentiated HL-60 leukemia cells and in some bone marrow cell subsets) [228]. In addition, C/EBPe mRNA is greatly induced during the in vitro granulocytic differentiation of human primary CD34þ cells or of HL-60 cells [229]. While suggestive, these findings will nonetheless have to be confirmed in mature, terminally differentiated human neutrophils. In this respect, it was recently reported that neither C/EBPa nor C/EBPb proteins are detectable in neutrophil nuclear extracts, be it in resting cells or following stimulation with either TNF-a or LPS [199]. Accordingly, no DNA binding to a C/EBP probe is detected in EMSA under the same conditions [160,199]. However, it must be pointed out that these extracts were prepared following cell disruption by detergent lysis in the absence of elastase-class protease inhibitors, a procedure that results in the proteolysis of many neutrophil transcription factors [64]. Thus, it cannot be ruled out that the absence of detectable nuclear C/EBP proteins might have been consequent to proteolytic degradation. Clearly, more studies are required to determine if C/EBP transcription factors are expressed and functional in neutrophils. C. The AP-1 Family of Transcription Factors The AP-1 transcription factor typically consists of combinations of Jun and Fos family proteins, which bind as dimers to a common enhancer sequence on gene promoters. Members of the Jun family of proteins (c-Jun, JunB, JunD) have the ability to homo- and heterodimerize among themselves, or to dimerize with Fos family proteins, resulting in multiple AP-1 variants [230]. By contrast, Fos family proteins (c-Fos, FosB, Fra-1, Fra-2) must associate with Jun proteins to form AP-1 complexes [230]. Dimerization occurs via a basic leucine zipper domain within the Jun/Fos proteins, which also makes it possible for them to associate with members of different transcription factor families, such as NF-kB/Rel proteins [231], ATF/CREB proteins [232,233], and a host of others [234,235]. Most cell types contain some Jun/Fos proteins, and upon cell stimulation, their genes are actively transcribed—an immediate-early
32
PATRICK P. MCDONALD
process that usually yields more Jun/Fos proteins, which are then available to form DNA-binding dimers in the nucleus. Posttranslational modification also plays a key role with regard to AP-1 function. Phosphorylation of c-Jun by c-Jun N-terminal kinase (JNK) and of c-Fos by for Fyn-related kinase (FRK) has been shown to greatly potentiate the transactivation potential of AP-1 complexes [236–238]. Similarly, it has been recently reported that JNKs can phosphorylate JunD and thereby enhance its ability to activate transcription [239], even though JunD was observed to bind JNK with much lower affinity than c-Jun. Several inflammatory mediators generated by activated neutrophils are known to contain AP-1 elements within their upstream regulatory region that are either essential or needed for full promoter activity. A partial list of these mediators includes TNF, IL-1, IL-1ra, IL-8, and ICAM [240–244]. Moreover, these mediators are often induced in response to known AP-1 activators in neutrophils. These observations therefore stress the importance of establishing whether AP-1 components are expressed and activable in neutrophils. In this regard, neutrophils have been reported to constitutively express the genes encoding c-Fos, c-Jun, JunB, and JunD [245–247]. The steady-state level of these transcripts can be rapidly up-regulated in response to formyl peptides, GM-CSF, G-CSF, TNF, or phorbol esters in the case of c-Fos [5–7], or in response to LPS, TNF, and phorbol esters in the case of the Jun family members [247,248]. Whether the corresponding proteins are also susceptible to modulation (or whether they are expressed at all), and whether they can form functional AP-1 complexes, remains to be demonstrated in neutrophils. In the latter case, Page et al. observed a moderate increase in AP-1 binding in EMSA when neutrophils were stimulated for 60 min with LPS [160]. Because no supershifts were performed, and because the inducible AP-1 complex was partially displaced by a nonspecific competitor, it is difficult to determine whether the DNA-binding activity was made up of AP-1 dimers or of other constituents. Nevertheless, these results constitute another hint that AP-1 might be activatable in neutrophils. Again, more studies are needed before any conclusion can be drawn. D. Concluding Remarks and Future Directions Our current knowledge of the transcription factors covered in this section in relation to neutrophil activation is fragmentary at best. In the case of the AP-1 and C/EBP families, it seems reasonable to suggest that they may play a role in the expression of many inflammatory cytokines and chemokines that are inducible in neutrophils, and that have already been shown in other cellular models to depend, at least in part, on these factors. By contrast, it would appear that although the PU-1 factor is expressed in neutrophils, its involvement in the regulation of gene expression is less certain, given that most of the PU.1-driven genes are usually expressed during neutrophil differentiation
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
33
and maturation. Nevertheless, there exist circumstances (such as the induction by IFN-g of the gp91phox gene, as mentioned earlier) in which a role for PU.1 cannot be ruled out. Clearly, further investigation is required to start evaluating the potential impact of the previous transcription factors on gene expression in mature neutrophils. V. Conclusion
Progress made in the last decade has made it clear that the traditional representation of neutrophils as professional phagocytes is outdated, and that in addition to this important role, neutrophils can also influence the course of inflammatory and immune reactions through their ability to express genes that encode numerous prominent inflammatory mediators. In fact, transcriptional control of neutrophil gene expression represents a new facet of neutrophil biology that is gradually emerging as a research field in its own right. Despite the substantial advances that have been made, however, much more remains to be learned. One of the most important questions that remains unaddressed is whether transcription factor activation has an impact on inducible gene expression in neutrophils. Even though this is highly probable, only a direct demonstration can satisfactorily settle the issue. Other research areas that will necessitate further investigation in neutrophils include the identification of transcriptional regulators other than NF-kB and STAT proteins, which are either activated or mobilized upon neutrophil stimulation; the possibility that there can be crosstalk or cooperation between transcription factor families; the characterization of the signaling cascades lying upstream of transcription factors; and the elucidation of the mechanisms responsible for terminating transcriptional activation. In a broader perspective, a better understanding of transcriptional regulation in neutrophils could help unveil new pharmacological targets for the treatment of a number of chronic inflammatory conditions in which neutrophils clearly predominate over other cells types, and therefore constitute a likely source of many inflammatory mediators that are typically found in high levels in these conditions. If the strides made in the past 6 years are any indication of future developments in this new field of investigation, then the years ahead should prove quite exciting indeed. Acknowledgments The author would like to thank Mr. Alexandre Cloutier for some preparatory work done toward the completion of this review, and to address special thanks to Dr. Marco Cassatella for his critical reading of the manuscript. P. P. McDonald is a Scholar of the Canadian Institutes for Health Research.
34
PATRICK P. MCDONALD
References 1. Mary, J. Y. (1984). Normal human granulopoiesis revisited. I. Blood data. Biomed. Pharmacother. 38, 33–43. 2. Borregaard, N., and Cowland, J. B. (1997). Granules of the human neutrophilic polymorphonuclear leukocyte. Blood 89, 3503–3521. 3. Tiku, K., Tiku, M. L., and Skosey, J. L. (1986). Interleukin 1 production by human polymorphonuclear neutrophils. J. Immunol. 136, 3677–3685. 4. Lindemann, A., Riedel, D., Oster, W., Meuer, S. C., Blohm, D., Mertelsmann, R. H., and Herrmann, F. (1988). Granulocyte/macrophage colony-stimulating factor induces interleukin 1 production by human polymorphonuclear neutrophils. J. Immunol. 140, 837–839. 5. Colotta, F., Wang, J. M., Polentarutti, N., and Mantovani, A. (1987). Expression of c-fos protooncogene in normal human peripheral blood granulocytes. J. Exp. Med. 165, 1224–1229. 6. Itami, M., Kuroki, T., and Nose, K. (1987). Induction of c-fos proto-oncogene by a chemotactic peptide in human peripheral granulocytes. FEBS Lett. 222, 289–292. 7. McColl, S. R., Kreis, C., DiPersio, J. F., Borgeat, P., and Naccache, P. H. (1989). Involvement of guanine nucleotide binding proteins in neutrophil activation and priming by GM-CSF. Blood 73, 588–591. 8. Cassatella, M. A. (1999). Neutrophil-derived proteins: Selling cytokines by the pound. Adv. Immunol. 73, 369–509. 9. Marucha, P. T., Zeff, R. A., and Kreutzer, D. L. (1991). Cytokine-induced IL-1 beta gene expression in the human polymorphonuclear leukocyte: Transcriptional and post-transcriptional regulation by tumor necrosis factor and IL-1. J. Immunol. 147, 2603–2608. 10. Cassatella, M. A., Gasperini, S., Calzetti, F., McDonald, P. P., and Trinchieri, G. (1995). Lipopolysaccharide-induced interleukin-8 gene expression in human granulocytes: Transcriptiorial inhibition by interferon-gamma. Biochem. J. 310, 751–755(pt. 3). 11. Cassatella, M. A. (1996). Interferon-gamma inhibits the lipopolysaccharide-induced macrophage inflammatory protein-1 alpha gene transcription in human neutrophils. Immunol. Lett. 49, 79–82. 12. Darnell, J. E., Jr. (1997). STATs and gene regulation. Science 277, 1630–1635. 13. Kisseleva, T., Bhattacharya, S., Braunstein, J., and Schindler, C. W. (2002). Signaling through the JAK/STAT pathway, recent advances and future challenges. Gene 285, 1–24. 14. Leonard, W. J., and O’Shea, J. J. (1998). Jaks and STATs: Biological implications. Annu. Rev. Immunol. 16, 293–322. 15. Schindler, C., Fu, X. Y., Improta, T., Aebersold, R., and Darnell, J. E., Jr. (1992). Proteins of transcription factor ISGF-3: One gene encodes the 91- and 84-kDa ISGF-3 proteins that are activated by interferon alpha. Proc. Natl. Acad. Sci. USA 89, 7836–7839. 16. Sugiyama, T., Nishio, Y., Kishimoto, T., and Akira, S. (1996). Identification of alternative splicing form of Stat2. FEBS Lett. 381, 191–194. 17. Schaefer, T. S., Sanders, L. K., and Nathans, D. (1995). Cooperative transcriptional activity of Jun and Stat3 beta, a short form of Stat3. Proc. Natl. Acad. Sci. USA 92, 9097–9101. 18. Caldenhoven, E., van, Dijk T. B., Solari, R., Armstrong, J., Raaijrnakers, J. A., Lammers, J. W., Koenderman, L., and de, Groot R. P. (1996). STAT3beta, a splice variant of transcription factor STAT3, is a dominant negative regulator of transcription. J. Biol. Chem. 271, 13221–13227. 19. Horvath, C. M. (2000). STAT proteins and transcriptional responses to extracellular signals. Trends Biochem. Sci. 25, 496–502. 20. Wang, D., Stravopodis, D., Teglund, S., Kitazawa, J., and Ihle, I. N. (1996). Naturally occurring dominant negative variants of Stat5. Mol. Cell Biol. 16, 6141–6148.
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
35
21. Patel, B. K., Pierce, J. H., and LaRochelle, W. J. (1998). Regulation of interleukin 4-mediated signaling by naturally occurring dominant negative and attenuated forms of human Stat6. Proc. Natl. Acad. Sci. USA 95, 172–177. 22. Azam, M., Lee, C., Strehlow, I., and Schindler, C. (1997). Functionally distinct isoforms of STAT5 are generated by protein processing. Immunity 6, 691–701. 23. Meyer, J., Jucker, M., Ostertag, W., and Stocking, C. (1998). Carboxyl-truncated STAT5beta is generated by a nucleus-associated serine protease in early hematopoietic progenitors. Blood 91, 1901–1908. 24. Chakraborty, A., and Tweardy, D. J. (1998). Granulocyte colony-stimulating factor activates a 72-kDa isoform of STAT3 in human neutrophils. J. Leukoc. Biol. 64, 675–680. 25. Mtiller, M., Laxton, C., Briscoe, J., Schindler, C., Improta, T., Darnell, J. E., Jr., Stark, G. R., and Kerr, I. M. (1993). Complementation of a mutant cell line: Central role of the 91 kDa polypeptide of ISGF3 in the interferon-alpha and -gamma signal transduction pathways. EMBO J. 12, 4221–4228. 26. Mui, A. L., Wakao, H., Kinoshita, T., Kitamura, T., and Miyajima, A. (1996). Suppression of interleukin-3-induced gene expression by a C-terminal truncated Stat5: Role of Stat5 in proliferation. EMBO J. 15, 2425–2433. 27. Moriggl, R., Gouilleux-Gruart, V., Jahne, R., Berchtold, S., Gartmann, C., Liu, X., Hennighausen, L., Sotiropoulos, A., Groner, B., and Gouilleux, F. (1996). Deletion of the carboxylterminal transactivation domain of MGF-Stat5 results in sustained DNA binding and a dominant negative phenotype. Mol. Cell Biol. 16, 5691–5700. 28. Kim, H., and Baumann, H. (1997). The carboxyl-terminal region of STAT3 controls gene induction by the mouse haptoglobin promoter. J. Biol. Chem. 272, 14571–14579. 29. Sherman, M. A., Secor, V. H., and Brown, M. A. (1999). IL-4 preferentially activates a novel STAT6 isoform in mast cells. J. Immunol. 162, 2703–2708. 30. Sasse, J., Hemmann, U., Schwartz, C., Schniertshauer, U., Heesel, B., Landgraf, C., Schneider-Mergener, J., Heinrich, P. C., and Horn, F. (1997). Mutational analysis of acute-phase response factor/Stat3 activation and dimerization. Mol. Cell Biol. 17, 4677–4686. 31. Heinrich, P. C., Behrmann, I., Muller-Newen, G., Schaper, F., and Graeve, L. (1998). Interleukin-6-type cytokine signalling through the gp130/Jak/STAT pathway. Biochem. J. 334, 297–314(pt. 2). 32. Leaman, D. W., Leung, S., Li, X., and Stark, G. R. (1996). Regulation of STAT-dependent pathways by growth factors and cytokines. FASEB J. 10, 1578–1588. 33. Paukku, K., Valgeirsdottir, S., Saharinen, P., Bergman, M., Heldin, C. H., and Silvennoinen, O. (2000). Platelet-derived growth factor (PDGF)-induced activation of signal transducer and activator of transcription (Stat) 5 is mediated by PDGF beta-receptor and is not dependent on c-src, fyn, jak1 or jak2 kinases. Biochem. J. 345, 759–766(pt. 3). 34. Levy, D. E., Kessler, D. S., Pine, R., and Darnell, J. E., Jr. (1989). Cytoplasmic activation of ISGF3, the positive regulator of interferon-alpha-stimulated transcription, reconstituted in vitro. Genes Dev. 3, 1362–1371. 35. Kessler, D. S., Veals, S. A., Fu, X. Y., and Levy, D. E. (1990). Interferon-alpha regulates nuclear translocation and DNA-binding affinity of ISGF3, a multimeric transcriptional activator. Genes Dev. 4, 1753–1765. 36. Qureshi, S. A., Salditt-Georgieff, M., and Darnell, J. E., Jr. (1995). Tyrosine-phosphorylated Stat1 and Stat2 plus a 48-kDa protein all contact DNA in forming interferon-stimulated-gene factor 3. Proc. Natl. Acad. Sci. USA 92, 3829–3833. 37. Paulson, M., Pisharody, S., Pan, L., Guadagno, S., Mui, A. L., and Levy, D. E. (1999). Stat protein transactivation domains recruit p300/CBP through widely divergent sequences. J. Biol. Chem. 274, 25343–25349. 38. Decker, T., and Kovarik, P. (2000). Serine phosphorylation of STATs. Oncogene 19, 2628–2637.
36
PATRICK P. MCDONALD
39. Jiao, H., Berrada, K., Yang, W., Tabrizi, M., Platanias, L. C., and Yi, T. (1996). Direct association with and dephosphorylation of Jak2 kinase by the SH2-domain-containing protein tyrosine phosphatase SHP-1. Mol. Cell Biol. 16, 6985–6992. 40. David, M., Chen, H. E., Goelz, S., Lamer, A. C., and Neel, B. G. (1995). Differential regulation of the alpha/beta interferon-stimulated Jak/Stat pathway by the SH2 domaincontaining tyrosine phosphatase SHPTP1. Mol. Cell Biol. 15, 7050–7058. 41. Aoki, N., and Matsuda, T. (2000). A cytosolic protein-tyrosine phosphatase PTP1B specifically dephosphorylates and deactivates prolactin-activated STAT5a and STAT5b. J. Biol. Chem. 275, 39718–39726. 42. Tanuma, N., Shima, H., Nakamura, K., and Kikuchi, K. (2001). Protein tyrosine phosphatase epsilonC selectively inhibits interleukin-6- and interleukin-10-induced JAK-STAT signaling. Blood 98, 3030–3034. 43. Tanuma, N., Nakamura, K., Shima, H., and Kikuchi, K. (2000). Protein-tyrosine phosphatase PTPepsilon C inhibits Jak-STAT signaling and differentiation induced by interleukin-6 and leukemia inhibitory factor in M1 leukemia cells. J. Biol. Chem. 275, 28216–28221. 44. You, M., Yu, D. H., and Feng, G. S. (1999). Shp-2 tyrosine phosphatase functions as a negative regulator of the interferon-stimulated Jak/STAT pathway. Mol. Cell Biol. 19, 2416–2424. 45. Haspel, R. L., Salditt-Georgieff, M., and Darnell, J. E., Jr. (1996). The rapid inactivation of nuclear tyrosine phosphorylated StatI depends upon a protein tyrosine phosphatase. EMBO J. 15, 6262–6268. 46. Haspel, R. L., and Darnell, J. E., Jr. (1999). A nuclear protein tyrosine phosphatase is required for the inactivation of Stat1. Proc. Natl. Acad. Sci. USA 96, 10188–10193. 47. Naka, T., Narazaki, M., Hirata, M., Matsumoto, T., Minamoto, S., Aono, A., Nishimoto, N., Kajita, T., Taga, T., Yoshizaki, K., Akira, S., and Kishimoto, T. (1997). Structure and function of a new STAT-induced STAT inhibitor. Nature 387, 924–929. 48. Starr, R., Willson, T. A., Viney, E. M., Murray, L. J., Rayner, J. R., Jenkins, B. J., Gonda, T. J., Alexander, W. S., Metcalf, D., Nicola, N. A., and Hilton, D. J. (1997). A family of cytokineinducible inhibitors of signalling. Nature 387, 917–921. 49. Endo, T. A., Masuhara, M., Yokouchi, M., Suzuki, R., Sakamoto, H., Mitsui, K., Matsumoto, A., Tanimura, S., Ohtsubo, M., Misawa, H., Miyazaki, T., Leonor, N., Taniguchi, T., Fujita, T., Kanakura, Y., Komiya, S., and Yoshimura, A. (1997). A new protein containing an SH2 domain that inhibits JAK kinases. Nature 387, 921–924. 50. Krebs, D. L., and Hilton, D. J. (2001). SOCS proteins: Negative regulators of cytokine signaling. Stem Cells 19, 378–387. 51. Yasukawa, H., Sasaki, A., and Yoshimura, A. (2000). Negative regulation of cytokine signaling pathways. Annu. Rev. Immunol. 18, 143–164. 52. Shuai, K. (2000). Modulation of STAT signaling by STAT-interacting proteins. Oncogene 19, 2638–2644. 53. Chung, C. D., Liao, J., Uu, B., Rao, X., Jay, P., Berta, P., and Shuai, K. (1997). Specific inhibition of Stat3 signal transduction by PIAS3. Science 278, 1803–1805. 54. Liu, B., Liao, J., Rao, X., Kushner, S. A., Chung, C. D., Chang, D. D., and Shuai, K. (1998). Inhibition of Stat1-mediated gene activation by PIAS1. Proc. Natl. Acad. Sci. USA 95, 10626– 10631. 55. Tweardy, D. J., Wright, T. M., Ziegler, S. F., Baumann, H., Chakraborty, A., White, S. M., Dyer, K. F., and Rubin, K. A. (1995). Granulocyte colony-stimulating factor rapidly activates a distinct STAT-like protein in normal myeloid cells. Blood 86, 4409–4416. 56. Lieschke, G. J., Grail, D., Hodgson, G., Metcalf, D., Stanley, E., Cheers, C., Fowler, K. J., Basu, S., Zhan, Y. F., and Dunn, A. R. (1994). Mice lacking granulocyte colony-stimulating factor have chronic neutropenia, granulocyte and macrophage progenitor cell deficiency, and impaired neutrophil mobilization. Blood 84, 1737–1746.
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
37
57. Bovolenta, C., Gasperini, S., and Cassatella, M. A. (1996). Granulocyte colony-stimulating factor induces the binding of STAT1 and STAT3 to the IFNgamma response region within the promoter of the Fc(gamma)RI/CD64 gene in human neutrophils. FEBS Lett. 386, 239–242. 58. van de Winkel, J. G., and Capel, P. J. (1993). Human IgG Fc receptor heterogeneity: Molecular aspects and clinical implications. Immunol. Today 14, 215–221. 59. Ravetch, N., and Kinet, J. P. (1991). Fc receptors. Annu. Rev. Immunol. 9, 457–492. 60. Pearse, R. N., Feinman, R., and Ravetch, J. V. (1991). Characterization of the promoter of the human gene encoding the high-affinity IgG receptor: Transcriptional induction by gammainterferon is mediated through common DNA response elements. Proc. Natl. Acad. Sci. USA. 88, 11305–11309. 61. Kuroki, M., and O’Flaherty, J. T. (1999). Extracellular signal-regulated protein kinase (ERK)dependent and ERK-independent pathways target STAT3 on serine-727 in human neutrophils stimulated by chemotactic factors and cytokines. Biochem. J. 341, 691–696(pt. 3). 62. Borregaard, N., Heiple, J. M., Simons, E. R., and Clark, R. A. (1983). Subcellular localization of the b-cytochrome component of the human neutrophil microbicidal oxidase: Translocation during activation. J. Cell Biol. 97, 52–61. 63. Nachman, R., Hirsch, J. G., and Baggiolini, M. (1972). Studies on isolated membranes of azurophil and specific granules from rabbit polymorphonuclear leukocytes. J. Cell Biol. 54, 133–140. 64. McDonald, P. P., Bovolenta, C., and Cassatella, M. A. (1998). Activation of distinct transcription factors in neutrophils by bacterial LPS, interferon-gamma, and GM-CSF and the necessity to overcome the action of endogenous proteases. Biochemistry 37, 13165–13173. 65. Brizzi, M. F., Aronica, M. G., Rosso, A., Bagnara, G. P., Yarden, Y., and Pegoraro, L. (1996). Granulocyte-macrophage colony-stimulating factor stimulates JAK2 signaling pathway and rapidly activates p93fes, STATI p91, and STAT3 p92 in polymorphonuclear leukocytes. J. Biol. Chem. 271, 3562–3567. 66. Al-Shami, A., Mahanna, W., and Naccache, P. H. (1998). Granulocyte-macrophage colonystimulating factor-activated signaling pathways in human neutrophils. Selective activation of Jak2, Stat3, and Stat5b. J. Biol. Chem. 273, 1058–1063. 67. Wilson, K. C., and Finbloom, D. S. (1992). Interferon gamma rapidly induces in human monocytes a DNA-binding factor that recognizes the gamma response region within the promoter of the gene for the high-affinity Fc gamma receptor. Proc. Natl. Acad. Sci. USA 89, 11964–11968. 68. Pearse, R. N., Feinman, R., Shuai, K., Darnell, J. E., Jr., and Ravetch, N. (1993). Interferon gamma-induced transcription of the high-affinity Fc receptor for IgG requires assembly of a complex that includes the 91-kDa subunit of transcription factor ISGF3. Proc. Natl. Acad. Sci. USA 90, 4314–4318. 69. Perez, C., Wietzerbin, J., and Benech, P. D. (1993). Two cis-DNA elements involved in myeloid-cell-specific expression and gamma interferon (IFN-gamma) activation of the human high-affinity Fc gamma receptor gene: A novel IFN regulatory mechanism. Mol. Cell Biol. 13, 2182–2192. 70. Hartner, M., Nielsch, D., Mayr, L. M., Johnston, J. A., Heinrich, P. C., and Haan, S. (2002). Suppressor of cytokine signaling-3 is recruited to the activated granulocyte-colony stimulating factor receptor and modulates its signal transduction. J. Immunol. 169, 1219–1227. 71. Cooney, R. N. (2002). Suppressors of cytokine signaling (SaCS): Inhibitors of the JAK/STAT pathway. Shock 17, 83–90. 72. Fleischmann, J., Golde, D. W., Weisbart, R. H., and Gasson, J. C. (1986). Granulocytemacrophage colony-stimulating factor enhances phagocytosis of bacteria by human neutrophils. Blood 68, 708–711.
38
PATRICK P. MCDONALD
73. Lopez, A. F., Williamson, D. J., Gamble, J. R., Begley, C. G., Harlan, J. M., Klebanoff, S. J., Waltersdorph, A., Wong, G., Clark, S. C., and Vadas, M. A. (1986). Recombinant human granulocyte-macrophage colony-stimulating factor stimulates in vitro mature human neutrophil and eosinophil function, surface receptor expression, and survival. J. Clin. Invest. 78, 1220–1228. 74. Weisbart, R. H., Golde, D. W., and Gasson, J. C. (1986). Biosynthetic human GM-CSF modulates the number and affinity of neutrophil f-Met-Leu-Phe receptors. J. Immunol. 137, 3584–3587. 75. McColl, S. R., Beauseigle, D., Gilbert, C., and Naccache, P. H. (1990). Priming of the human neutrophil respiratory burst by granulocyte-macrophage colony-stimulating factor and tumor necrosis factor-alpha involves regulation at a post-cell surface receptor level. Enhancement of the effect of agents which directly activate G proteins. J. Immunol. 145, 3047–3053. 76. Wirthmueller, D., De Weck, A. L., and Dahinden, C. A. (1989). Platelet-activating factor production in human neutrophils by sequential stimulation with granulocyte-macrophage colony-stimulating factor and the chemotactic factors C5A or formyl-methionyl-leucyl-phenylalanine. J. Immunol. 142, 3213–3218. 77. DiPersio, J. R., Naccache, P. H., Borgeat, P., Gasson, J. C., Nguyen, M. H., and McColl, S. R. (1988). Characterization of the priming effects of human granulocyte-macrophage colonystimulating factor on human neutrophil leukotriene synthesis. Prostaglandins 36, 673–691. 78. McColl, S. R., Krump, E., Naccache, P. H., Poubelle, P. E., Braquet, P., Braquet, M., and Borgeat, P. (1991). Granulocyte-macrophage colony-stimulating factor increases the synthesis of leukotriene B4 by human neutrophils in response to platelet-activating factor. Enhancement of both arachidonic acid availability and 5-lipoxygenase activation. J. Immunol. 146, 1204–1211. 79. Dahinden, C. A., Zingg, J., Maly, F. E., and de, Weck A. L. (1988). Leukotriene production in human neutrophils primed by recombinant human granulocyte/macrophage colony-stimulating factor and stimulated with the complement component C5A and FMLP as second signals. J. Exp. Med. 167, 1281–1295. 80. Neuman, E., Huleatt, J. W., and Jack, R. M. (1990). Granulocyte-macrophage colony-stimulating factor increases synthesis and expression of CR1 and CR3 by human peripheral blood neutrophils. J. Immunol. 145, 3325–3332. 81. Kurt-Jones, E. A., Mandell, L., Whitney, C., Padgett, A., Gosselin, K., Newburger, P. E., and Finberg, R. W. (2002). Role of toll-like receptor 2 (TLR2) in neutrophil activation: GM-CSF enhances TLR2 expression and TLR2-mediated interleukin 8 responses in neutrophils. Blood 100, 1860–1868. 82. Nagase, H., Miyamasu, M., Yamaguchi, M., Imanishi, M., Tsuno, N. H., Matsushima, K., Yamamoto, K., Morita, Y., and Hirai, K. (2002). Cytokine-mediated regulation of CXCR4 expression in human neutrophils. J. Leukoc. Biol. 71, 711–717. 83. Cheng, S. S., Lai, J. J., Lukacs, N. W., and Kunkel, S. L. (2001). Granulocyte-macrophage colony stimulating factor up-regulates CCR1 in human neutrophils. J. Immunol. 166, 1178– 1184. 84. Clark, S. C. (1988). Biological activities of human granulocyte-macrophage colony-stimulating factor. Int. J. Cell Cloning. 6, 365–377. 85. Pouliot, M., McDonald, P. P., Borgeat, P., and McColl, S. R. (1994). Granulocyte/macrophage colony-stimulating factor stimulates the expression of the 5-lipoxygenase-activating protein (FLAP) in human neutrophils. J. Exp. Med. 179, 1225–1232. 86. Gosselin, E. J., Wardwell, K., Rigby, W. F., and Guyre, P. M. (1993). Induction ofMHC class II on human polymorphonuclear neutrophils by granulocyte/macrophage colony-stimulating factor, IFN-gamma, and IL-3. J. Immunol. 151, 1482–1490.
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
39
87. Colotta, F., Re, F., Polentarutti, N., Sozzani, S., and Mantovani, A. (1992). Modulation of granulocyte survival and programmed cell death by cytokines and bacterial products. Blood 80, 2012–2020. 88. Cox, G., Gauldie, J., and Jordana, M. (1992). Bronchial epithelial cell-derived cytokines (GCSF and GM-CSF) promote the survival of peripheral blood neutrophils in vitro. Am J. Respir Cell Mol. Biol. 7, 507–513. 89. Brach, M. A., de Vos, S., Gruss, H. J., and Herrmann, F. (1992). Prolongation of survival of human polymorphonuclear neutrophils by granulocyte-macrophage colony-stimulating factor is caused by inhibition of programmed cell death. Blood 80, 2920–2924. 90. Lee, A., Whyte, M. K., and Haslett, C. (1993). Inhibition of apoptosis and prolongation of neutrophil functional longevity by inflammatory mediators. J. Leukoc. Biol. 54, 283–288. 91. Wen, Z., and Darnell, J. E., Jr. (1997). Mapping of Stat3 serine phosphorylation to a single residue (727) and evidence that serine phosphorylation has no influence on DNA binding of Statl and Stat3. Nucleic Acids Res. 25, 2062–2067. 92. Caldenhoven, E., van Dijk, T. B., Raaijmakers, J. A., Lammers, J. W., Koenderman, L., and de, Groot R. P. (1999). Activation of a functionally distinct 80-kDa STAT5 isoform by IL-5 and GM-CSF in human eosinophils and neutrophils. Mol. Cell Biol. Res Commun. 1, 95–101. 93. Epling-Burnette, P. K., Zhong, B., Bai, F., Jiang, K., Bailey, R. D., Garcia, R., Jove, R., Djeu, J. Y., Loughran, T. P., Jr., and Wei, S. (2001). Cooperative regulation of Mcl-1 by Janus kinase/ stat and phosphatidylinositol 3-kinase contribute to granulocyte-macrophage colonystimulating factor-delayed apoptosis in human neutrophils. J. Immunol. 166, 7486–7495. 94. Bovolenta, C., Gasperini, S., McDonald, P. P., and Cassatella, M. A. (1998). High affinity receptor for IgG (Fc gamma RI/CD64) gene and STAT protein binding to the IFN-gamma response region (GRR) are regulated differentially in human neutrophils and monocytes by IL-10. J. Immunol. 160, 911–919. 95. Rosen, R. L., Winestock, K. D., Chen, G., Liu, X., Hennighausen, L., and Finbloom, D. S. (1996). Granulocyte-macrophage colony-stimulating factor preferentially activates the 94-kD STAT5A and an 80-kD STAT5A isoform in human peripheral blood monocytes. Blood 88, 1206–1214. 96. Moulding, D. A., Quayle, J. A., Hart, C. A., and Edwards, S. W. (1998). Mcl-1 expression in human neutrophils: Regulation by cytokines and correlation with cell survival. Blood 92, 2495– 2502. 97. Moulding, D. A., Akgul, C., Derouet, M., White, M. R., and Edwards, S. W. (2001). BCL-2 family expression in human neutrophils during delayed and accelerated apoptosis. J. Leukoc. Biol. 70, 783–792. 98. Wang, J. M., Chao, J. R., Chen, W., Kuo, M. L., Yen, J. J., and Yang-Yen, H. F. (1999). The antiapoptotic gene mcl-1 is up-regulated by the phosphatidylinositol 3-kinase/Akt signaling pathway through a transcription factor complex containing CREB. Mol. Cell Biol. 19, 6195–6206. 99. Cassatella, M. A., Gasperini, S., Bovolenta, C., Calzetti, F., Vollebregt, M., Scapini, P., Marchi, M., Suzuki, R., Suzuki, A., and Yoshimura, A. (1999). Interleukin-10 (IL-10) selectively enhances CIS3/S0CS3 mRNA expression in human neutrophils: evidence for an IL-10induced pathway that is independent of STAT protein activation. Blood 94, 2880–2889. 100. Matsumoto, A., Masuhara, M., Mitsui, K., Yokouchi, M., Ohtsubo, M., Misawa, H., Miyajima, A., and Yoshimura, A. (1997). CIS, a cytokine inducible SH2 protein, is a target of the JAK-STAT5 pathway and modulates STAT5 activation. Blood 89, 3148–3154. 101. Feldman, G. M., Rosenthal, L. A., Liu, X., Hayes, M. P., Wynshaw-Boris, A., Leonard, W. J., Hennighausen, L., and Finbloom, D. S. (1997). STAT5-deficient mice demonstrate a defect in granulocyte-macrophage colony-stimulating factor-induced proliferation and gene expression. Blood 90, 1768–1776.
40
PATRICK P. MCDONALD
102. Cassatella, M. A., Bazzoni, F., Calzetti, F., Guasparri, I., Rossi, F., and Trinchieri, G. (1991). Interferon-gamma transcriptionally modulates the expression of the genes for the high affinity IgG-Fc receptor and the 47-kDa cytosolic component of NADPH oxidase in human polymorphonuclear leukocytes. J. Biol. Chem. 266, 22079–22082. 103. Huizinga, T. W., Van der Schoot, C. E., Roos, D., and Weening, R. S. (1991). Induction of neutrophil FC-gamma receptor I expression can be used as a marker for biologic activity of recombinant interferon-gamma in vivo. Blood 77, 2088–2090. 104. Perussia, B., Dayton, E. T., Lazarus, R., Fanning, V., and Trinchieri, G. (1983). Immune interferon induces the receptor for monomeric IgG1 on human monocytic and myeloid cells. J. Exp. Med. 158, 1092–1113. 105. Eilers, A., Seegert, D., Schindler, C., Baccarini, M., and Decker, T. (1993). The response of gamma interferon activation factor is under developmental control in cells of the macrophage lineage. Mol. Cell Biol. 13, 3245–3254. 106. Lamer, A. C., David, M., Feldman, G. M., Igarashi, K., Hackett, R. H., Webb, D. S., Sweitzer, S. M., Petricoin, E. F., 3rd, and Finbloom, D. S. (1993). Tyrosine phosphorylation of DNA binding proteins by multiple cytokines. Science 261, 1730–1733. 107. Eilers, A., Baccarini, M., Horn, F., Hipskind, R. A., Schindler, C., and Decker, T. (1994). A factor induced by differentiation signals in cells of the macrophage lineage binds to the gamma interferon activation site. Mol. Cell Biol. 14, 1364–1373. 108. Eilers, A., Georgellis, D., Klose, B., Schindler, C., Ziemiecki, A., Harpur, A. G., Wilks, A. F., and Decker, T. (1995). Differentiation-regulated serine phosphorylation of STAT1 promotes GAF activation in macrophages. Mol. Cell Biol. 15, 3579–3586. 109. Xu, X., Sun, Y. L., and Hoey, T. (1996). Cooperative DNA binding and sequence-selective recognition conferred by the STAT amino-terminal domain. Science 273, 794–797. 110. Vinkemeier, U., Cohen, S. L., Moarefi, I., Chait, B. T., Kuriyan, J., and Darnell, J. E., Jr. (1996). DNA binding of in vitro activated Stat1 alpha, Stat1 beta and truncated Stat1: Interaction between NH2-terminal domains stabilizes binding of two dimers to tandem DNA sites. EMBO J. 15, 5616–5626. 111. Sakamoto, H., Yasukawa, H., Masuhara, M., Tanimura, S., Sasaki, A., Yuge, K., Ohtsubo, M., Ohtsuka, A., Fujita, T., Ohta, T., Furukawa, Y., Iwase, S., Yamada, H., and Yoshimura, A. (1998). A Janus kinase inhibitor, JAB, is an interferon-gamma-inducible gene and confers resistance to interferons. Blood 92, 1668–1676. 112. Saito, H., Morita, Y., Fujimoto, M., Narazaki, M., Naka, T., and Kishimoto, T. (2000). IFN regulatory factor-I-mediated transcriptional activation of mouse STAT-induced STAT inhibitor-1 gene promoter by IFN-gamma. J. Immunol. 164, 5833–5843. 113. Cassatella, M. A., Meda, L., Bonora, S., Ceska, M., and Constantin, G. (1993). Interleukin 10 (IL-10) inhibits the release of pro inflammatory cytokines from human polymorphonuclear leukocytes. Evidence for an autocrine role of tumor necrosis factor and IL-1 beta in mediating the production of IL-8 triggered by lipopolysaccharide. J. Exp. Med. 178, 2207–2211. 114. Cassatella, M. A., Meda, L., Gasperini, S., Calzetti, F., and Bonora, S. (1994). Interleukin 10 (IL-10) upregulates IL-1 receptor antagonist production from lipopolysaccharide-stimulated human polymorphonuclear leukocytes by delaying mRNA degradation. J. Exp. Med. 179, 1695–1699. 115. Kasama, T., Strieter, R. M., Lukacs, N. W., Burdick, M. D., and Kunkel, S. L. (1994). Regulation of neutrophil-derived chemokine expression by IL-10. J. Immunol. 152, 3559–3569. 116. Wang, P., Wu, P., Anthes, J. C., Siegel, M. I., Egan, R. W., and Billah, M. M. (1994). Interleukin-10 inhibits interleukin-8 production in human neutrophils. Blood 83, 2678–2683.
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
41
117. Jenkins, J. K., Malyak, M., and Arend, W. P. (1994). The effects of interleukin-10 on interleukin-1 receptor antagonist and interleukin-1 beta production in human monocytes and neutrophils. Lymphokine Cytokine Res. 13, 47–54. 118. Cassatella, M. A., Meda, L., Gasperini, S., D’Andrea, A., Ma, X., and Trinchieri, G. (1995). Interleukin-12 production by human polymorphonuclear leukocytes. Eur. J. Immunol. 25, 1–5. 119. Marie, C., Pitton, C., Fitting, C., and Cavaillon, J. M. (1996). IL-10 and IL-4 synergize with TNF-alpha to induce IL-lraproduction by human neutrophils. Cytokine 8, 147–151. 120. te, Velde A. A., de, Waal Malefijt R., Huijbens, R. J., de, Vries I. E., and Figdor, C. G. (1992). IL-10 stimulates monocyte Fc gamma R surface expression and cytotoxic activity. Distinct regulation of antibody-dependent cellular cytotoxicity by lPN-gamma, IL-4, and IL-10. J. Immunol. 149, 4048–4052. 121. Lehmann, J., Seegert, D., Strehlow, I., Schindler, C., Lohmann-Matthes, M. L., and Decker, T. (1994). IL-10-induced factors belonging to the p91 family of proteins bind to IFN-gammaresponsive promoter elements. J. Immunol. 153, 165–172. 122. Heijnen, I. A., and Van de Winkel, J. G. (1995). A human Fc gamma RI/CD64 transgenic model for in vivo analysis of (bispecific) antibody therapeutics. J. Hematother. 4, 351–356. 123. Capsoni, F., Minonzio, F., Ongari, A. M., Carbonelli, V., Galli, A., and Zanussi, C. (1997). Interleukin-10 down-regulates oxidative metabolism and antibody-dependent cellular cytotoxicity of human neutrophils. Scand. J. Immunol. 45, 269–275. 124. Crepaldi, L., Gasperini, S., Lapinet, J. A., Calzetti, F., Pinardi, C., Liu, Y., Zurawski, S., de, Waal Malefyt R., Moore, K. W., and Cassatella, M. A. (2001). Up-regulation of IL-1 OR1 expression is required to render human neutrophils fully responsive to IL-10. J. Immunol. 167, 2312–2322. 125. Donnelly, R. P., Dickensheets, H., and Finbloom, D. S. (1999). The interleukin-10 signal transduction pathway and regulation of gene expression in mononuclear phagocytes. J. Interferon Cytokine Res. 19, 563–573. 126. Pallard, C., Gouilleux, F., Charon, M., Groner, B., Gisselbrecht, S., and Dusanter-Fourt, I. (1995). Interleukin-3, erythropoietin, and prolactin activate a STAT5-like factor in lymphoid cells. J. Biol. Chem. 270, 15942–15945. 127. Clevenger, C. V., Freier, D. O., and Kline, J. B. (1998). Prolactin receptor signal transduction in cells of the immune system. J. Endocrinol. 157, 187–197. 128. Dogusan, Z., Hooghe, R., Verdood, P., and Hooghe-Peters, E. L. (2001). Cytokine-like effects of prolactin in human mononuclear and polymorphonuclear leukocytes. J. Neuroimmunol. 120, 58–66. 129. Ryu, H., Lee, J. H., Kim, K. S., Jeong, S. M., Kim, P. H., and Chung, H. T. (2000). Regulation of neutrophil adhesion by pituitary growth hormone accompanies tyrosine phosphorylation of Jak2, p125FAK, and paxillin. J. Immunol. 165, 2116–2123. 130. Ethuin, F., Delarche, C., Benslama, S., Gougerot-Pocidalo, M. A., Jacob, L., and CholletMartin, S. (2001). Interleukin-12 increases interleukin 8 production and release by human polymorphonuclear neutrophils. J. Leukoc. Biol. 70, 439–446. 131. Chen, F. E., and Ghosh, G. (1999). Regulation of DNA binding by Rel/NF-kappaB transcription factors: Structural views. Oncogene 18, 6845–6852. 132. Israel, A. (2000). The IKK complex: An integrator of all signals that activate NF-kappaB? Trends Cell Biol. 10, 129–133. 133. Karin, M. (1999). How NF-kappaB is activated: The role of the IkappaB kinase (IKK) complex. Oncogene 18, 6867–6874. 134. Finco, T. S., and Baldwin, A. S. (1995). Mechanistic aspects of NF-kappa B regulation: The emerging role of phosphorylation and proteolysis. Immunity 3, 263–272.
42
PATRICK P. MCDONALD
135. Zhong, H., Su, Yang H., Erdjument-Bromage, H., Tempst, P., and Ghosh, S. (1997). The transcriptional activity of NF-kappaB is regulated by the IkappaB-associated PKAc subunit through a cyclic AMP-independent mechanism. Cell 89, 413–424. 136. Sizemore, N., Leung, S., and Stark, G. R. (1999). Activation of phosphatidylinositol 3-kinase in response to interleukin-I leads to phosphorylation and activation of the NF-kappaB p65/ RelA subunit. Mol. Cell Biol. 19, 4798–4805. 137. Sun, S. C., Ganchi, P. A., Ballard, D. W., and Greene, W. E. (1993). NF-kappa B controls expression of inhibitor I kappa B alpha: Evidence for an inducible autoregulatory pathway. Science 259, 1912–1915. 138. Brown, K., Park, S., Kanno, T., Franzoso, G., and Siebenlist, U. (1993). Mutual regulation of the transcriptional activator NF-kappa B and its inhibitor, I kappa B-alpha. Proc. Natl. Acad. Sci. USA 90, 2532–2536. 139. Arenzana-Seisdedos, F., Thompson, J., Rodriguez, M. S., Bachelerie, F., Thomas, D., and Hay, R. T. (1995). Inducible nuclear expression of newly synthesized I kappa B alpha negatively regulates DNA-binding and transcriptional activities of NF-kappa B. Mol. Cell Biol. 15, 2689–2696. 140. Pahl, H. L. (1999). Activators and target genes of ReVNF-kappaB transcription factors. Oncogene 18, 6853–6866. 141. Sugita, S., Kohno, T., Yamamoto, K., Imaizumi, Y., Nakajima, H., Ishimaru, T., and Matsuyama, T. (2002). Induction of macrophage-inflammatory protein-3alpha gene expression by TNF-dependent NF-kappaB activation. J. Immunol. 168, 5621–5628. 142. Druker, B. J., Neumann, M., Okuda, K., Franza, B. R., Jr, and Griffin, J. D. (1994). Rel is rapidly tyrosine-phosphorylated following granulocyte-colony stimulating factor treatment of human neutrophils. J. Biol. Chem. 269, 5387–5390. 143. McDonald, P. P., Bald, A., and Cassatella, M. A. (1997). Activation of the NF-kappaB pathway by inflammatory stimuli in human neutrophils. Blood 89, 3421–3433. 144. Vancurova, I., Miskolci, V., and Davidson, D. (2001). NF-kappa B activation in tumor necrosis factor alpha-stimulated neutrophils is mediated by protein kinase Cdelta. Correlation to nuclear Ikappa Balpha. J. Biol. Chem. 276, 19746–19752. 145. Browning, D. D., Pan, Z. K., Prossnitz, E. R., and Ye, R. D. (1997). Cell type- and developmental stage-specific activation of NF-kappaB by fMet-Leu-Phe in myeloid cells. J. Biol. Chem. 272, 7995–8001. 146. Ishii, N., Takeshita, T., Kimura, Y., Tada, K., Kondo, M., Nakamura, M., and Sugamura, K. (1994). Expression of the IL-2 receptor gamma chain on various populations in human peripheral blood. Int. Immunol. 6, 1273–1277. 147. Uu, J. H., Wei, S., Ussery, D., Epling-Burnette, P. K., Leonard, W. J., and Djeu, J. Y. (1994). Expression of interleukin-2 receptor gamma chain on human neutrophils. Blood 84, 3870– 3875. 148. Girard, D., Gosselin, J., Heitz, D., Paquin, R., and Beaulieu, A. D. (1995). Effects of interleukin-2 on gene expression in human neutrophils. Blood 86, 1170–1176. 149. Djeu, J. Y., Liu, J. H., Wei, S., Rui, H., Pearson, C. A., Leonard, W. J., and Blanchard, D. K. (1993). Function associated with IL-2 receptor-beta on human neutrophils. Mechanism of activation of antifungal activity against Candida albicans by IL-2. J. Immunol. 150, 960–970. 150. Schumann, R. R., Nakarai, T., Gruss, H. J., Brach, M. A., von, Arnim U., Kirschning, C., Karawajew, L., Ludwig, W. D., Renauld, J. C., Ritz, J., and Herrmann, F. (1996). Transcript synthesis and surface expression of the interleukin-2 receptor (alpha-, beta-, and gammachain) by normal and malignant myeloid cells. Blood 87, 2419–2427. 151. McDonald, P. P., Russo, M. P., Ferrini, S., and Cassatella, M. A. (1998). Interleukin-15 (IL15) induces NF-kappaB activation and IL-8 production in human neutrophils. Blood 92, 4828–4835.
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
43
152. Bazzoni, F., Cassatella, M. A., Laudanna, C., and Rossi, F. (1991). Phagocytosis of opsonized yeast induces tumor necrosis factor-alpha mRNA accumulation and protein release by human polymorphonuclear leukocytes. J. Leukoc. Biol. 50, 223–228. 153. Bazzoni, F., Cassatella, M. A., Rossi, F., Ceska, M., Dewald, B., and Baggiolini, M. (1991). Phagocytosing neutrophils produce and release high amounts of the neutrophil-activating peptide 1/interleukin 8. J. Exp. Med. 173, 771–774. 154. Takeichi, O., Saito, I., Tsurumachi, T., Saito, T., and Moro, I. (1994). Human polymorphonuclear leukocytes derived from chronically inflamed tissue express inflammatory cytokines in vivo. Cell Immunol. 156, 296–309. 155. McDonald, P. P., and Cassatella, M. A. (1997). Activation of transcription factor NF-kappa B by phagocytic stimuli in human neutrophils. FEBS Lett. 412, 583–586. 156. Wakshull, E., Brunke-Reese, D., Lindermuth, J., Fisette, L., Nathans, R. S., Crowley, J. J., Tufts, J. C., Zimmerman, J., Mackin, W., and Adams, D. S. (1999). PGG-glucan, a soluble beta-(1,3)-glucan, enhances the oxidative burst response, microbicidal activity, and activates an NF-kappa B-like factor in human PMN: Evidence for a glycosphingolipid beta-(1,3)glucan receptor. Immunopharmacology 41, 89–107. 157. Vollebregt, M., Hampton, M. B., and Winterboum, C. C. (1998). Activation of NF-kappaB in human neutrophils during phagocytosis of bacteria independently of oxidant generation. FEBS Lett. 432, 40–44. 158. Sugita, N., Kimura, A., Matsuki, Y., Yamamoto, T., Yoshie, H., and Hara, K. (1998). Activation of transcription factors and IL-8 expression in neutrophils stimulated with lipopolysaccharide from Porphyromonas gingivalis. Inflammation 22, 253–267. 159. Wright, S. D., Ramos, R. A., Tobias, P. S., Ulevitch, R. J., and Mathison, J. C. (1990). CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding protein. Science 249, 1431–1433. 160. Page, S. M., Gleich, G. J., Roebuck, K. A., and Thomas, L. L. (1999). Stimulation of neutrophil interleukin-8 production by eosinophil granule major basic protein. Am. J. Respir. Cell Mol. Biol. 21, 230–237. 161. Vancurova, I., Bellani, P., and Davidson, D. (2001). Activation of nuclear factor-kappaB and its suppression by dexamethasone in polymorphonuclear leukocytes: Newborn versus adult. Pediatr. Res. 49, 257–262. 162. Contrino, J., Krause, P. J., Slover, N., and Kreutzer, D. (1993). Elevated interleukin-1 expression in human neonatal neutrophils. Pediatr. Res. 34, 249–252. 163. Zentay, Z., Sharaf, M., Qadir, M., Drafta, D., and Davidson, D. (1999). Mechanism for dexamethasone inhibition of neutrophil migration upon exposure to lipopolysaccharide in vitro: Role of neutrophil interleukin-8 release. Pediatr. Res. 46, 406–410. 164. Palombella, V. J., Rando, O. J., Goldberg, A. L., and Maniatis, T. (1994). The ubiquitinproteasome pathway is required for processing the NF-kappa B 1 precursor protein and the activation of NF-kappa B. Cell 78, 773–785. 165. Janssen-Heininger, Y. M., Poynter, M. E., and Baeuerle, P. A. (2000). Recent advances towards understanding redox mechanisms in the activation of nuclear factor kappaB. Free Radic. Biol. Med. 28, 1317–1327. 166. Arrigo, A. P. (1999). Gene expression and the thiol redox state. Free Radic. Biol. Med. 27, 936–944. 167. Pietarinen-Runtti, P., Lakari, E., Raivio, K. O., and Kinnula, V. L. (2000). Expression of antioxidant enzymes in human inflammatory cells. Am. J. Physiol. Cell Physiol. 278, C118–125. 168. Carballo, M., Marquez, G., Conde, M., Martin-Nieto, J., Monteseirin, J., Conde, J., Pintado, E., and Sobrino, F. (1999). Characterization of calcineurin in human neutrophils. Inhibitory effect of hydrogen peroxide on its enzyme activity and on NF-kappaB DNA binding. J. Biol. Chem. 274, 93–100.
44
PATRICK P. MCDONALD
169. Schmidt, A., Hennighausen, L., and Siebenlist, U. (1990). Inducible nuclear factor binding to the kappa B elements of the human immunodeficiency virus enhancer in T cells can be blocked by cyclosporin A in a signal-dependent manner. J. Virol. 64, 4037–4041. 170. Frantz, B., Nordby, E. C., Bren, G., Steffan, N., Paya, C. V., Kincaid, R. L., Tocci, M. J., O’Keefe, S. J., and O’Neill, E. A. (1994). Calcineurin acts in synergy with PMA to inactivate I kappa B/MAD3, an inhibitor of NF-kappa B. EMBO J. 13, 861–870. 171. Gualberto, A., Marquez, G., Carballo, M., Youngblood, G. L., Hunt, S. W., 3rd, Baldwin, A. S., and Sobrino, F. (1998). p53 transactivation of the HIV-1 long terminal repeat is blocked by PD 144795, a calcineurin-inhibitor with anti-HIV properties. J. Biol. Chem. 273, 7088–7093. 172. Zouki, C., Jozsef, L., Ouellet, S., Paquette, Y., and Filep, J. G. (2001). Peroxynitrite mediates cytokine-induced IL-8 gene expression and production by human leukocytes. J. Leukoc. Biol. 69, 815–824. 173. Schreck, R., Meier, B., Mannel, D. N., Droge, W., and Baeuerle, P. A. (1992). Dithiocarbamates as potent inhibitors of nuclear factor kappa B activation in intact cells. J. Exp. Med. 175, 1181–1194. 174. Niwa, M., Hara, A., Kanamori, Y., Hatakeyama, D., Saio, M., Takami, T., Matsuno, H., Kozawa, O., and Uematsu, T. (2000). Nuclear factor-kappaB activates dual inhibition sites in the regulation of tumor necrosis factor-alpha-induced neutrophil apoptosis. Eur. J. Pharmacol. 407, 211–219. 175. Stefano, G. B., Hartman, A., Bilfinger, T. V., Magazine, H. I., Liu, Y., Casares, F., and Goligorsky, M. S. (1995). Presence of the mu3 opiate receptor in endothelial cells. Coupling to nitric oxide production and vasodilation. J. Biol. Chem. 270, 30290–30293. 176. Magazine, H. I., Liu, Y., Bilfinger, T. V., Fricchione, G. L., and Stefano, G. B. (1996). Morphine-induced conformational changes in human monocytes, granulocytes, and endothelial cells and in invertebrate immunocytes and microglia are mediated by nitric oxide. J. Immunol. 156, 4845–4850. 177. Matthews, J. R., Botting, C. H., Panico, M., Morris, H. R., and Hay, R. T. (1996). Inhibition of NF-kappaB DNA binding by nitric oxide. Nucleic Acids Res. 24, 2236–2242. 178. Cuthbertson, B. H., Galley, H. F., and Webster, N. R. (1997). Effect of exogenous nitric oxide and superoxide on interleukin-8 from human polymorphonuclear leucocytes. Br. J. Anaesth. 78, 714–717. 179. Marcoli, M., Ricevuti, G., Mazzone, A., Bekkering, M., Lecchini, S., and Frigo, G. M. (1988). Opioid-induced modification of granulocyte function. Int. J. Immunopharmacol. 10, 425–433. 180. Stefano, G. B., Scharrer, B., Smith, E. M., Hughes, T. K., Jr., Magazine, H. I., Bilfinger, T. V., Hartman, A. R., Fricchione, G. L., Liu, Y., and Makman, M. H. (1996). Opioid and opiate immunoregulatory processes. Crit. Rev. Immunol. 16, 109–144. 181. Welters, I. D., Menzebach, A., Goumon, Y., Cadet, P., Menges, T., Hughes, T. K., Hempelmann, G., and Stefano, G. B. (2000). Morphine inhibits NF-kappaB nuclear binding in human neutrophils and monocytes by a nitric oxide-dependent mechanism. Anesthesiology 92, 1677–1684. 182. Akgul, C., Moulding, D. A., and Edwards, S. W. (2001). Molecular control of neutrophil apoptosis. FEBS Lett. 487, 318–322. 183. Karin, M., and Lin, A. (2002). NF-kappaB at the crossroads of life and death. Nat. Immunol. 3, 221–227. 184. Ward, C., Chilvers, E. R., Lawson, M. F., Pryde, J. G., Fujihara, S., Farrow, S. N., Haslett, C., and Rossi, A. G. (1999). NF-kappaB activation is a critical regulator of human granulocyte apoptosis in vitro. J. Biol. Chem. 274, 4309–4318. 185. Pahl, H. L., Krauss, B., Schulze-Osthoff, K., Decker, T., Traenckner, E. B., Vogt, M., Myers, C., Parks, T., Warring, P., Muhlbacher, A., Czernilofsky, A. P., and Baeuerle, P. A. (1996). The
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
186.
187.
188.
189. 190.
191. 192.
193.
194.
195. 196.
197.
198.
199.
200.
201.
45
immunosuppressive fungal metabolite gliotoxin specifically inhibits transcription factor NFkappaB. J. Exp. Med. 183, 1829–1840. Nolan, B., Kim, R., Duffy, A., Sheth, K., De, M., Miller, C., Chari, R., and Bankey, P. (2000). Inhibited neutrophil apoptosis: Proteasome dependent NF-kappaB translocation is required for TRAF-1 synthesis. Shock 14, 290–294. Dunican, A. L., Leuenroth, S. J., Grutkoski, P., Ayala, A., and Simms, H. H. (2000). TNFalpha-induced suppression of PMN apoptosis is mediated through interleukin-8 production. Shock 14, 284–288; discussion 288–289. Ward, C., Dransfield, I., Murray, J., Farrow, S. N., Haslett, C., and Rossi, A. G. (2002). Prostaglandin D2 and its metabolites induce caspase-dependent granulocyte apoptosis that is mediated via inhibition of I kappa B alpha degradation using a peroxisome proliferatoractivated receptor-gamma-independent mechanism. J. Immunol. 168, 6232–6243. Sementchenko, V. I., and Watson, D. K. (2000). Ets target genes: Past, present and future. Oncogene 19, 6533–6548. Klemsz, M. J., McKercher, S. R., Celada, A., Van Beveren, C., and Maki, R. A. (1990). The macrophage and B cell-specific transcription factor PU.1 is related to the Ets oncogene. Cell 61, 113–124. Eisenbeis, C. F., Singh, H., and Storb, U. (1995). Pip, a novel IRF family member, is a lymphoid-specific, PU.1-dependent transcriptional activator. Genes Dev. 9, 1377–1387. Meraro, D., Hashmueli, S., Koren, B., Azriel, A., Oumard, A., Kirchhoff, S., Hauser, H., Nagulapalli, S., Atchison, M. L., and Levi, B. Z. (1999). Protein-protein and DNA-protein interactions affect the activity of lymphoid-specific IFN regulatory factors. J. Immunol. 163, 6468–6478. Marecki, S., Riendeau, C. J., Liang, M. D., and Fenton, M. J. (2001). PUI and multiple IFN regulatory factor proteins synergize to mediate transcriptional activation of the human IL-1 beta gene. J. Immunol. 166, 6829–6838. McKercher, S. R., Torbett, B. E., Anderson, K. L., Henkel, G. W., Vestal, D. J., Baribault, H., Klemsz, M., Feeney, A. J., Wu, G. E., Paige, C. J., and Maki, R. A. (1996). Targeted disruption of the PUI gene results in multiple hematopoietic abnormalities. EMBO J. 15, 5647–5658. Anderson, K. L., Smith, K. A., Conners, K., McKercher, S. R., Maki, R. A., and Torbett, B. E. (1998). Myeloid development is selectively disrupted in PU.1 null mice. Blood 91, 3702–3710. Anderson, K. L., Smith, K. A., Pio, F., Torbett, B. E., and Maki, R. A. (1998). Neutrophils deficient in PUI do not terminally differentiate or become functionally competent. Blood 92, 1576–1585. Chen, H. M., Zhang, P., Voso, M. T., Hohaus, S., Gonzalez, D. A., Glass, C. K., Zhang, D. E., and Tenen, D. G. (1995). Neutrophils and monocytes express high levels of PU1 (Spi-1) but not Spi-B. Blood 85, 2918–2928. Suzuki, S., Kumatori, A., Haagen, I. A., Fujii, Y., Sadat, M. A., Jun, H. L., Tsuji, Y., Roos, D., and Nakamura, M. (1998). PU1 as an essential activator for the expression of gp91 (phox) gene in human peripheral neutrophils, monocytes, and B lymphocytes. Proc. Natl. Acad. Sci. USA 95, 6085–6090. Tsutsumi-Ishii, Y., Hasebe, T., and Nagaoka, I. (2000). Role of CCAAT/enhancer-binding protein site in transcription of human neutrophil peptide-1 and -3 defensin genes. J. Immunol. 164, 3264–3273. Eklund, E. A., Jalava, A., and Kakar, R. (1998). PD.I, interferon regulatory factor 1, and interferon consensus sequence-binding protein cooperate to increase gp91 (phox) expression. J. Biol. Chem. 273, 13957–13965. Eklund, E. A., and Kakar, R. (1999). Recruitment of CREB-binding protein by PU.1, IFNregulatory factor-I, and the IFN consensus sequence-binding protein is necessary for IFNgamma-induced p67phox and gp91phox expression. J. Immunol. 163, 6095–6105.
46
PATRICK P. MCDONALD
202. Friedman, A. D. (2002). Transcriptional regulation of granulocyte and monocyte development. Oncogene 21, 3377–3390. 203. Smith, L. T., Hohaus, S., Gonzalez, D. A., Dziennis, S. E., and Tenen, D. G. (1996). PU.1 (Spi-1) and C/EBP alpha regulate the granulocyte colony-stimulating factor receptor promoter in myeloid cells. Blood 88, 1234–1247. 204. Hohaus, S., Petrovick, M. S., Voso, M. T., Sun, Z., Zhang, D. E., and Tenen, D. G. (1995). PU.1 (Spi-1) and C/EBP alpha regulate expression of the granulocyte-macrophage colonystimulating factor receptor alpha gene. Mol. Cell Biol. 15, 5830–5845. 205. Poli, V. (1998). The role of C/EBP isoforms in the control of inflammatory and native immunity functions. J. Biol. Chem. 273, 29279–29282. 206. Stein, B., and Baldwin, A. S., Jr. (1993). Distinct mechanisms for regulation of the interleukin-8 gene involve synergism and cooperativity between C/EBP and NF-kappa B. Mol. Cell Biol. 13, 7191–7198. 207. Matsusaka, T., Fujikawa, K., Nishio, Y., Mukaida, N., Matsushima, K., Kishimoto, T., and Akira, S. (1993). Transcription factors NF-IL6 and NF-kappa B synergistically activate transcription of the inflammatory cytokines, interleukin 6 and interleukin 8. Proc. Natl. Acad. Sci. USA 90, 10193–10197. 208. Grove, M., and Plumb, M. (1993). C/EBP, NF-kappa B, and c-Ets family members and transcriptional regulation of the cell-specific and inducible macrophage inflammatory protein 1 alpha immediate-early gene. Mol. Cell Biol. 13, 5276–5289. 209. Chang, M. S., McNinch, J., Basu, R., and Simonet, S. (1994). Cloning and characterization of the human neutrophil-activating peptide (ENA-78) gene. J. Biol. Chem. 269, 25277–25282. 210. Wood, L. D., and Richmond, A. (1995). Constitutive and cytokine-induced expression of the melanoma growth stimulatory activity/GRO alpha gene requires both NF-kappa B and novel constitutive factors. J. Biol. Chem. 270, 30619–30626. 211. Descombes, P., Chojkier, M., Lichtsteiner, S., Falvey, E., and Schibler, U. (1990). LAP, a novel member of the C/EBP gene family, encodes a liver-enriched transcriptional activator protein. Genes Dev. 4, 1541–1551. 212. LeClair, K. P., Blanar, M. A., and Sharp, P. A. (1992). The p50 subunit of NF-kappa B associates with the NF-IL6 transcription factor. Proc. Natl. Acad. Sci. USA 89, 8145–8149. 213. Kowenz-Leutz, E., Twamley, G., Ansieau, S., and Leutz, A. (1994). Novel mechanism of C/EBP beta (NF-M) transcriptional control: Activation through derepression. Genes Dev. 8, 2781–2791. 214. Twamley-Stein, G., Kowenz-Leutz, E., Ansieau, S., and Leutz, A. (1996). Regulation of C/EBP beta/NF-M activity by kinase oncogenes. Curr. Top Microb. Immunol. 211, 129–136. 215. Nakajima, T., Kinoshita, S., Sasagawa, T., Sasaki, K., Naruto, M., Kishimoto, T., and Akira, S. (1993). Phosphorylation at threonine-235 by a ras-dependent mitogen-activated protein kinase cascade is essential for transcription factor NF-IL6. Proc. Natl. Acad. Sci. USA 90, 2207–2211. 216. Zhang, D. E., Zhang, P., Wang, N. D., Hetherington, C. J., Darlington, G. J., and Tenen, D. G. (1997). Absence of granulocyte colony-stimulating factor signaling and neutrophil development in CCAAT enhancer binding protein alpha-deficient mice. Proc. Natl. Acad. Sci. USA 94, 569–574. 217. Yamanaka, R., Barlow, C., Lekstrom-Himes, J., Castilla, L. H., Liu, P. P., Eckhaus, M., Decker, T., Wynshaw-Boris, A., and Xanthopoulos, K. G. (1997). Impaired granulopoiesis, myelodysplasia, and early lethality in CCAAT/enhancer binding protein epsilon-deficient mice. Proc. Natl. Acad. Sci. USA 94, 13187–13192. 218. Lekstrom-Himes, J. A. (2001). The role of C/EBP(epsilon) in the terminal stages of granulocyte differentiation. Stem Cells 19, 125–133.
TRANSCRIPTIONAL REGULATION IN NEUTROPHILS
47
219. Screpanti, I., Romani, L., Musiani, P., Modesti, A., Fattori, E., Lazzaro, D., Sellitto, C., Scarpa, S., Bellavia, D., Lattanzio, G., et al. (1995). Lymphoproliferative disorder and imbalanced T-helper response in C/EBP beta-deficient mice. EMBO J. 14, 1932–1941. 220. Tanaka, T., Akira, S., Yoshida, K., Umemoto, M., Yoneda, Y., Shirafuji, N., Fujiwara, H., Suematsu, S., Yoshida, N., and Kishimoto, T. (1995). Targeted disruption of the NF-IL6 gene discloses its essential role in bacteria killing and tumor cytotoxicity by macrophages. Cell 80, 353–361. 221. Tanaka, T., Yoshida, N., Kishimoto, T., and Akira, S. (1997). Defective adipocyte differentiation in mice lacking the C/EBPbeta and/or C/EBPdelta gene. EMBO J. 16, 7432–7443. 222. Wedel, A., Sulski, G., and Ziegler-Heitbrock, H. W. (1996). CCAAT/enhancer binding protein is involved in the expression of the tumor necrosis factor gene in human monocytes. Cytokine 8, 335–341. 223. Pope, R. M., Leutz, A., and Ness, S. A. (1994). C/EBP beta regulation of the tumor necrosis factor alpha gene. J. Clin. Invest. 94, 1449–1455. 224. Shirakawa, F., Saito, K., Bonagura, C. A., Galson, D. L., Fenton, M. J., Webb, A. C., and Auron, P. E. (1993). The human prointerleukin 1 beta gene requires DNA sequences both proximal and distal to the transcription start site for tissue-specific induction. Mol. Cell Biol. 13, 1332–1344. 225. Plevy, S. E., Gemberling, J. H., Hsu, S., Dorner, A. J., and Smale, S. T. (1997). Multiple control elements mediate activation of the murine and human interleukin 12 p40 promoters: Evidence of functional synergy between C/EBP and Rel proteins. Mol. Cell Biol. 17, 4572–4588. 226. Williams, S. C., Du, Y., Schwartz, R. C., Weiler, S. R., Ortiz, M., Keller, J. R., and Johnson, P. F. (1998). C/EBPepsilon is a myeloid-specific activator of cytokine, chemokine, and macrophage-colony-stimulating factor receptor genes. J. Biol. Chem. 273, 13493–13501. 227. Ford, A. M., Bennett, C. A., Healy, L. E., Towatari, M., Greaves, M. F., and Enver, T. (1996). Regulation of the myeloperoxidase enhancer binding proteins Pu1, C-EBP alpha, -beta, and -delta during granulocyte-lineage specification. Proc. Natl. Acad. Sci. USA 93, 10838–10843. 228. Scott, L. M., Civin, C. I., Rorth, P., and Friedman, A. D. (1992). A novel temporal expression pattern of three C/EBP family members in differentiating myelomonocytic cells. Blood 80, 1725–1735. 229. Yamanaka, R., Kim, G. D., Radomska, H. S., Lekstrom-Himes, J., Smith, L. T., Antonson, P., Tenen, D. G., and Xanthopoulos, K. G. (1997). CCAAT/enhancer binding protein epsilon is preferentially up-regulated during granulocytic differentiation and its functional versatility is determined by alternative use of promoters and differential splicing. Proc. Natl. Acad. Sci. USA 94, 6462–6467. 230. Karin, M., Liu, Z., and Zandi, E. (1997). AP-1 function and regulation. Curr. Opin. Cell Biol. 9, 240–246. 231. Stein, B., Baldwin, A. S., Jr., Ballard, D. W., Greene, W. C., Angel, P., and Herrlich, P. (1993). Cross-coupling of the NF-kappa B p65 and Fos/Jun transcription factors produces potentiated biological function. EMBO J. 12, 3879–3891. 232. Hai, T., and Curran, T. (1991). Cross-family dimerization of transcription factors Fos/Jun and ATF/CREB alters DNA binding specificity. Proc. Natl. Acad. Sci. USA 88, 3720–3724. 233. Kerppola, T. K., and Curran, T. (1993). Selective DNA bending by a variety of bZIP proteins. Mol. Cell Biol. 13, 5479–5489. 234. Mechta-Grigoriou, F., Gerald, D., and Yaniv, M. (2001). The mammalian Jun proteins: Redundancy and specificity. Oncogene 20, 2378–2389. 235. Chinenov, Y., and Kerppola, T. K. (2001). Close encounters of many kinds: Fos-Jun interactions that mediate transcription regulatory specificity. Oncogene 20, 2438–2452.
48
PATRICK P. MCDONALD
236. Deng, T., and Karin, M. (1994). c-Fos transcriptional activity stimulated by H-Ras-activated protein kinase distinct from JNK and ERK. Nature 371, 171–175. 237. Smeal, T., Binetruy, B., Mercola, D. A., Birrer, M., and Karin, M. (1991). Oncogenic and transcriptional cooperation with Ha-Ras requires phosphorylation of c-Jun on serines 63 and 73. Nature 354, 494–496. 238. Pulverer, B. J., Kyriakis, J. M., Avruch, J., Nikolakaki, E., and Woodgett, J. R. (2002). Phosphorylation of c-jun mediated by MAP kinases. Nature 353, 670–674. 239. Yazgan, O., and Pfarr, C. M. (2002). Regulation of two JunD isoforms by Jun N-terminal kinases. J. Biol. Chem. 277, 29710–29718. 240. Sung, S. J., Walters, J. A., Hudson, J., and Gimble, J. M. (1991). Tumor necrosis factor-alpha mRNA accumulation in human myelomonocytic cell lines. Role of transcriptional regulation by DNA sequence motifs and mRNA stabilization. J. Immunol. 147, 2047–2054. 241. Hurme, M., and Matikainen, S. (1993). Okadaic acid, a phosphatase inhibitor, enhances the phorbol ester-induced interleukin-1 beta expression via an AP-1-mediated mechanism. Scand. J. Immunol. 38, 570–574. 242. Smith, M. F., Jr., Eidlen, D., Brewer, M. T., Eisenberg, S. P., Arend, W. P., and GutierrezHartmann, A. (1992). Human IL-1 receptor antagonist promoter. Cell type-specific activity and identification of regulatory regions. J. Immunol. 149, 2000–2007. 243. Roebuck, K. A., Rahman, A., Lakshminarayanan, V., Janakidevi, K., and Malik, A. B. (1995). H2O2 and tumor necrosis factor-alpha activate intercellular adhesion molecule 1 (ICAM-1) gene transcription through distinct cis-regulatory elements within the ICAM-1 promoter. J. Biol. Chem. 270, 18966–18974. 244. Okamoto, S., Mukaida, N., Yasumoto, K., Rice, N., Ishikawa, Y., Horiguchi, H., Murakami, S., and Matsushima, K. (1994). The interleukin-8 AP-1 and kappa B-like sites are genetic end targets of FK506-sensitive pathway accompanied by calcium mobilization. J. Biol. Chem. 269, 8582–8589. 245. Kreipe, H., Radzun, H. J., Heidorn, K., Parwaresch, M. R., Verrier, B., and Muller, R. (1986). Lineage-specific expression of c-fos and c-fms in human hematopoietic cells: Discrepancies with the in vitro differentiation of leukemia cells. Differentiation 33, 56–60. 246. Heidorn, K., Kreipe, H., Radzun, H. J., Muller, R., and Parwaresch, M. R. (1987). The protooncogene c-fos is transcriptionally active in normal human granulocytes. Blood 70, 456–459. 247. Mollinedo, F., Vaquerizo, M. J., and Naranjo, J. R. (1991). Expression of c-jun, jun B and jun D proto-oncogenes in human peripheral-blood granulocytes. Biochem. J. 273, 477–479. 248. Bertani, A., Polentarutti, N., Sica, A., Rambaldi, A., Mantovani, A., and Colotta, F. (1989). Expression of c-jun protooncogene in human myelomonocytic cells. Blood 74, 1811–1816. 249. Caldenhoven, E., Buitenhuis, M., van Dijk, T. B., Raaijmakers, J. A., Lammers, J. W., Koenderman, L., and de Groot, R. P. (1999). Lineage-specific activation of STAT3 by interferon-gamma in human neutrophils. J. Leukoc. Biol. 65, 391–396. 250. Watson, R. W., Rotstein, O. D., Parodo, J., Bitar, R., Marshall, J. C., William, R., and Watson, G. (1998). The IL-1 beta-converting enzyme (caspase-1) inhibits apoptosis of inflammatory neutrophils through activation of IL-1 beta. J. Immunol. 161, 957–962.
advances in immunology, vol. 82
Tumor Vaccines FREDA K. STEVENSON, JASON RICE, AND DELIN ZHU Molecular Immunology Group, Tenovus Laboratory, Cancer Sciences Division Southampton University Hospitals Trust, Southampton SO16 6YD, United Kingdom
I. Introduction
Following the development of a vaccine against smallpox by Edward Jenner, the impact of preventive vaccination against infectious diseases on public health has been clearly demonstrated. The incidence of common infections has fallen dramatically, confirming the effectiveness of this strategy of activating immunity against pathogens. For most infections, the goal of vaccination has been to induce an antibody response, which will remove the invading organism before it has had a chance to enter its favored niche. The situation is different for vaccination against cancer where the tumor cells are already in place. The challenge then is to activate immune pathways capable of attacking tumor cells in situ. While this approach of therapeutic vaccination has rarely been used against infection, it was the setting faced by Louis Pasteur in 1885, who successfully vaccinated individuals infected with rabies virus by injecting spinal cord from an infected rabbit. It is also the situation for patients with chronic viral infections such as human immunodeficiency virus (HIV), where not only is virus already in place, but there is accompanying damage to the immune system. Many parallels can be drawn between infection and cancer, the most obvious being that 10–20% of cancers are known to be associated with infectious organisms (Fig. 1). It is clearly possible to vaccinate against those organisms and thereby prevent the associated cancer. The success of this approach is already evident in Taiwan where vaccination of children against hepatitis B is reducing the incidence of hepatocellular carcinoma (Chang et al., 1997). However, most vaccines against hepatitis B involve the surface antigen, which is relatively simple to prepare, and is now available as a recombinant protein. Vaccines against other cancer-associated organisms are less straightforward, although candidates are emerging. A. Passive Immunotherapy A common approach to treatment of infection and cancer lies in the use of passive immunotherapy, with antibody playing an important role in both these settings. The importance of natural passive antibodies, acquired from maternal transfer, in protecting newborn human babies from infection is clear (Zinkernagel, 2001). The use of passively administered immune serum for 49 Copyright 2004, Elsevier Inc. All rights reserved. 0065-2776/04 $35.00
50
FREDA K. STEVENSON ET AL.
Fig 1 Cancers associated with potentially preventable infections.
treatment of infections has a long history, being first developed by Paul Ehrlich for diphtheria and by O. T. Avery for pneumonia. This strategy still has a place for certain infections (Krause, 1999). Treatment of cancer with passive monoclonal antibodies is also finally succeeding, with the therapeutic use of anti-CD20 for patients with lymphoma being a shining example. One reason for the success of anti-CD20 may be that a relatively large dose of antibody was used in the first exploratory tests (Maloney et al., 1994). The importance of administering a large dose of passive antibody, and the requirement for standardization, had been made clear by O. T. Avery in his treatment of pneumonia. Standardization should be more straightforward for monoclonal antibodies, and dose levels remain relevant for cancer. Passive immunotherapy with specific T cells also is gaining a place in the treatment of infectious disease and cancer. For the former, it has been applied mainly to treat immunosuppressed patients at risk of reactivation or infection with Epstein–Barr virus (EBV) or cytomegalovirus (CMV). The clinical setting usually involves patients who have undergone intensive immunosuppressive treatment for leukemia, and who will be receiving an allogeneic transplant from an MHC class I matched donor. In the case of CMV, reactivation can occur during the vulnerable period of immune reconstitution, and adoptive transfer of CD8þ T cells cultured in vitro can provide protection (Riddell and Greenberg, 2000). An alternative approach is to purify the CMV-specific
TUMOR VACCINES
51
CD8þ T cells from the transplant by tetramer selection prior to transfer (Keenan et al., 2001). Similar adoptive immunotherapeutic strategies appear to be effective against EBV, and can suppress growth of EBV-associated posttransplant lymphomas (Rooney et al., 2001). Although the results are encouraging, there are many complex questions surrounding the fate and survival of the transferred effector T cells. A particular question concerns the requirement for CD4þ Th cells to maintain the transferred CD8þ T cell population. Application of adoptive transfer to treat patients with HIV, this time using autologous CD8þ HIV-specific T cell clones expanded in vitro, has revealed a rapid loss of transferred cells (Riddell and Greenberg, 2000). The likely reason for this loss is the deficiency of IL-2-producing recipient CD4þ Th cells, underlining the requirement for this support. Similar questions apply to the passive transfer of T cells aimed against antigens expressed by tumor cells in the recipient. Once again the clinical setting is commonly that of allogeneic transplantation for patients who have undergone high-dose chemotherapy for leukemia. The potential for T cells in the allograft to do more than simply reconstitute immune capacity was revealed by the clinical observation of a graft-versus-leukemia (GvL) effect. The antileukemic activity was particularly evident when recipients in relapse were being treated by donor lymphocyte infusion (DLI). In many cases, DLI led to long-term remission and recovery of the recipients, although the everpresent danger of graft-versus-host (GvH) disease had to be managed. The antigenic target molecules, known as minor histocompatibility antigens, can be expressed at a higher level by the leukemic cells as compared to normal cells, resulting in the desired separation of GvL from GvH. They appear to be polymorphic antigens that differ between donor and recipient and some, such as HA-1 and HA-2, are limited to the hematopoietic system (Goulmy, 1997). There is clear potential for the tactic of transferring specific cellular immunity to patients, in order to combat both infection and cancer. A refinement of the approach is to engineer T cell receptors in the transferred cells to recognize specific target peptides (Stanislawski et al., 2001). Two important general points have been highlighted by the recent observation that transferred CD8þ T cells aimed against melanoma antigens perform optimally both when accompanied by CD4þ T cells and when transfer follows a nonmyeloablative conditioning regimen (Dudley et al., 2002). Creating ‘‘space’’ for transferred cells was also found to be important in a mouse model (Borrello et al., 2000). Cellular therapy, at least for leukemia, has grown out of clinical observation and is now being refined and improved, but it is more technically demanding than antibody therapy and at present not possible to standardize. Although passive immunotherapy can be effective against infection or cancer, monoclonal antibodies alone are unlikely to cure cancer, and the
52
FREDA K. STEVENSON ET AL.
potential for adoptively transferred cells remains undetermined. For antibody, combination with drug treatment is one way of increasing efficacy, and many clinical trials of combinatorial strategies are underway for lymphoma and leukemias (Thieblemont and Coiffier, 2002). However, the attraction of active vaccination, with induction of a variety of immune effector pathways, and a memory response able to continually monitor emergent infection or cancer cell growth is obvious. For infection, this is the way both to prevent invasion and to attack resident organisms. For cancer, residual tumor cells remaining after conventional treatment are the major target. It is a curiosity that even while the strategy of antibody therapy has been expanding, it has been assumed by many of those involved in vaccine design that only CD8þ cytotoxic T cells have relevance for removing tumor cells. Clearly, the effector mechanism depends on the molecular nature of the expressed target antigen, and multiple vaccine designs will be required. Dissection of the effector pathways involved in passive therapy will be helpful in designing vaccines to induce appropriate responses. B. Active Vaccination There are several problems facing the use of vaccination as a strategy to treat cancer. The first is that apart from the cancer-associated viral antigens, most tumor antigens are relatively weak immunogens. Many candidates are, in fact, autoantigens, which may be overexpressed by tumor cells, but are evident only at low levels in normal cells. It is a truism that no effective immunity has been activated spontaneously in a patient with a tumor, although low levels of antibody and T cell responses can be measured in some cases (see later). Induction of a natural immune response to tumor antigens is likely to depend on the site of the tumor (Ochsenbein et al., 2001) and on its ability to evade or suppress immune recognition. It is possible that immune control of tumors does occur in some cases and that we see only the failures of this control. This point was made by Burnett who commented that ‘‘if there were tumor immunity, it would be invisible’’ (Burnett, 1957). In human subjects, it is evident for virus-associated tumors such as EBV-associated posttransplant lymphomas, which develop in patients treated with immunosuppressive drugs that lower levels of antiviral cytotoxic T cells. It has always been more difficult to reveal a role for the immune system in tumors not associated with viral infection, but an intriguing set of observations pointing to immune control has been recently summarized (Darnell and Posner, 2003). The tumors include breast, ovarian, and lung malignancies, which are revealed at an early stage by an accompanying paraneoplastic neurological degeneration due to autoimmune attack. Intriguingly, antibodies and cytotoxic T lymphocytes (CTL) against ectopically expressed neuronal antigens on the tumor cells cause the autoimmune manifestations. Importantly the strength of these responses correlates with the
TUMOR VACCINES
53
suppression of tumor growth, offering direct evidence of naturally occurring successful tumor immunity in humans (Darnell and Posner, 2003). In mice, a role for adaptive immunity is indicated by the increased incidence of cancer in aged Rag-2/ knockout models (Shankaran et al., 2001). However, disentangling the relative abilities of the innate and adaptive immune systems to control tumor growth can be difficult. In the double mutant mouse, Rag-2/ STAT1/, which is also deficient in interferon (IFN) signaling, the incidence of carcinomas is increased, but the types of tumor differ from the single Rag-2/ mutant, indicating an influence of innate immunity, possibly mediated by natural killer (NK) cells (Shankaran et al., 2001). Certainly some tumors can downregulate MHC class I molecules, or express stress-associated molecules such as MHC class I chain-related proteins A and B (MICA/B) (human), or Rae1 family members (mouse), which should be recognized by NKG2D-activating receptors on NK or gd T cells (Diefenbach and Raulet, 2002). Knockout mice are a powerful resource to dissect out the role of molecules involved in innate or adaptive immunity in controlling tumor development. However, caution is required before extrapolating to human cancer. Many mouse strains and most murine tumor lines carry retroviruses, and retroviral proteins can act as strong tumor-associated antigens (Huang et al., 1996). Involvement of innate and adaptive immunity in controlling these tumors may differ from immune control of spontaneous human cancer, which, as far as we understand at present, is not commonly associated with retroviruses (Nelson et al., 2003). A further problem for vaccination against cancer is that patients may have a reduced immune capacity, due either to disease or to chemotherapy/radiotherapy. Increasing age per se is associated with a fall in immune capacity (Ershler, 1993), and surgery can induce a transient immunosuppression (Faist et al., 1986; Hensler et al., 1997). Hematological malignancies are a special case since they occupy the immune system and can cause damage. The level of damage from the tumor varies according to its nature and duration. The reduction in normal serum immunoglobulin observed in untreated chronic lymphocytic leukemia (CLL) and myeloma is clear evidence of disease-related effects on antibody production. In CLL this may be due in part to production of growth inhibitory factors such as transforming growth factor-b (TGF-b) by the tumor cells (Lotz et al., 1994). Similar cytokine-mediated effects may operate in myeloma, but patients with non-Hodgkin’s lymphoma appear less immunosuppressed unless treated (King et al., 1979). New treatments of B cell malignancies with antibodies such as anti-CD20, while capable of removing tumor cells, are also likely to affect immune status. Other antibodies such as CamPath1 are even more immunosuppressive (Rebello et al., 2001). Interestingly, anti-CD20 apparently does not affect serum immunoglobulin (lg) levels, but does impair secondary humoral immune responses (van der Kolk et al., 2002). One question for vaccination concerns the period of immune deficit following treatment, and
54
FREDA K. STEVENSON ET AL.
whether recovery precedes relapse. The fact that anti-CD20 is now being used in combination with chemotherapy makes assessments of its effect on immunity more difficult. For hematological malignancies, the success of autologous and allogeneic transplantation has facilitated the use of more immunosuppressive regimens in patients. Because the recipients are highly vulnerable to infection after conditioning treatment, there has been concern about the rate of reconstitution of immunity (Singhal and Mehta, 1999), and this has relevance for vaccination as a strategy to prevent resurgence of tumor. Based largely on recovery of antibody responses, the recommended time to vaccinate against infectious organisms in children is 1 year posttransplantation (Singhal and Mehta, 1999). However, investigations in mice suggest that vaccination should be early in order to capture the expanding T cells that are filling the space created by chemotherapy (Borrello et al., 2000). For patients, the level of immunity in the donor, the cell populations transferred, and the conditioning used all influence the rate of recovery of responsiveness. Objective criteria of immune status would be useful to inform the selection of cancer patients for vaccination protocols. However, studies have been generally limited to the practical clinical questions of how to ensure that patients are protected against infection. More subtle questions on the ability of patients to mount CD4þ or CD8þ T cell responses against specific antigen are rarely posed. Perhaps that is not surprising, given that there is no bedrock of information on immune responses in normal human subjects for comparison. In most studies of conventional vaccination against infectious diseases, the effect on the incidence of infection is the primary measure. Since antibody is important in protection, assessment of efficacy has also centered on serum antibody levels, commonly taken at 4–6 weeks following vaccination. There have been few studies of cellular responses in normal adults or in children undergoing routine vaccination, partly because of the technical difficulties involved, but also because the relevance was unclear. The paucity of information on the nature and kinetics of T cell responses in normal subjects undermines our ability to select cancer patients for vaccination. It also creates difficulties for assessing performance of new vaccination protocols. Again, some information is available from vaccination against infectious disease, with one example being clinical testing of a new DNA vaccine against hepatitis B. A surprisingly slow development of immune responses over a period of up to a year was observed (Roy et al., 2000). Whether this was due to the mode of delivery, or reflected a feature of human immune responses remains unclear. However, the fact that comparably slow responses are being observed in monkeys and in clinical trials with a variety of vaccines (Robinson and Pertmer, 2000), suggests that this may be the usual response to vaccination in primates.
TUMOR VACCINES
55
C. Tolerance and Evasion Another potential problem for vaccination against cancer is that of tolerance or evasion of immunity. While the two terms have distinct meanings to immunologists, both processes lead to the same outcome of failure of immune cells to clear tumor. This failure can contribute to tumor progression by allowing local invasion and metastasis. The mechanisms involved in mediating tolerance and evasion are complex and overlapping. Central tolerance reflects deletional loss of responding T cells in the thymus, but in adult mice the efficiency of deletion depends on the avidity of T cells, and may not be complete. Partial tolerance to disparate skin grafts, induced by intrathymic injection of peptide, can, however, be maintained by CD25þ , CD4þ immunoregulatory T cells, demonstrating a multiplicity of tolerizing processes (Trani et al., 2003). Apart from the cancer-associated viral antigens, there are few tumor antigens that can be considered to be new to the immune system. For many, tolerance induction in the thymus may have occurred prior to cancer development. Proteins arising only in the cancer cell include idiotypic determinants expressed by B cell or T cell tumors, mutated protooncogene products, splice variants, and novel peptides generated by chromosomal translocations. Antigens with expression restricted to the testis and the cancer cell, the so-called cancer testis antigens should also have not been exposed to early thymic tolerogenic processes. However, expression by tumor cells during the course of the disease could lead to purging of specific T cells in the residual thymus or in the periphery. This could occur by clonal exhaustion of effector cells similar to that induced by infection with noncytopathic viruses, retroviral superantigens, or bacterial superantigens. The same process may also be involved in tolerance to transplantation antigens (Starzl and Zinkernagel, 1998). For cancer, an important question is whether there has been engagement between the tumor cells and the immune system. If the cancer is localized, and tumor cells do not express immune recognition molecular patterns or stress signals, they may not attract the attention of patrolling dendritic cells, and will therefore fail to reach the lymphoid tissue (Starzl and Zinkernagel, 1998). Under these circumstances, it is unlikely that naı¨ve T cells will be activated. Consequently, tolerance will not result and the T cell repertoire will be retained. Unless this setting is perturbed either by progressive growth, outstripping oxygen or nutrients, or by surgical or radio/chemotherapeutic intervention, these tumors may remain invisible. If so, they could represent the best targets for vaccination. Hematological malignancies occupy the lymphoid system and are therefore exposed, but, even here unless stress proteins or other alerting proteins are expressed and recognized, tumor cells may not be sensed by dendritic cells. For all tumors, changes are likely during progression, and genetic dysregulation should expose new antigenic
56
FREDA K. STEVENSON ET AL.
targets. The question is whether these are naturally immunogenic, requiring tumor cells to escape from immunosurveillance (Khong and Restifo, 2002). If tolerance does occur, it may involve deletion, anergy, or regulation of responding T cells (Goodnow, 1997). Transgenic (tg) mouse models are vital in revealing the process of tolerance induction, although each model is necessarily limited to analysis of the fate of a specific Tcell receptor. In a tg mouse expressing antiidiotypic CD4þ T cells, thymic deletion was evident when the level of serum idiotypic immunoglobulin (Id-Ig) was raised to 300 mg/ml (Bogen et al., 1993). It could be concluded from this that patients with myeloma accompanied by a high level of Id-Ig would have deleted the T cell repertoire. However, the data apply only to a single T cell receptor (TCR) of a particular avidity, and leave open the question of the fate of T cells of lower avidity that could be enrolled following vaccination (Bogen, 1996; Bogen et al., 1993). In fact, considerations of tolerance have not stopped the testing of the ability of Id-Ig to activate immunity in patients with myeloma, and T cell responses are being reported (Li et al., 2000; Osterborg et al., 1998). The same is true for follicular lymphoma, where again it might be expected that tumor cells could have deleted the T cell repertoire. In clinical trials of these patients, vaccination with Id-Ig linked to keyhole limpet hemocyanin (KLH) was able to induce anti-Id antibody and proliferative T cells, consistent with retention or regeneration of Id-specific T cells (Hsu et al., 1997). It is certainly possible that tumor in contact with immune cells can prevent induction of immunity via multiple strategies, including low expression of costimulatory molecules, tumor-induced dysfunction of antigen-presenting cells, and secretion of immunosuppressive cytokines (Khong and Restifo, 2002). These blocks could be circumvented by a range of vaccination strategies, particularly if applied in a setting of minimal residual disease. Once immune effector mechanisms are induced, the next problem is that tumor cells may also evade attack. There is highly suggestive evidence that some of the apparent evasion of tumor-infiltrating T cells may simply be due to the fact that there are insufficient numbers of effector T cells (Perez-Diez et al., 2002). This is reminiscent of the situation for antibody therapy where critical levels have to be attained to successfully attack either pathogens or tumor cells (PerezDiez et al., 2002). Tumor cells are dynamic targets, and are capable of evasion, especially if they are given time during the course of development in the host. This is an argument for early treatment, which requires improved methods for detection, together with an increased understanding of tumor behavior. Evasion strategies include down-regulation of MHC class I molecules (Algarra et al., 1997), expression of receptors such as CD94/NKG2A able to inhibit attack by NK or CD8þ cells (Lukacher, 2002), expression of FasL, production of immunosuppressive cytokines, and antiapoptotic mechanisms (reviewed in Khong and Restifo, 2002). Interestingly, NKG2A receptors have been detected in patients on a high proportion of CD8þ T cells infiltrating melanomas and astrocytomas
TUMOR VACCINES
57
(Perrin et al., 2002; Vetter et al., 2000). For tumors associated with g-herpes viruses, such as EBV, a role for viral proteins in tumor development, or in evasion of immune recognition, is strongly indicated (Rickinson and Kieff, 1996). Whether viral immune evasion molecules or cellular analogues contribute to tumor escape from immune effector pathways is not known. However, we showed that introduction of a viral immune evasion gene coding for a chemokine receptor homologue into tumor cells blocked attack by specific CTL in vivo (Rice et al., 2002b). Not only does this illustrate the requirements for CTL movement to tumor sites, but it does point to potential evasion tactics in tumors. There may also be a role for regulatory T cells in suppressing natural or induced immunity against tumor cells. The concept of ‘‘suppressor’’ cells has been resuscitated with the observation that a subset of CD4þ 25þ T cells is involved in maintaining immune homeostasis, and in protecting the host against autoimmune disease (Shevach, 2001). This poorly proliferative population constitutively expresses CD45RO and CD152 (CTLA4), inhibits proliferation of CD4þ 25 and CD8þ T cells, and secretes TGF-b and interleukin (IL)-10 (Shevach, 2001). Several subsets of regulatory T cells with different phenotypes and properties are now being defined (Weiner, 2001). There is evidence in mouse models that regulatory T cells are involved in suppression of natural immunity against transplanted tumors (Golgher et al., 2002; Shimizu et al., 1999; Sutmuller et al., 2001). While their role in modulating tumor immunity in mice is complicated by the fact that natural immunity may be directed against retroviral antigens, the reality of involvement of regulatory T cells in cancer is confirmed by human studies. The prevalence of CD4þ 25þ T cells with suppressor function was found to be increased in the blood of patients with pancreatic and breast adenocarcinoma, and these cells were also present in the tumor microenvironment (Liyanage et al., 2002). Similarly increased numbers of functional CD4þ 25þ T cells have been detected in the blood of patients with other epithelial cancers (Wolf et al., 2003). In one case, it was possible to show that regulatory T cells could inhibit growth of an autologous cytotoxic T cell line specific for tumor cells (Somasundaram et al., 2002). It is interesting that human primary and memory CTL responses can be induced in vitro using either dendritic cells or activated B cells loaded with peptide (von Bergwelt-Baildon et al., 2002). Successful generation of CTL responses against autoantigens such as telomerase suggests that deletional tolerance is not complete. Perhaps surprisingly, the avidity of the CTL was sufficient to kill target telomerase-expressing tumor cells. It appears therefore that conditions in vitro have allowed expansion of the CTL, and it remains unclear which regulatory processes were involved in controlling these cells in vivo. We are now in a position to exploit our knowledge of immunology, and our command of molecular technology, to design strategies to engage the powerful immune system against cancer. As has been found for drug therapy, there will
58
FREDA K. STEVENSON ET AL.
be a variety of approaches, depending on the nature and behavior of the cancer. Encouragingly, the immune repertoire appears to be available or restorable, and our task is to deliver tumor antigens in a form that will induce effector pathways able to recognize tumor cells without significant collateral damage to normal cells. The next sections will describe selected strategies that focus on defined target molecules. We will not discuss in any detail the alternative approach of activating immunity against unknown tumor antigens by manipulating tumor cells or by using whole cell extracts, since that has been described recently elsewhere (Rivoltini et al., 2002). Sections will be concerned with methods of delivering candidate antigens and the outcome, using examples taken from different tumors. Although mouse models will provide the rational basis for vaccine designs, we are conscious of the fact that translation to the clinic has to cross unknown territory. Data from pilot clinical trials are therefore of immense importance and will be introduced where possible. II. Tumor Antigens
Tumor antigens include proteins, commonly glycoproteins, peptides, carbohydrates, and glycolipids, expressed by tumor cells and able to act as targets for immune effector mechanisms. Carbohydrate and lipid antigens provide important specialized targets on certain cancers, and have been reviewed elsewhere (Bitton et al., 2002). This review will focus mainly on (glyco)protein and peptide targets. Apart from virus-associated tumor antigens, descriptions of tumor antigens sometimes refer only to expression, without comment on immunogenicity. However, in many cases, immunogenicity has been tested by the detection of naturally occurring antibodies or by natural or expanded T cell responses in patients, indicative of at least some immune recognition (Rosenberg, 1998). The presence of serum antibodies in patients against potential tumor antigens has led to the technology of SEREX (serological identification of antigens by recombinant expression cloning) whereby target antigens are identified by reacting the sera with cDNA libraries derived from tumor cells (Preuss et al., 2002). For antigens recognized by T cells, isolation and cloning of tumor-infiltrating T cells has facilitated identification of expressed tumor-derived peptides that have apparently induced a local response (Rosenberg, 2001). Glycoprotein antigens may be expressed at the surface of the tumor cell, or be secreted, while the peptides will be bound to either the MHC class I or class II molecules. For certain molecules, such as idiotypic Ig, all three ways of expression may be possible. Each molecular form will require a different mode of attack, with surface glycoproteins potentially accessible to antibody, and peptides to T cell attack. Within these three categories, there is a range of specificity of expression, from those detected only on tumor cells to those that are also expressed by normal cells, but that may be up-regulated in tumor cells.
TUMOR VACCINES
59
One promising category of antigen arises from products of specialized cells, with melanocytes being a good example. Tumor cells of melanoma preserve expression of a range of melanocyte lineage proteins, including tyrosinase, gp100, MART-1/Melan-A, TRP-1 (gp75), and TRP-2 (reviewed in Van den Eynde and van der Bruggen, 1997; Rosenberg, 1999). Lineage-specific proteins are also expressed by tumor cells of prostate, breast, and other cancers. These proteins can be targeted by vaccination with minimal side effects anticipated. For melanoma, vaccines successful in inducing antitumor immunity have been noted also to induce vitiligo, indicative of an attack on normal melanocytes, which may be tolerable for patients (Rosenberg et al., 1996). Hematological malignancies express clonally restricted products of the specialized cells of origin, including idiotypic Ig or clonal T cell receptors (George and Stevenson, 1989; Stevenson et al., 1990). These are perfect tumor-specific antigens with no autoimmune outcome expected. Partly for this reason, the Id-Ig of B cell malignancies has been a focus of investigation of immunotherapeutic strategies (Stevenson et al., 1990). However, a disadvantage of clone-specific targets is that vaccines have to be individual for each patient. Cells of hematological malignancies have the advantage that they are generally available for study from blood or accessible tissues. In addition, the clinical success of allogeneic stem cell transplantation for leukemia has shown that the tumor cells are susceptible to attack by passively administered T cells specific for target peptides (see earlier). Candidate tumor antigens expressed by hematological malignancies, several of which are also expressed by other nonhematological tumors, are illustrated in Fig. 2. Cell surface glycoproteins potentially accessible to antibody attack include clone-specific idiotypic Ig of B cell neoplasms and clonotypic T cell receptors of T cell tumors. Mucins and differentiation antigens such as CD20 are also possible targets for induced antibodies. Cell surface glycoproteins may also be expressed as MHC-associated peptides and therefore could also be susceptible to T cell attack. Proteins confined to the intracellular compartment can be attacked only by this route, and these include a large number of candidates (Fig. 2). For B cell malignancies, MHC Class IIassociated peptides could be targets for direct attack by CD4þ T cells. Peptides derived from mutated protooncogenes, such as p53 (Vierboom et al., 1997), are mainly clone specific raising the possibility that patient-specific vaccines may be required. For the vast array of peptide antigens expressed in the context of MHC class I molecules, there is the complication of haplotype specificity, with most knowledge of human peptide targets being derived from the relatively common HLA-A*0201 allele. Prediction of the ability of peptide sequences to bind to human or murine MHC class I molecules is derived from algorithms based on anchor residues and known peptide sequences (Rammensee et al., 1999). Certain tumor antigens may be expressed by groups of patients who share MHC class I haplotypes, with one example being the chimeric gene,
60
FREDA K. STEVENSON ET AL.
Fig 2 Target tumor antigens of hematological malignancies.
BCR-ABL, generated by the reciprocal translocation t(9;22)(q34;q11) and characteristic of patients with chronic myeloid leukemia (CML) (reviewed in Melo, 1996). For this junctional peptide, several restriction elements have been identified, and CTL able to kill tumor cells have been detected in vivo and expanded in vitro (Yotnda et al., 1998). CML is also susceptible to infused donor lymphocytes via a GvL effect (Kolb et al., 1990, 1995), directed against minor histocompatibility antigens (Fig. 2). Additional target peptides for CTL attack are likely to be identified by the extensive analysis of gene expression using microarray technology (Staudt and Brown, 2000). It is already clear that tumor cells, particularly at late stage, have genomic instability and this may generate new candidate antigens. The search for a ‘‘universal’’ tumor antigen is intense, since the ability to target a single protein expressed by a wide range of tumors would be very attractive. Cancer-testis antigens approach this category. Although they were first identified in melanoma (Boon et al., 1994), some at least have also been detected in other tumors, including hematological malignancies (van Baren et al., 1999). The relative tumor specificity arises from the fact that they are expressed only by normal testis or placenta. They include the MAGE, BAGE, GAGE, RAGE, and LAGE families, NY-ESO-1 and PRAME, and their properties and patterns of expression in melanoma and other tumors have been widely reviewed (Boon et al., 1994; Van den Eynde and van der Bruggen, 1997; Wang and Rosenberg, 1999). There has been considerable success in identifying peptides associated
TUMOR VACCINES
61
with several human MHC class I haplotypes, and clinical trials of vaccines against them, mainly for melanoma, are in progress (see later). Another tumor antigen with restricted expression in normal tissues is the mucin, MUC1, a transmembrane type I molecule found on almost all human epithelial adenocarcinomas, including breast, pancreas, ovary, lung, urinary bladder, prostate, and endometrium (Finn et al., 1995). Tumor cells express high levels of MUC1, with changes in glycosylation patterns that reveal core protein epitopes, masked in normal mucin. MUC1 is also expressed by myeloma cells, with levels apparently up-regulated by certain drug therapies (Treon et al., 1999), opening the possibility of enhancing the efficacy of immune attack. Cytotoxic T cells specific for the mucin have been detected in cancer patients, and can apparently act in an MHC-unrestricted manner (Magarian-Blander et al., 1998). There have been several attempts to develop vaccines against MUC-1, using a variety of vaccine vehicles (see later). Carcinoembryonic antigen (CEA), a molecule involved in cell adhesion, and related molecules also represent widely expressed potential tumor antigenic targets. CEA is a 180-kD glycoprotein expressed via a glycosylphosphatidylinositol linkage on the surface of adenocarcinomas of colon, breast, lung, and other epithelial cancers. There are suggestive data pointing to expression of CEA on a proportion of tumor cells in patients with chronic lymphocytic leukemia (Penezina et al., 1998). CEA represents an example of the category of oncofetal antigens (Hammarstrom, 1999), and antiCEA monoclonal antibodies (MAbs) and derivatives have been widely used in imaging and treatment (Murray and Unger, 1988). Vaccines containing CEA in a range of delivery systems have been tested in preclinical models and in clinical trials (see later) (Berinstein, 2002). The final category of tumor antigen includes proteins that are expressed by normal cells, but that may be overexpressed by tumor cells. A large proportion of potential tumor antigens lies in this group, and, if powerful vaccines are able to break tolerance and induce effective immunity, care will need to be taken to exclude damaging autoimmunity. They include antiapoptotic proteins such as survivin (Altieri, 2001, 2003), transcription factors such as WTI (Huang et al., 1990; Inoue et al., 1997), and telomerase reverse transcriptase (hTERT) (Vonderheide, 2002), all of which are expressed by both hematological and other tumors.
III. Vaccination Strategies
A. Proteins and Glycoproteins The widely held view that tumor cells can be attacked most efficiently by CTL (Melief and Kast, 1995), and that antibody is not only useless, but may in fact inhibit T cell attack (Prehn, 1994), has diverted attention away from using proteins as vaccines against tumors. The exceptions to this view include
62
FREDA K. STEVENSON ET AL.
cell-surface carbohydrate antigens, such as TF, Tn, STn, and globo H antigens expressed in breast cancer, and glycolipids and gangliosides preferentially expressed on the cell sufaces of a range of cancers (Bitton et al., 2002). Exogenous protein or carbohydrate vaccines would be expected to induce antibody responses rather than CTL, and this may be appropriate for cell surface antigenic targets. However, CD4þ T cells can be induced by exogenous protein vaccines, and diversion of the response to Th1 or CD8þ T cell responses may be achievable using appropriate adjuvants. Clearly, CTL have immense potential to attack cancer cells, but the ability of passive monoclonal antibodies to clear malignant B cells points to a role for antibody against certain target antigens. One of these is HER2 (also known as Neu, ErbB2), a member of the epidermal growth factor receptor family of receptor tyrosine kinases, which is overexpressed by a subset of breast cancer. A monoclonal antibody, Herceptin, is currently showing effectiveness in combination treatment (Thomssen, 2001). A second example is Id-Ig expressed by B cell tumors, where protein vaccines, commonly linked to KLH, can induce protective and therapeutic immunity (Hsu et al., 1997; Stevenson et al., 1990). In this case, protection appears to be mediated largely by antibody, and vaccine efficacy parallels the success of passively administered monoclonal anti-Id antibody. As for all antibodies, the mechanism of attack may be multiple and remains unclear. However, there is intriguing evidence for a direct inhibitory signal via the B cell receptor (Uhr et al., 1996; Vuist et al., 1994). Another more recently raised possibility for follicular lymphoma is that anti-Id antibody may dislodge tumor cells from the site in the germinal center where maintenance signaling via the B cell receptor is occurring (Zhu et al., 2002). The finding that specific protective immunity against a T cell lymphoma is mediated by anticlonotypic antibodies against the T cell receptor indicates that Id Ig is not exceptional in being susceptible to antibody attack (Thirdborough et al., 2002). Adjuvants are of obvious importance for protein antigens, and there are several safe choices available of variable efficacy (Hunter, 2002; Schijns, 2001). It was always frustrating that the powerful immunostimulatory effect of complete Freund’s adjuvant (CFA) could not be easily harnessed for human use due to the accompanying damaging inflammatory response. More recently it has become clear that the major active component of CFA is bacterial DNA (Krieg, 2002). Within the DNA sequence there is a prevalence of unmethylated CpG motifs, which activate innate immunity in mammalian hosts. These findings have opened a new area of research into the potential of CpG as an adjuvant for vaccines against infectious diseases or cancer and have been well reviewed (Krieg, 2002). Most investigations are centered on using synthetic stabilized phosphorothioate oligonucleotides, which may behave differently from the natural diester-linked oligonucleotides. Studies of the influence of surrounding base sequence patterns have revealed
TUMOR VACCINES
63
differences in activity with CpG-A motifs activating NK cells while CpG-B are potent B cell activators (Krieg, 2002). The cellular receptor for CpG appears to be the TLR-9 expressed by most mouse dendritic cells (DCs), but in humans expressed only by plasmacytoid DC (Krug et al., 2001). Mouse DCs are therefore directly activated by CpG to produce an array of cytokines, including TNFa (Sparwasser et al., 1998). CpG treatment of human blood secondarily activates monocytes to produce IL-6 and TNF-a, and CpG acts as a stimulatory or costimulatory molecule for a wide range of cells, including NK, B, and T cells (Wagner, 1999). Production of IFN-g and IL12 tends to drive induction of Th1 CD4þ T cells. CpG can promote antibody and CTL responses against diverse antigens in human and in mouse models, with the exception of polysaccharide antigens (Krieg and Davis, 2001). When compared with CFA for the ability to induce a response against a lymphoma Id-Ig, CpG performed at least as well (Weiner et al., 1997). Interestingly, CpG tends to induce higher levels of IgG2a than CFA, consistent with the activation of Th1 CD4þ T cells. In terms of virus-associated cancer, CpG was able to activate immunity against E7 protein from human papilloma virus (HPV), and this provided protection against challenge with HPV-immortalized tumor cells (Kim et al., 2002). A possible role for alternative adjuvant molecules in redirecting induced immune effector function has also been suggested by the data from a clinical trial of patients with follicular lymphoma given Id-Ig KLH protein vaccines delivered with local subcutaneous injection of granulocyte macrophage colony-stimulating factor (GM-CSF) (Bendandi et al., 1999). GM-CSF has multiple effects on hematopoietic and nonhematopoietic cells, with one important function being to stimulate differentiation and outgrowth of DCs. In this trial, patients developed T cell responses, both CD4þ and CD8þ , which appeared to be Id specific, and immune responses were accompanied by eradication of minimal residual disease (Bendandi et al., 1999). The recognition of the pivotal role of DCs in taking up antigen and presenting it to naı¨ve T cells has led to attempts to improve performance of protein vaccines by direct loading of DCs in vitro, followed by maturation and reinfusion (Dhodapkar et al., 1999; Inaba et al., 1998). The complexity and heterogeneity of DCs both within one species and between species make modeling of DCbased vaccine strategies difficult, and argues for testing in pilot clinical trials. Trials of autologous DCs loaded with Id-Ig or Id-Ig KLH have been carried out in patients with lymphoma and myeloma, with evidence for modest beneficial immune responses (Timmerman et al., 2002). A similar approach has been used to vaccinate patients with myeloma, a considerably more challenging disease since patients have high levels of circulating Id-Ig and are immunosuppressed by disease and treatment. Responses as measured by Id-specific T cell proliferation varied from a low (15% of patients) (Liso
64
FREDA K. STEVENSON ET AL.
et al., 2000) to a high incidence (83%). A more optimistic setting for myeloma is to vaccinate the donors of allogeneic stem cell transplants and then to transfer immune cells. This was carried out initially using an Id-Ig KLH vaccine with demonstration of transfer of Id-specific T cell immunity (Kwak et al., 1995). Further donors were vaccinated with Id-Ig KLH plus GM-CSF with apparent induction of Id-specific CD4þ and CD8þ T cell responses (Li et al., 2000). CD8þ T cells were also induced from a patient with myeloma by stimulation with autologous DCs pulsed with Id-Ig. In mouse models, it is difficult to induce significant levels of CD8þ T cells using protein, even if loaded onto DCs (Shibagaki and Udey, 2002). Using ovalbumin as an exogenous protein, it was found that in order to gain entry to the processing pathway for MHC class I-restricted peptides, it was necessary to add a transduction domain from HIV to the protein. However, certain proteins can apparently enter the MHC class I processing pathway from an exogenous route, since b–galactosidase-loaded DCs induced CTL able to lyse transfected target cells (Paglia et al., 1996). The heterogeneity of these findings and the lack of comparability between human and mouse may reflect different DC subsets, or the different behavior of proteins. MUC1 has also been used as an exogenous protein vaccine, in one trial consisting of identical tandem repeats of its repetitive peptide sequence linked to KLH and conjugated to oxidized mannan (Vaughan et al., 2000). In mice, this vaccine induces CD8þ responses and protection against tumor challenge. However, in primates, and apparently in human subjects, only a humoral response could be induced, possibly because of preexisting natural antibodies against a 1-3-linked galactose (Apostolopoulos et al., 1998). This result emphasizes the importance of species-specific oligosaccharide patterns on the behavior of glycoproteins or carbohydrate adjuvants present in exogenous vaccines. B. Peptides Peptides represent a simple source of tumor antigen, and can be readily manufactured under good manufacturing practice (GMP) conditions. Candidate epitopes for binding to murine or human MHC class I molecules can be predicted by algorithms of increasing power (Rammensee et al., 1999). For human subjects, early attention was focused on the HLA-A*0201 haplotye expressed by 40% of the white population, but binding motifs for other alleles are becoming increasingly available. Importantly, the ability to bind to MHC molecules can then be tested, followed by scrutiny for immunogenicity based on the induction and expansion of functional CTL in vitro (Schultze and Vonderheide, 2001). The next hurdle is whether the CTL generated can kill target tumor cells. This requires processing of the candidate peptide via the proteasomal degradative pathways in the tumor cell. Again, prediction of
TUMOR VACCINES
65
cleavage sites is available via the databases (http://www.uni-tuebingen.de/uni/kxi/). However, it is becoming clear that degradative pathways can be mediated via either the proteasome or the immunoproteasome, depending on the cytokine environment, adding another complication to the selection procedure (Schultze and Vonderheide, 2001). There are many traps in peptide selection, since peptides may not only be cryptic, but analogue peptides may also be antagonistic (Grey et al., 1993). In contrast, modification of anchor residues to improve binding toMHC class I molecules can create agonistic peptides, which are better able to induce CTL (Abrams and Schlom, 2000). The problem for peptide vaccines is their inherent lack of immunogenicity and their biodegradability (Celis, 2002). In spite of this, vaccination of 25 melanoma patients with the MAGE-3A1 nonapeptide with no additions or adjuvants led to significant tumor regression in seven (Marchand et al., 1999). However, in the four patients analyzed, which included two with regression, no CTL activity was detectable. To raise measurable CTL activity, adjuvants may be desirable both to provide stimulation of innate immunity and possibly to create a protected depot. CpG oligonucleotides (see earlier) may also provide useful adjuvants. Delivery of peptide antigens via heat shock proteins is another intriguing possibility, and has been shown to induce protective peptide-specific immunity against tumor cells in models (Blachere et al., 1997). However, the real strength of this approach is in delivering unknown peptides from the intracellular milieu of the tumor cells as vaccines (Przepiorka and Srivastava, 1998). Alternatively, specific peptides can be loaded onto DCs in vitro and cells reinfused. This strategy has been effective in mouse models (Celluzzi et al., 1996; Porgador and Gilboa, 1995). What is remarkable in human subjects is the relative ease of inducing primary and memory CTL responses in vitro using either DCs or CD40-activated B cells as stimulators (von BergweltBaildon et al., 2002). CTL of significant avidity can be generated from both healthy donors and cancer patients, indicating that a repertoire is available. The potential for activating this repertoire by injection of loaded DCs is clear, but there are pitfalls. There are obvious problems in selecting the DCs and in maturing and loading them optimally in vitro under conditions of GMP. The route of injection and the migratory potential of the DCs are other variables, and modeling in mice is not always helpful. A further question concerns the concentration of peptide, which may be critical, since too high a concentration could lead to ‘‘low-quality’’ CTL unable to kill tumor cells (Celis, 2002). However, in spite of these difficulties, clinical trials of peptide-loaded DCs have already been carried out in patients with cancer, including melanoma (Nestle et al., 1998; Thurner et al., 1999), with encouraging immune responses, and possible clinical responses, detected. Patients with advanced breast and ovarian cancer have also been vaccinated with DCs loaded with
66
FREDA K. STEVENSON ET AL.
peptides from HER-2/neu or MUC1 (Brossart et al., 2000). Peptide-specific IFN-g-producing CTL were detected in 5 of 10 patients, and cytolytic activity against a tumor cell line could be mediated by expanded CTL. It will take time to assess if the immense effort and cost involved in ex vivo culture and loading of DCs justify the outcome. However, the approach has clear potential, and, as knowledge accumulates, it should be possible to harness the power of these antigen-presenting cells to activate useful immunity. C. Gene-Based Vaccines The availability of genomic information from normal and tumor cells, together with the development of molecular biological technology able to exploit this information, has opened the possibility of gene-based vaccine designs. The simplicity of the concept of injecting selected gene sequences into patients so that the encoded proteins can be presented to the immune system in situ is attractive. As always, initial enthusiasm has been tempered by the reality of performance. Clearly, time is needed to understand the mechanisms by which the encoded proteins activate immunity, so that the potential of the approach can be fully realized. Translation of observations in preclinical models to patients will also require careful modifications. However, the opportunities to manipulate design and delivery for the whole range of antigens are immense. 1. Naked DNA Vaccines Wolff et al. (1990) first discovered that intramuscular injection of naked plasmid DNA resulted in transfection of muscle cells and production of the plasmid encoded protein, b-galactosidase. It was subsequently demonstrated that vaccination by intramuscular delivery of plasmid DNA encoding the nucleoprotein of influenza A virus could induce specific humoral and cellular immune responses able to protect animals from influenza challenge (Ulmer et al., 1993). In an independent study, Tang et al. (1992) showed that delivery of plasmid DNA coated onto gold particles directly into the skin of mice using a biolistic device (the ‘‘gene gun’’) could elicit a specific immune response against the encoded protein. These pioneering studies have paved the way for the development of a simple and potentially very powerful vaccination technology. Extensive studies in various preclinical disease models have since established that DNA vaccines have the ability to induce a full spectrum of immunological activities and are capable of generating protective immunity in a prophylactic setting. In some cases, DNA vaccines have also been shown to be therapeutically effective against established diseases. The mechanisms by which DNA vaccines elicit immune responses are gradually being revealed (Fig. 3). It is now clear that the bacterial plasmid DNA not only ensures antigen production in transfected host cells, but also
TUMOR VACCINES
67
Fig 3 Induction of immunity by DNA vaccines.
has intrinsic adjuvant properties owing to the presence of immunostimulatory sequences (ISSs) (Krieg, 2002). These ISSs, also known as CpG motifs, are composed of unmethylated CpG dinucleotides with specific flanking nucleotides. Such motifs are present in microbial DNA at a 20-fold greater frequency than in vertebrate genomes and represent one form of pathogen-associated molecular patterns. Recently, the Toll-like receptor 9 (TLR9) has been identified as the ‘‘pattern recognition receptor’’ for bacterial DNA in both mice and humans (Bauer et al., 2001; Hemmi et al., 2000; Takeshita et al., 2001). TLR9 is expressed by B cells, dendritic cells, and other cells of the innate immune system and CpG DNA–TLR9 interaction activates the Toll/IL-1 signaling pathway and triggers an inflammatory responses characterized by the production of IL-6, IL-12, IL-18, TNF-a, IFN-a and -g, and chemokines. The ISSs also induce polyclonal B cell proliferation and stimulate DC maturation with up-regulated expression of MHC class II and costimulatory molecules. In addition, it was recently shown that IL-6 and some other factor(s) produced by CpG- or LPS-stimulated DCs could block the suppressive effect of CD4þ CD25þ regulatory T cells on T cell activation (Pasare and Medzhitov, 2003). It appears that the vertebrate immune system has evolved the ability to recognize the presence of bacterial DNA and activate a coordinated innate and adaptive immune response. DNA vaccines are usually administered by intramuscular (i.m.) or intradermal (i.d.) injection, or into skin cells with a gene gun, and antigens are produced endogenously by transfected cells. When DNA vaccines are
68
FREDA K. STEVENSON ET AL.
delivered through the i.d. route, the skin-derived DCs (Langerhans cells) are directly transfected and probably play a key role in antigen presentation to T cells (Akbari et al., 1999; Condon et al., 1996; Porgador et al., 1998). Vaccine delivery by i.m injection predominantly results in transfection of myocytes (Fig. 3). However, muscle cells do not express MHC class II and costimulatory molecules that are critical for effective T cell priming. Instead, there is compelling evidence that the bone marrow-derived antigen-presenting cells, presumably DCs, play a predominant role in the activation of T cells through a process termed ‘‘cross-priming,’’ whereby the antigen released by muscle cells is taken up, processed, and presented by the bone marrow-derived professional APCs (Corr et al., 1996, 1999; Fu et al., 1997; Iwasaki et al., 1997). The mechanism of this transfer is unclear but could involve ferrying of proteins or peptides via heat shock proteins (Srivastava, 2000). However, the i.m. delivery route may also result in transfection of a small number of bone marrow-derived APCs, which could directly activate T cells (Casares et al., 1997). The outcome of DNA vaccination is to induce all arms of the immune response with a tendency for dominance of the Th1 subset of the CD4þ T cell population. 2. DNA Tumor Vaccines The efficacy of DNA vaccines against infectious diseases has clearly been demonstrated in a variety of animal species for a large number of preclinical models (Gurunathan et al., 2000; Robinson and Pertmer, 2000). Several recent clinical trials have further shown that DNA vaccination is safe and capable of inducing humoral and/or cellular immune responses in humans (MacGregor et al., 1998; Roy et al., 2000; Wang et al., 1998, 2001). It is evident that DNA vaccines are capable of inducing high levels of CTL responses, but are generally weaker than protein vaccines in inducing antibody responses. This may be due to the low levels of antigen produced, but it also may derive from the preferential activation of CD4þ Th1 cells leading to production of IgG2a rather than IgG1 antibodies, as noted in mouse models. Development of effective DNA vaccines to attack cancer cells is a much greater challenge. As discussed earlier, most tumor-associated antigens are nonmutated selfproteins of low immunogenicity, and immunological tolerance or evasion may be real issues. It is therefore not surprising that vaccination simply with a eukaryotic expression plasmid encoding the tumor antigen of interest delivered in saline is rarely effective. Efforts are therefore being made to develop various strategies to overcome these hurdles. One widely exploited and simple approach is coadministration of plasmids encoding cytokines, chemokines, or costimulatory molecules. The usage of these genetic adjuvants can not only increase the magnitude but also modulate the type of immune responses to DNA vaccines. However, a dynamic immune response is likely
TUMOR VACCINES
69
to involve these molecules at different stages, and it may be difficult to deliver coactivators at the time required. The strategy of codelivery has recently been reviewed in detail elsewhere (Berzofsky et al., 2001; Scheerlinck, 2001) and will not be further discussed here. We will focus on various DNA fusion vaccine strategies that have been designed to break tolerance and to enhance antigen presentation and immune cell activation. In our view, provision of additional T cell help for weak or tolerizing antigens is a key component for successful induction of antitumor immunity. We will also discuss recent developments in strategies for improving in vivo antigen delivery and production. In addition, DNA vaccines can be combined with various viral delivery systems in a primeboost fashion. Although the latter strategies are largely being investigated for DNA vaccines against infectious diseases, they should also be applicable to cancer. 3. Genetic Fusion Vaccine Strategies The ease of DNA manipulation has facilitated construction and evaluation of various DNA vaccines that genetically link antigens to a range of immunomodulating molecules. These fusion vaccine strategies have been designed to provide cognate CD4þ T cell help critical for antigen-specific B and T cell activation, to activate the innate immune system, to target antigen delivery to professional antigen-presenting cells, and to direct antigen to more relevant processing and presentation pathways. a. Provision of T Cell Help. Conjugation of an immunogenic carrier protein or a Th epitope has proved to be an effective strategy to improve antibody responses to polysaccharide and peptide/protein-based vaccines through provision of cognate CD4þ T cell help. It is also clear that CD4þ T cells play a key role in the induction and maintenance of CD8þ cytotoxic T cell responses (Bourgeois et al., 2002; Clarke, 2000; Janssen et al., 2003). For effective activation of antitumor immunity, it is essential to incorporate strategies capable of breaking tolerance in the CD4þ T cell compartment that likely have occurred in the tumor-bearing host (Bogen, 1996). DNA fusion vaccine strategies designed to provide cognate CD4 T cell help have been shown to be effective in promoting the induction of B cell and T cell responses in a number of preclinical models. In murine B cell lymphomas, vaccination using DNA constructs encoding Id of immunoglobulin antigens genetically linked to immunogenic carrier sequences, either xenogeneic human Ig constant regions (Syrengelas et al., 1996) or a pathogen-derived sequence, fragment C (FrC) of tetanus toxin (King et al., 1998), can generate protective anti-Id antibody responses against lethal tumor challenge. The linkage of immunogenic carrier sequences is absolutely required, because immunization with vaccines containing Id sequence only or fused to autologous mouse Ig constant regions fails
70
FREDA K. STEVENSON ET AL.
to induce an anti-Id antibody response (King et al., 1998; Spellerberg et al., 1997; Syrengelas et al., 1996). Importantly, fusion was required because separate genes, even within the same plasmid construct, were ineffective (King et al., 1998). The FrC gene fusion strategy also dramatically increased anti-Id immunity against T cell lymphoma using a DNA vaccine encoding the TCRa/b idiotypic sequences (Thirdborough et al., 2002). Recently, a DNA vaccine trial in patients with stage III or IV follicular B cell lymphoma has been reported (Timmerman et al., 2002). Patients were immunized with plasmid constructs encoding a chimeric immunoglobulin molecule consisting of a tumor-specific Id sequence linked to xenogeneic (murine) constant regions with and without codelivery of GM-CSF-encoding plasmid. A majority of patients mounted B and/or T cell responses to the murine Ig component, while one patient had an Id-specific T cell response and several patients had immune responses that were cross-reactive with other patients’ Id proteins. The data showed that DNA vaccination was safe and potentially a useful approach to anti-Id immunotherapy. The modest antitumor immune responses may be improved by inclusion of more potent carrier sequences. Our laboratory is currently conducting a Phase I/II clinical trial in patients with follicular lymphoma, using idiotype-encoding variable region gene sequences fused to the FrC sequence. Preliminary data show that this strategy is capable of inducing strong proliferative and antibody responses against the FrC component and proliferative IFN-gproducing T cell responses specific to the idiotypic determinants in a majority of patients (unpublished observation). It is possible that the high level of T cell help against FrC is particularly effective in amplifying antitumor immunity in human patients. Importantly, both the clinical trials of DNA idiotypic fusion vaccines show that naked DNA is capable of inducing immune responses even in patients with lymphoma, thus providing a platform for further clinical development of the approach. Cognate T cell help provided by fusion to FrC (King et al., 1998) or to the coat protein of potato virus X (PVXCP) (Savelyeva et al., 2001) also promotes anti-Id immune responses that are protective against an Ig secreting but surface Ig-negative myeloma, with CD4þ T cells as major mediators of protection. The induction of this enhanced Id-specific CD4þ T cell response is probably by a Th–Th cooperation mechanism as proposed by Gerloni et al. (2000); namely, CD4þ T cells reactive with a dominant determinant provide help to other CD4þ T cells recognizing weaker epitopes through up-regulating the costimulatory ability of APCs. Interestingly, the Id–PVXCP fusion vaccine design is also able to induce protective immunity against a surface Ig-positive lymphoma, but unlike the Id–FrC fusion vaccine, the protection is mediated by CD4þ T cells (Savelyeva et al., 2001). Moreover, PVXCP appears to preferentially direct anti-Id antibody response to the IgG2a subclass, indicating a Th-1-type response, whereas FrC also induces IgG1 (King et al., 1998), a
TUMOR VACCINES
71
Th-2-type response. It appears therefore that the molecular nature of the encoded immunoenhancing protein influences the immune pathways activated, and therefore the protective mechanism. Fusion of immunogenic carrier sequences also promotes induction of CD8þ CTL responses targeted to HIV envelope protein (Le Borgne et al., 1998), influenza NP and HPV E7 CTL epitopes (Wolkers et al., 2002), and to a tumor rejection antigen from an endogenous retrovirus (Takeda et al., 2000). However, for amplifying CTL responses, care is required when adding immunoenhancing sequences to avoid the introduction of potentially competitive MHC class I-binding epitopes. For our vaccines that exploit FrC of tetanus toxin as an immunoenhancer, we have engineered a truncated FrC single domain sequence to remove potentially immunodominant CTL epitopes identified in the second domain, while still retaining the known Th epitope within the first domain (Rice et al., 2001, 2002a). Another modification was to place the encoding CTL epitope sequences at the 30 end of the FrC domain. This design led to accelerated and increased CTL responses to the fused epitopes, and protective or therapeutic immunity against tumor. In contrast, the response is less if unmodified, full-length two domain FrC is present, consistent with the phenomenon of immunodominance (Rice et al., 2001). Interestingly, reduction of the truncated FrC to a single helper epitope provided only a modest increase of the CTL response against the fused peptide sequence, indicating a requirement for the full domain. It is unclear if this is due to additional unknown CD4þ epitopes in the domain, or to another property of the domain that could affect the cross-priming process. These results suggest that it is possible to activate an optimal effector mechanism for the target antigen by selecting an appropriate carrier sequence. b. Complement. The complement system plays a key role in innate immunity and also represents an important link between innate and acquired immunity (Fearon, 1998; Ochsenbein and Zinkernagel, 2000). A recombinant protein vaccine consisting of the hen egg lysozyme (HEL) fused to murine C3d component of complement has been shown to dramatically increase (>1000-fold) the efficiency of antibody response against HEL in mice (Dempsey et al., 1996). Ross et al. (2000, 2001) have further demonstrated that the fusion of C3d can also augment the humoral immune response against hemagglutinin (HA) of influenza and against HIV gp120 delivered as DNA vaccines. Immunization with the HA–C3d fusion vaccine induces HA-specific antibodies with higher avidities more rapidly. The accelerated antibody response correlates with a faster appearance of protective immunity. Remarkably, a single vaccination with a low dose of fusion vaccine is able to confer complete protection against live virus challenge (Ross et al., 2000). The mechanism by which C3d promotes the humoral response is likely to be by enhanced stimulation and expansion of antigen-specific B cells. C3d binds to its receptor, CD21, that is expressed and
72
FREDA K. STEVENSON ET AL.
associated with the costimulatory molecule CD19 on B cells and follicular dendritic cells. The fusion of an Ag to C3d should allow for a cross-linking of Ag-specific B cell receptor and CD21, thereby delivering a more effective signal for activation. CD21 is also expressed on follicular dendritic cells (FDCs) in germinal centers where affinity maturation takes place. It is conceivable that C3d could facilitate Ag binding to FDCs, and the trapped Ag then attracts and stimulates affinity maturation in responding B cells (Fearon and Carroll, 2000). The humoral immune response plays an important role in the control of certain tumors. Enhancing B cell signaling by the C3d fusion strategy may therefore be a useful approach for developing cancer vaccines with improved humoral response. c. Defensins and Chemokines. Defensins are a family of small, cationic peptides that directly contributes to the killing of microbes, and a recent study by Zhang et al. (2002) has identified human a-defensins as the major components of the antiviral factor secreted by CD8 T cells of HIV-infected patients. Moreover, it has been shown that human b-defensins are also chemotactic for immature DCs and memory T cells (Biragyn et al., 2001; Yang et al., 1999), while murine b-defensin 2 acts directly on immature DCs via TLR-4, inducing up-regulation of costimulatory molecules and DC maturation (Biragyn et al., 2002b). Thus defensins also play a role in bridging innate immunity and adaptive immune responses. Biragyn et al. (2001) show that murine b-defensins 2 and 3 bind CCR6 preferentially expressed on immature DCs and that DNA fusion vaccines consisting of nonimmunogenic B cell idiotypic antigens and murine b-defensin 2 induce protective and therapeutic immunity against two different syngeneic mouse lymphomas. The same study shows that fusion vaccines consisting of tumor Id and inflammatory chemokines MIP-3a or MCP-3, which are chemotactic for immature DCs, are also effective. The same fusion vaccine strategy also enhances immune responses against HIV-1 gp120 (Biragyn et al., 2002a). Importantly, both chemotactic activities and the covalent linkage are required, since neither fusion of inactive pro-b-defensin or mutant chemokine nor coimmunization with a mixture of plasmids encoding unlinked Id antigens and defensin or chemokines is effective. These results suggest that targeting the delivery of antigens to immature DCs is a general applicable approach to increasing the efficacy of DNA vaccines. d. Targeting to Antigen-Presenting Cells. One problem for naked DNA vaccination is that only a small number of host cells is transfected and a relatively low level of antigen is produced. The small amount of antigen available for processing and presentation by APCs could therefore be a limiting factor for the potency of DNA vaccines (Heath and Carbone, 2001).
TUMOR VACCINES
73
This may be a particularly important issue for intramuscular DNA vaccination where antigen transfer from transfected muscle cells to professional APCs is required for effective induction of immune responses. Therefore, various strategies for targeting antigen delivery to professional APCs are currently being explored. Boyle et al. (1998) have shown that i.m. vaccination in mice with a DNA construct encoding a human IgG (hIgG) and the extracellular domain of CTLA-4 fusion protein can generate a faster and higher (up to 10,000-fold) anti-hIgG antibody response than DNA vaccines encoding hIgG alone. The CTLA-4 fusion also enhances the T cell proliferative response (up to 8-fold). In the same study the authors showed that fusion to L-selectin, which binds to CD34 on high endothelial venule cells of lymph node, could also improve the immune response but to a lesser extent. Subsequent studies further show that CTLA-4 DNA fusion vaccines are able to induce accelerated and increased antibody responses against HA of influenza, chicken ovalbumin, and malaria apical membrane antigen (AMA1) (Deliyannis et al., 2000; Lew et al., 2000). More importantly, the CTLA-4 fusion designs are capable of providing more effective protection against influenza virus challenge in a mouse model (Deliyannis et al., 2000) and against Corynebacterium pseudotuberculosis challenge in sheep (Chaplin et al., 1999). A recent report showed that the CTLA-4 fusion protein vaccination approach was also able to generate protective immunity against a self-tumor antigen (Huang et al., 2000). CTLA-4 is expressed by activated T cells and binds with high affinity to CD80/CD86 surface receptors expressed by APCs. The findings that CTLA-4–HA retains the ability to bind to CD80 and CTLA-4-Ig fusion protein and is at a 4-fold higher level than nonfused protein in the draining lymph nodes within 2–24 hr after immunization (Deliyannis et al., 2000) suggest that the mechanism of enhanced immune responses by the CTLA-4 fusion constructs is more efficient antigen capture and presentation by APCs. Another approach for improving antigen delivery to APCs is to target the Fcg receptor. A DNA vaccine encoding a fusion protein consisting of a hepatitis B virus e antigen (HBeAg) and a human IgG Fc has been shown to induce enhanced antigen-specific B cell, CD4þ Th and CD8þ CTL responses in mice as compared to a DNA vaccine encoding HBeAg alone (You et al., 2001). DCs express IgG Fc receptors (FcgRs), and interactions between FcgRs and antigen–Ig complexes result in antigen internalization. Soluble antigen uptake by receptor-mediated endocytosis is 1000- to 10,000-fold more efficient than fluid phase pinocytosis (Regnault et al., 1999; Sallusto and Lanzavecchia, 1994), and the endocytosed antigen is processed via both the MHC class II pathway for CD4þ T cell induction and via MHC class I pathway for CD8þ CTL induction (Regnault et al., 1999). Furthermore, the interaction of Fc with FcgRs stimulates DC maturation by up-regulating expression of surface molecules that are important for antigen presentation
74
FREDA K. STEVENSON ET AL.
(Regnault et al., 1999). It should be noted that in this study a human IgG Fc fragment is used; it is therefore possible that the xenogeneic sequence has contributed to the increased immune responses by providing cognate T cell help as discussed earlier. Since cross-presentation is an important pathway for DNA vaccination, increasing antigen transfer from transfected somatic cells to professional APCs may improve the potency of DNA vaccines. This hypothesis has been tested by fusing antigens to the herpes simplex virus type 1 (HSV1) VP22 protein that possesses an intercellular trafficking property (Elliott and O’Hare, 1997). Hung et al. (2001) constructed DNA vaccines with the HPV E7 antigen fused either to the full-length VP22 or a C-terminally truncated VP22 (DVP22) that is defective for intercellular spreading. Vaccination with VP22-E7 induced a markedly improved CD8þ CTL response and tumor immunity compared to the DNA vaccine encoding E7 only. The DVP22-E7 vaccine did not show improvement. In a similar study, Michel et al. (2002) also showed that fusion of VP22 to E7 led to dramatic enhancement of CTL induction. However, the latter study found that fusion to DVP22 was equally effective, indicating that other mechanism(s) may also contribute to the enhancement, such as increased stability of E7 and more efficient secretion. In addition, as discussed earlier, the virus-derived VP22 protein may contain Th epitope(s) that would provide cognate help to the fused antigen. Further investigations will be required to elucidate the exact mechanism(s). One mechanism of cross-priming by professional APCs is acquisition of exogenous antigens by phagocytosis of apoptotic cells (Albert et al., 1998). There is also evidence that DNA vaccination causes injury, local inflammation, and cell death, which may account, at least in part, for the ability of DNA vaccines to induce T cell activation. Two recent reports have reinforced this notion. First, Chattergoon et al. (2000) showed that intramuscular codelivery of a Fas-expressing plasmid with DNA vaccines encoding HIV-1 env or gag/pol generated an increased antigen-specific CTL response, as compared with vaccines without Fas coexpression. They found that anti-Fas antibody treatment of cells transfected in vitro with Fas-expressing plasmid underwent rapid apoptosis, and Fas expression in situ resulted in significant infiltration into the muscle of CD4þ and CD8þ T cells and CD11cþ phagocytic cells. Furthermore, coinjection of Fas and EGFP plasmids led to an 5-fold increase in the number of APCs (macrophages and DCs) containing apoptotic EGFP positive cells in the proximal draining lymph node. Second, Sasaki et al. (2001) constructed dual-expression plasmids encoding influenza virus HA or NP antigens and partially inactivated caspases 2 and 3. These dual vectors were still able to induce apoptosis in vitro but at a level about 10-fold lower than their fully active counterparts. Vaccination with the dual vectors-encoding
TUMOR VACCINES
75
mutant, but not intact caspases, could elicit markedly enhanced T cell responses, indicating that sufficient antigen production in transfected cells before cell death was required. One difference between the two studies is that Fas can be delivered in a separate plasmid from the vaccine, whereas expression of mutant caspases and immunogens in the same cells is required. Nevertheless, the two studies suggest that deliberate induction of apoptosis of antigen-producing cells can enhance antigen presentation by professional APCs and increase the potency of DNA vaccines. e. Somatic Transgene Immunization. Perhaps the most direct approach to optimize antigen presentation by professional APCs within organized lymphoid organs is injection of DNA vaccines into spleen or into lymph nodes. The intrasplenic approach was first described by Zanetti and colleagues (Gerloni et al., 1997), who showed that inoculation of a plasmid containing a chimeric immunoglobulin heavy-chain gene under the control of immunoglobulin promoter and enhancer elements resulted in chromosomal integration in B cells of the plasmid, expression of the transgene, and induction of a specific antibody response. This report of integration from a DNA vaccine is unique, since no significant integration has been found in exhaustive studies of more conventional injection sites (Manam et al., 2000). In this procedure, termed somatic transgene immunization, antigen presentation and T cell priming are dependent on the transgenic B cells that produce the antigen, and apparently do not require bone marrow-derived DCs (Castiglioni et al., 2003). In a further study they inserted a weak CD4þ T cell epitope from MUC1 and a strong epitope from a malaria antigen in the CDRs of the VH gene. Immunization with this construct could induce a strong CD4þ T cell response against the MUC1 epitope (Gerloni et al., 2000). In the study by White et al. (2000), a plasmid encoding CEA under the control of the CMV promoter/enhancer was used for intrasplenic (i.spl) or intramuscular immunization, and antibody responses and tumor protection were assessed. Immunization by the i.spl route induced CEA-specific immune responses comparable to that elicited by i.m. inoculation. Vaccination by intra-lymph node (i.ln) delivery of plasmid DNA was reported by Maloy et al. (2001) to be more effective than conventional i.m. or i.d injection. They immunized mice with a plasmid containing an LCMV gp33 CTL epitope linked to GFP by various routes, including i.m., i.d., i.ln, and i.spl injection. They found that i.ln injection enhanced immunogenicity by 100- to 1000-fold and induced superior CT8þ CTL immunity against viral challenge, and could eradicate the peripherally transplanted tumor pieces expressing the gp33 epitope. There was evidence that a small fraction of CD11cþ lymph node cells expressed GFP protein 24 hr after i.ln vaccination,
76
FREDA K. STEVENSON ET AL.
suggesting that the enhancement could be attributed, at least in part, to increased uptake and expression of naked DNA vaccines in DCs and more efficient priming of T cell responses. 4. Manipulating Antigen Presentation and Processing Pathways The ease of recombinant DNA manipulation provides an opportunity to test various strategies to enhance antigen processing via particular pathways. For example, optimal antibody responses are often achieved by linking antigens to secretory signal sequences. There is also evidence that targeting of proteins to the endoplasmic reticulum may increase levels of induced CTL, possibly via retrograde transport into the cytosol (Ciernik et al., 1996). Alternatively, increasing intracellular antigen degradation may enhance CTL induction. Fusion of ubiquitin to several antigens reduces or abolishes antibody responses but enhances cell-mediated immunity with increased CTL activities and IFNg production, and results in improved protection against viral, bacterial, and tumor challenge (Delogu et al., 2000; Rodriguez et al., 1998; Velders et al., 2001b). Ubiquitin plays an important role in proteasome-dependent degradation of intracellular proteins. In some of these reports in vitro studies have confirmed that linkage of ubiquitin leads to more rapid protein degradation, suggesting that the enhanced cellular immune responses are the results of more efficient intracellular antigen processing and presentation via the MHC Class I pathway. However, the influence of ubiquitination on protein degradation and induction of a CTL response appears to be antigen dependent, as the stability of influenza nuclear protein (NP) is not affected by conjugation to ubiquitin in vitro (Fu et al., 1998), and vaccination with ubiquitin fusion vaccines does not increase induction of CTL responses against NP (Fu et al., 1998) or against hepatitis C core antigen in vivo (Vidalin et al., 1999). Heat shock proteins (HSPs), including HSP70, HSP90, gp96, and calreticulin (CRT), play multiple roles in antigen presentation by MHC class I molecules (Basu and Srivastava, 2000). These chaperone proteins form complexes with antigenic peptides, and are involved both in the transport of these peptides to the endoplasmic reticulum (ER) and in the folding and assembly of MHC class I molecules. Furthermore, complexes of HSPs and antigenic peptides released by dying cells, such as virus-infected cells or tumor cells, are efficiently taken up by APCs via the a2-macroglobulin receptor CD91 and re-presented by MHC class I molecules (Basu et al., 2001). Considering the importance of cross-presentation in the induction of CD8þ T cell response by DNA vaccines, linking an antigen to HSP may potentially be an effective approach for enhancing the induction of a CTL response. The CD8þ CTL response specific for HPV16 E7 antigen could be improved by linkage to HSP70 of Mycobacterium tuberculosis (Chen et al., 2000) and to CRT (Cheng et al., 2001a). Similarly, E7-HSP70 fusion vaccines based on
TUMOR VACCINES
77
Sindbis virus self-replicating RNA vector and Semliki Forest virus RNA vector also generated enhanced E7-specific Tcell-mediated immune responses (Cheng et al., 2001b; Hsu et al., 2001). In addition, CRT possesses antiangiogenic activity that inhibits tumor growth in vivo (Pike et al., 1999). Intradermal delivery of the CRT–HPV17 E7 fusion vaccine not only induced increased CTL activities, but also showed antiangiogenesis activities that further enhanced the vaccine potency (Cheng et al., 2001a). In addition to their role in antigen presentation, HSPs are considered as endogenous ‘‘danger’’ signals (Gallucci and Matzinger, 2001), in that they can activate DCs and macrophages, and stimulate production of proinflammatory cytokines (Asea et al., 2000; Basu et al., 2000; Chen et al., 1999; Kol et al., 2000; Singh-Jasuja et al., 2000). The powerful immunostimulatory activities of HSPs make them potentially excellent adjuvants for DNA vaccines. It is increasingly clear that CD4þ T cells are critical in controlling tumors in vivo, mediating direct and indirect killing of tumor cells, as well as providing help for B cells and CD8þ CTLs. Rational vaccine design can also direct antigen processing and presentation to the MHC class II pathway in order to enhance CD4þ T cell responses. The MHC class II-associated invariant chain (li) has recently emerged as an attractive molecule for enhancing MHC class II presentation (Koch et al., 2000; van Bergen et al., 1999). The li ensures correct assembly of MHC class II a- and b-chains in the ER, directs them to the endosomal pathway, and facilitates loading of antigenic peptides (Pieters, 2000). In one DNA vaccine design, the MHC class II-associated invariant chain peptide (CLIP) sequence was replaced by an MHC class II-restricted epitope derived from HSP60 (van Tienhoven et al., 2001). Immunization of rats using this fusion construct induced an HSP-specific T cell proliferative response, while immunization with a plasmid encoding full-length HSP60 failed. The ability of the fusion construct to induce a CD4þ T cell response was correlated with improved antigen presentation by MHC class II molecules as demonstrated by in vitro studies. The CLIP substitution approach has previously been shown to increase MHC class II presentation of other antigens in vitro (Fujii et al., 1998; van Bergen et al., 1997). More recently, a cell-based vaccine with similar CLIP substitution design has been shown to induce a strong CD4þ T cell response that is protective against a highly aggressive tumor (van Bergen et al., 2000). Enhancement of MHC class II presentation has also been achieved by targeting antigen to specific organelles. In one approach, antigens were linked to lysosomal-associated membrane protein LAMP-1 (Rowell et al., 1995; Wu et al., 1995). This approach uses the sorting signal in the cytoplasmic tail of LAMP-1 to target antigens to the endosomal/lysosome compartment where MHC class II-restricted peptides are generated. DNA fusion vaccines consisting of HPV E7 and LAMP-1 conferred superior tumor protection in a mouse
78
FREDA K. STEVENSON ET AL.
model (Ji et al., 1999; Smahel et al., 2001), and increased CD4þ and CD8þ T cell responses (Ji et al., 1999). In another approach, minigenes encoding Th epitopes from LCMV were linked to the 20-amino acid C-terminal tail of lysosomal integral membrane protein-II (LIMP-II) (Rodriguez et al., 2001). The potential advantage of LIMP-II is that it moves directly from ER to the lysosomal compartment, directed by residues in its C-terminal tail. Using this strategy, induction of a CD4þ T cell response to the NP309–328 epitope of LCMV was greatly enhanced as assessed by IFN-g production. However, the same design did not improve the CD4þ T cell response to the GP61–80 epitope of LCMV, possibly due to rapid degradation of this epitope by cathepsin D in lysosomes. 5. Technologies for Enhanced DNA Vaccine Delivery Conventional DNA vaccine delivery by injection of naked plasmid DNA in saline solution results in inefficient uptake of DNA by myocytes and poor transfection of APCs, especially in large animals and in primates (Dupuis et al., 2000). One approach for improving in vivo transfection is to use electroporation following inoculation of plasmid DNA. Electroporation has two consequences: first, formation of inverted hydrophilic pores in the cell membrane, and second, electrophoresis of the DNA into the cell (Satkauskas et al., 2002). The outcome is to increase DNA uptake and protein expression by cells of skin, muscle, liver, and tumors (Bigey et al., 2002; Smith and Nordstrom, 2000). When applied to DNA vaccination, electroporation improves vaccine potency with faster induction and a higher level of immune response against viral and bacterial antigens in mice, guinea pigs, rabbits, and pigs (Babiuk et al., 2002; Drabick et al., 2001; Dupuis et al., 2000; Selby et al., 2000; Tollefsen et al., 2002; Widera et al., 2000). Electroporation also increases the efficacy of antimelanoma DNA vaccines in murine models (Kalat et al., 2002; Mendiratta et al., 2001). Generally, there is a correlation between the level of gene expression and the magnitude of induced immunity. However, application of electrical current per se may contribute to the immune enhancement. Electroporation may induce some inflammatory responses with release of cytokines and migration of antigen-presenting cells to the site of increased antigen production, thereby acting synergistically to enhance immune responses. While electroporation is being used clinically to increase uptake of chemotherapeutic drugs (Mir, 2001), so far it has not been used in a clinical trial of DNA vaccination. Another strategy for improving in vivo plasmid DNA uptake is to use biodegradable microparticle (<10 mM) delivery technology. Plasmid DNA can be either encapsulated in a poly(lactide-co-glycolide) (PLG) microsphere or adsorbed onto the surface of cationic PLG microparticles. Both procedures are able to augment the antibody and cellular immune responses of DNA vaccines delivered via mucosal (Jones et al., 1997; Singh et al., 2001) or i.m.
TUMOR VACCINES
79
routes (McKeever et al., 2002; O’Hagan et al., 2001; Singh et al., 1997) in both mice and in larger animal species such as guinea pigs and rhesus macaques (O’Hagan et al., 2001). Adsorption of DNA to the microparticle surface appears to be less damaging to DNA and may be a preferable option (O’Hagan et al., 2001). Possible mechanisms for the enhancement include reduced enzymatic DNA degradation and prolonged antigen expression, and perhaps more importantly, phagocytosis of plasmid/PLG microparticles and subsequent antigen presentation by professional APCs (Denis-Mize et al., 2000; Lunsford et al., 2000). PLG is a biocompatible and biodegradable material with an extensive record of safe use in humans; DNA vaccine delivery via microparticles may thus offer a promising technology for clinical application. 6. Alphavirus-Based Cancer Vaccines Another strategy to increase performance of DNA vaccines is to deliver the genes via viral vectors. Alphaviruses are positive-strand RNA viruses capable of replicating in many vertebrate and invertebrate cells. The alphavirus genome is 12 kb in length and is divided into two open reading frames (ORFs). The first ORF located in the 50 two thirds of the genome encodes four nonstructural proteins (nsP1–nsP4) that function as a replicase to transcribe negativestrand as well as positive-strand viral RNA. The second ORF is located in the 30 one third of the genome and encodes the viral structural proteins. A subgenomic (26S) promoter is located between the two ORFs in the negative strand. RNA replication occurs via synthesis of a negative strand of the genomic RNA. This is used as the template for transcription of additional full-length genomic RNA, and for transcription of a positive-strand subgenomic RNA encompassing the second ORF by the nonstructural proteins from the 26S promoter. The subgenomic RNA can accumulate to levels as high as 106 copies per cell. For vaccine purpose, alphavirus replicon vectors retain the sequence coding for the nonstructural proteins, but a heterologous gene replaces the genes encoding the viral structural proteins. Most alphaviral replicon vaccines currently being developed are based on Sindbis (SIN) virus, Semliki Forest virus (SFV), or Venezuelan equine encephalitis (VEE) (Rayner et al., 2002). The recombinant replicon vaccines can be delivered as recombinant viral particles, in vitro transcribed RNA, or plasmid DNA with initial transcription of the RNA genome under the control of a eukaryotic promoter. Many studies have shown that replicon vaccines encoding tumor as well as microbial antigens are significantly more potent than conventional plasmid DNA vaccines. For example, Cheng et al. (2001b, 2002) showed, using HPV–E7 tumor antigen as a model and various gene fusion strategies, that delivery of the tumor antigen as Sindbis replicon RNA, DNA, or viral particles can markedly enhance antitumor immunity. The enhancement included increased CD4þ , CD8þ , and NK cell activities,
80
FREDA K. STEVENSON ET AL.
depending on the form of the vaccines. Daemen et al. (2000) constructed recombinant SFV expressing HPV E6 and E7 antigens (SFV–E6E7). They found that immunization of mice with SFV–E6E7 recombinant viral particles resulted in efficient activation of HPV-specific CTL activities, and protection from HPV–E7 expressing tumor cells. Recombinant SFV vectors expressing the P815A antigen (a murine cancer-testis antigen) were also able to induce strong CTL responses and protect mice against P815 tumor challenge (Colmenero et al., 1999). Similarly, a recombinant VEE expressing HPV–E7 was constructed and tested in mice (Velders et al., 2001a). Vaccination with the VEE–HPV–E7 vector induced CD8þ T cell responses as assessed by IFN-g production, tetramer staining, and cytotoxicity assays. Furthermore, vaccination completely prevented tumor growth and was even effective in eliminating 7-day established tumors. When b-galactosidase was used as a model tumor antigen, it was shown that vaccination with replicon vaccines in either RNA form (SFV–LacZ) (Ying et al., 1999) or as DNA plasmid (pSIN–LacZ) (Leitner et al., 2000) could elicit strong b-gal-specific humoral and cytotoxic T cell responses. Prevaccination protected mice from tumor challenge, and therapeutic treatment could significantly prolong the survival of mice with established tumors. Compared to conventional plasmid vaccines, DNA replicons were as effective at doses 100- to 1000-fold lower. More significantly, a recent report by Leitner et al. (2003) further showed that alphavirus-based DNA vaccination could effectively induce protective immunity against a nonmutated, tumor-associated self-antigen. Vaccination with a pSIN plasmid expressing the melanoma-associated antigen, tyrosinase-related protein 1 (TRP-1), was as effective as a plasmid encoding the xenogeneic human TRP-1 in breaking immunological tolerance (Leitner et al., 2003). It was initially thought that the self-replicating nature of alphaviruses would drive vigorous self-amplification of the viral RNA genome, leading to highlevel antigen production. There is indeed evidence for substantial amplification of viral RNA in in vitro transfected cells (Ying et al., 1999). However, antigen expression in vitro or in vivo was not necessarily increased (Hariharan et al., 1998; Leitner et al., 2003). The elegant study of Leitner et al. (2003) instead indicated that production of double-stranded RNA, a potent inducer of innate immunity, and increased apoptosis of transfected cells contributed more significantly to the enhanced vaccine potency. As discussed previously, apoptosis may enhance antigen presentation by dendritic cells and significantly improve the efficacy of DNA vaccines (Chattergoon et al., 2000; Sasaki et al., 2001). 7. Prime-Boost It is generally accepted that current immunization strategies, with protein or attenuated/inactivated virus, require booster vaccinations to maintain or improve immunity. In the context of gene-based vaccines there is now a
TUMOR VACCINES
81
large body of evidence in animal models demonstrating the success of heterologous prime-boost protocols (Ramshaw and Ramsay, 2000). This strategy entails priming with DNA and then boosting via sequential delivery of the same recombinant antigen delivered in a different vaccine vector (such as a replication-restricted recombinant virus), or as recombinant protein. In this way it is possible to generate higher levels of immunity when compared with homologous vaccination with DNA, recombinant viral vectors, or protein. Following DNA vaccination, improved humoral responses have been demonstrated when boosting with either recombinant protein (Babiuk et al., 2002; Jones et al., 2001; Tanghe et al., 2001) or recombinant viral vectors (Eo et al., 2001; Meseda et al., 2002; Ramshaw and Ramsay, 2000). Boosting with replication-restricted recombinant virus can also lead to enhanced T cell responses, including CD8þ CTL (Hanke and McMichael, 1999; Hanke et al., 1998, 1999; Robinson et al., 1999; Schneider et al., 1998). This may be particularly significant for the immunotherapy of cancer because the intensity of the immune response appears to determine tumor clearance (Perez-Diez et al., 2002). This vaccination strategy has led to enhanced immunogenicity and protective efficacy in a range of disease models, including HIV (Amara et al., 2001; Kent et al., 1998; Robinson et al., 1999), malaria (Jones et al., 2001; Schneider et al., 1998; Sedegah et al., 2000), and cancer (Greiner et al., 2002; Grosenbach et al., 2001). While the precise immunological mechanisms responsible for the efficacy of this approach have yet to be elucidated, several explanations have been proposed to account for the synergistic effect of this strategy. The simple, nonreplicating nature of DNA vaccines ensures that the immune response will become focused on vaccine-encoded antigens (Harrington et al., 2002). In addition, the immunostimulatory nature of CpG elements within the bacterial plasmid backbone will lead to preferential activation of CD4þ Th1 cells, thereby orienting the immune response in a particular direction. In the mouse, for example, this orientation will have a propensity to induce immunoglobulin of the IgG2a subclass and will also favor CD8þ CTL induction. It has also been suggested that the low, but persistent nature of antigen expression following DNA vaccination might select for responding T cells with higher affinity TCRpeptide/MHC interactions (Busch and Pamer, 1999; Estcourt et al., 2002; Rees et al., 1999). Subsequent immunization with a recombinant nonreplicating viral vector, expressing the same antigen, boosts this immunity. Again, the immunological mechanisms responsible for this effect are unclear, although they probably include a combination of increased antigen expression, effective delivery of antigen to the processing and presentation machinery of transfected or infected cells, and the inflammatory response associated with delivery of an immunogenic viral vector (Ramshaw and Ramsay, 2000; Zavala et al., 2001).
82
FREDA K. STEVENSON ET AL.
While this approach shows promise in the field of infectious diseases, with several immunization strategies entering clinical trials (Doolan and Hoffman, 2001; Hanke and McMichael, 2000), it is not without its limitations. DNA vaccines are considered to be safe for use in humans; however, the safety of many recombinant viral vaccines has not been sufficiently tested. Certain vectors (e.g., modified vaccinia virus Ankara [MVA]), do appear to be suitable for use in humans. MVA is replication restricted and appears to be safe following immunization of approximately 120,000 humans, including high-risk groups, during the global program to eradicate smallpox (Blanchard et al., 1998; Mahnel and Mayr, 1994; Mayr et al., 1978); testing in animal models, including irradiated and neonatal mice confirm the avirulence and safety of this vector (Mayr et al., 1978). Such safety concerns will be paramount in the setting of cancer immunotherapy, since many patients will be immunocompromised due to tumor or myeloablative treatment regimes. Schneider et al. (1998) demonstrated enhanced boosting of a DNA-primed CTL response when using recombinant MVA, compared with a virulent strain of recombinant vaccinia virus. This may be due to the deletion of a number of viral immune evasion genes from the MVA genome, including soluble cytokine receptors (e.g., IFN-g, TNF, and IFN-a/b) and the CC chemokine-binding receptor (Alcami and Koszinowski, 2000; Blanchard et al., 1998). Despite an inability to replicate in most mammalian cells, this could explain the enhanced immunogenicity of MVA, making it the vector of choice for many prime-boost studies in the field of infectious diseases. However, the immunogenicity of many viral vectors can act as a double-edged sword. The first problem is that preexisting immunity to vector-encoded antigens can neutralize the virus, impairing responses to vaccine-encoded antigen (Cooney et al., 1991; Kundig et al., 1993; Palmowski et al., 2002; Rooney et al., 1988; Schneider et al., 1998). This is particularly pertinent in the current climate, with many countries prepared to immunize their citizens against smallpox, to counter the threat of bioterrorism. This could negate the use of vaccinia-based vectors for human prime-boost vaccination programs of infectious diseases and cancer. Even where human subjects do not have preexisting immunity to a particular viral vector, its administration could negate further use. This factor could limit application of the prime-boost approach to cancer immunotherapy, where the immunosuppression of many patients and the low immunogenicity of many tumor antigens suggest that it will probably be necessary to give patients multiple booster vaccinations in order to generate and maintain tumor-specific responses. An additional concern is that highly immunogenic viral vectors may drive coexpression of viral proteins containing potentially immunodominant T cell epitopes. This could prevent the immune response from focusing exclusively on ‘‘weak’’ vaccine-encoded tumor antigens (Harrington et al., 2002; Yewdell and Bennink, 1999).
TUMOR VACCINES
83
For infectious diseases, a recent strategy to induce an even more effective CD8þ T cell response has been to prime with one recombinant viral vector and then boost with another, replacing the DNA vaccine component altogether (Gilbert et al., 2002; Hanke et al., 2003). Although successful, this strategy again relies on immunization with different recombinant viruses to overcome the problem of preexisting immunity to viral antigens. However, the development of alphaviral vectors may offer one solution, since removal of genes encoding viral structural proteins reduces vector immunogenicity, allowing the same vaccine to be used on multiple occasions, for both the priming and boosting phases (Hanke et al., 2003) (see earlier). 8. RNA Vaccines Vaccination with antigen-encoding RNA is a relatively new concept. It is possible to induce immune responses by direct in vivo delivery of RNA-encoding single antigens (Conry et al., 1995; Hoerr et al., 2000; Martinon et al., 1993; Qiu et al., 1996; Wolff et al., 1990; Zhou et al., 1999), entire virus (Vassilev et al., 2001), or in vitro-generated self-replicating recombinant RNA of alphaviruses (discussed earlier). Most recent work has, however, focused on using RNA to transfect DCs in vitro for subsequent vaccination (Mitchell and Nair, 2000; Sullenger and Gilboa, 2002). Dendritic cells are the most potent professional antigen-presenting cells, and clinical grade DCs are readily available (Fong and Engleman, 2000). To deliver antigens to DCs, peptides, proteins, DNA, tumor lysates, or even cell fusion have been used. Transfecting DCs with antigenencoding mRNA offers another simple but efficient method to load DCs with antigens (Boczkowski et al., 1996). RNA transfection can be achieved by incubation of DCs with RNA as cationic lipid complex or in naked form, and electroporation can further improve transfection efficiency (Saeboe-Larssen et al., 2002). mRNA can be isolated from tumor cells or synthesized in vitro from cDNA templates. Vaccination with tumor RNA-transfected DCs can induce CTL responses and antitumor immunity in several preclinical murine models (Ashley et al., 1997; Boczkowski et al., 1996, 2000; Granstein et al., 2000; Koido et al., 2000; Nair et al., 2000; Zhang et al., 1999). DCs generated from healthy human subjects or from cancer patients and transfected with tumor RNA can also activate CTLs and kill tumor cells in vitro (Boczkowski et al., 2000; Heiser et al., 2000, 2001a, 2001b; Nair et al., 1998, 2000; Saeboe-Larssen et al., 2002; Strobel et al., 2000; Su et al., 2001; Thornburg et al., 2000; Van Tendeloo et al., 2001; Weissman et al., 2000). A recent phase I clinical trial in patients with metastatic prostate cancer showed that vaccination with prostatespecific antigen (PSA) RNA-transfected DCs elicited PSA-specific T cell responses in all patients, and significantly reduced serum PSA velocities in some patients (Heiser et al., 2002). Apart from efficient antigen loading, another potential advantage of RNA transfection of DCs is that RNA encoding single
84
FREDA K. STEVENSON ET AL.
or multiple tumor antigens can be isolated and amplified from a small number of tumor cells. RNA transfection may thus represent a convenient and powerful technology for DC-based tumor vaccines.
IV. Concluding Summary
Tumor vaccines able to deliver specific antigen to the immune system are now available and are beginning to stand the test of performance in human subjects. Target antigens are being defined, and sophisticated methods to measure induced specific responses are available. Initially, therefore, outcome is measurable in terms of immune parameters. This allows rapid evaluation of one level of efficacy prior to assessing clinical value. Our task is to rouse immunity against weak or tolerizing tumor antigens, and, in order to succeed, we have to build vaccines that contain not only antigen, but also additional essential components. Adjuvants to activate innate immunity are clearly required, and the definition of the Toll-like receptors is opening possibilities for selective stimulation. Delivery of antigen to dendritic cells is a second goal and can be achieved either directly or via gene-based vaccines. Finally, there is a perceived need to activate high levels of T cell help to promote and maintain antibody, CD4þ , or CD8þ effector pathways. The easiest way to build components into a vaccine is to use genetic technology, and therefore gene-based vaccines are likely to have a real future. Physical means to improve delivery and potential combinations with viral vector-mediated delivery may be required to optimize performance. Opportunities to codeliver genes encoding cytokines, chemokines, and other molecules are there. It will take time to exploit the new genetic information and technology, and tailoring of vaccines for specific target antigens will be required. Provided flexible and rapid evaluation in pilot clinical trials is allowed at reasonable cost, design will progress to the stage where tumor vaccines will be a reality. Vaccination of patients is likely to succeed mainly in the setting of minimal residual disease. However, transfer of immunity from vaccinated donors of transplants, or from cells expanded in vitro, should also have a place. Vaccination could enhance the already successful passive immunotherapies being used to attack residual leukemic cells or dangerous viruses in immunosuppressed patients. A successful immune response should provide continuous surveillance against emergent tumor, and this would be a major contribution to the treatment of cancer.
Acknowledgments We thank the Leukaemia Research Fund, Tenovus, and Cancer Research UK for their support.
TUMOR VACCINES
85
References Abrams, S. I., and Schlom, J. (2000). Rational antigen modification as a strategy to upregulate or downregulate antigen recognition. Curr. Opin. Immunol. 12, 85–91. Akbari, O., Panjwani, N., Garcia, S., Tascon, R., Lowrie, D., and Stockinger, B. (1999). DNA vaccination: Transfection and activation of dendritic cells as key events for immunity. J. Exp. Med. 189, 169–178. Albert, M. L., Sauter, B., and Bhardwaj, N. (1998). Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs. Nature 392, 86–89. Alcami, A., and Koszinowski, U. H. (2000). Viral mechanisms of immune evasion. Mol. Med. Today 6, 365–372. Algarra, I., Collado, A., and Garrido, F. (1997). Altered MHC class I antigens in tumors. Int. J. Clin. Lab. Res. 27, 95–102. Altieri, D. C. (2001). The molecular basis and potential role of survivin in cancer diagnosis and therapy. Trends Mol. Med. 7, 542–547. Altieri, D. C. (2003). Validating survivin as a cancer therapeutic target. Nat. Rev. Cancer 3, 46–54. Amara, R. R., Villinger, F., Altman, J. D., Lydy, S. L., O’Neil, S. P., Staprans, S. I., Montefiori, D. C., Xu, Y., Herndon, J. G., Wyatt, L. S., Candido, M. A., Kozyr, N. L., Earl, P. L., Smith, J. M., Ma, H. L., Grimm, B. D., Hulsey, M. L., Miller, J., McClure, H. M., McNicholl, J. M., Moss, B., and Robinson, H. L. (2001). Control of a mucosal challenge and prevention of AIDS by a multiprotein DNA/MVA vaccine. Science 292, 69–74. Apostolopoulos, V., Osinski, C., and McKenzie, I. F. (1998). MUC1 cross-reactive Gal alpha(1,3)Gal antibodies in humans switch immune responses from cellular to humoral. Nat. Med. 4, 315–320. Asea, A., Kraeft, S. K., Kurt-Jones, E. A., Stevenson, M. A., Chen, L. B., Finberg, R. W., Koo, G. C., and Calderwood, S. K. (2000). HSP70 stimulates cytokine production through a CD14-dependent pathway, demonstrating its dual role as a chaperone and cytokine. Nat. Med. 6, 435–442. Ashley, D. M., Faiola, B., Nair, S., Hale, L. P., Bigner, D. D., and Gilboa, E. (1997). Bone marrowgenerated dendritic cells pulsed with tumor extracts or tumor RNA induce antitumor immunity against central nervous system tumors. J. Exp. Med. 186, 1177–1182. Babiuk, S., Baca-Estrada, M. E., Foldvari, M., Storms, M., Rabussay, D., Widera, G., and Babiuk, L. A. (2002). Electroporation improves the efficacy of DNA vaccines in large animals. Vaccine 20, 3399–3408. Basu, S., and Srivastava, P. K. (2000). Heat shock proteins: The fountainhead of innate and adaptive immune responses. Cell Stress Chaperones 5, 443–451. Basu, S., Binder, R. J., Suto, R., Anderson, K. M., and Srivastava, P. K. (2000). Necrotic but not apoptotic cell death releases heat shock proteins, which deliver a partial maturation signal to dendritic cells and activate the NF-kappa B pathway. Int. Immunol. 12, 1539–1546. Basu, S., Binder, R. J., Ramalingam, T., and Srivastava, P. K. (2001). CD91 is a common receptor for heat shock proteins gp96, hsp90, hsp70, and calreticulin. Immunity 14, 303–313. Bauer, S., Kirschning, C. J., Hacker, H., Redecke, V., Hausmann, S., Akira, S., Wagner, H., and Lipford, G. B. (2001). Human TLR9 confers responsiveness to bacterial DNA via speciesspecific CpG motif recognition. Proc. Natl. Acad. Sci. USA 98, 9237–9242. Bendandi, M., Gocke, C. D., Kobrin, C. B., Benko, F. A., Sternas, L. A., Pennington, R., Watson, T. M., Reynolds, C. W., Gause, B. L., Duffey, P. L., Jaffe, E. S., Creekmore, S. P., Longo, D. L., and Kwak, L. W. (1999). Complete molecular remissions induced by patient-specific vaccination plus granulocyte-monocyte colony-stimulating factor against lymphoma. Nat. Med. 5, 1171–1177. Berinstein, N. L. (2002). Carcinoembryonic antigen as a target for therapeutic anticancer vaccines: A review. J. Clin. Oncol. 20, 2197–2207.
86
FREDA K. STEVENSON ET AL.
Berzofsky, J. A., Ahlers, J. D., and Belyakov, I. M. (2001). Strategies for designing and optimizing new generation vaccines. Nat. Rev. Immunol. 1, 209–219. Bigey, P., Bureau, M. F., and Scherman, D. (2002). In vivo plasmid DNA electrotransfer. Curr. Opin. Biotechnol. 13, 443–447. Biragyn, A., Surenhu, M., Yang, D., Ruffini, P. A., Haines, B. A., Klyushnenkova, E., Oppenheim, J. J., and Kwak, L. W. (2001). Mediators of innate immunity that target immature, but not mature, dendritic cells induce antitumor immunity when genetically fused with nonimmunogenic tumor antigens. J. Immunol. 167, 6644–6653. Biragyn, A., Belyakov, I. M., Chow, Y. H., Dimitrov, D. S., Berzofsky, J. A., and Kwak, L. W. (2002a). DNA vaccines encoding human immunodeficiency virus-1 glycoprotein 120 fusions with proinflammatory chemoattractants induce systemic and mucosal immune responses. Blood 100, 1153–1159. Biragyn, A., Ruffini, P. A., Leifer, C. A., Klyushnenkova, E., Shakhov, A., Chertov, O., Shirakawa, A. K., Farber, J. M., Segal, D. M., Oppenheim, J. J., and Kwak, L. W. (2002b). Toll-like receptor 4-dependent activation of dendritic cells by beta-defensin 2. Science 298, 1025–1029. Bitton, R. J., Guthmann, M. D., Gabri, M. R., Carnero, A. J., Alonso, D. F., Fainboim, L., and Gomez, D. E. (2002). Cancer vaccines: An update with special focus on ganglioside antigens. Oncol. Rep. 9, 267–276. Blachere, N. E., Li, Z., Chandawarkar, R. Y., Suto, R., Jaikaria, N. S., Basu, S., Udono, H., and Srivastava, P. K. (1997). Heat shock protein-peptide complexes, reconstituted in vitro, elicit peptide-specific cytotoxic T lymphocyte response and tumor immunity. J. Exp. Med. 186, 1315–1322. Blanchard, T. J., Alcami, A., Andrea, P., and Smith, G. L. (1998). Modified vaccinia virus Ankara undergoes limited replication in human cells and lacks several immunomodulatory proteins: Implications for use as a human vaccine. J. Gen. Virol. 79(pt. 5), 1159–1167. Boczkowski, D., Nair, S. K., Snyder, D., and Gilboa, E. (1996). Dendritic cells pulsed with RNA are potent antigen-presenting cells in vitro and in vivo. J. Exp. Med. 184, 465–472. Boczkowski, D., Nair, S. K., Nam, J. H., Lyerly, H. K., and Gilboa, E. (2000). Induction of tumor immunity and cytotoxic T lymphocyte responses using dendritic cells transfected with messenger RNA amplified from tumor cells. Cancer Res. 60, 1028–1034. Bogen, B. (1996). Peripheral T cell tolerance as a tumor escape mechanism: Deletion of CD4þ T cells specific for a monoclonal immunoglobulin idiotype secreted by a plasmacytoma. Eur. J. Immunol. 26, 2671–2679. Bogen, B., Dembic, Z., and Weiss, S. (1993). Clonal deletion of specific thymocytes by an immunoglobulin idiotype. EMBO J. 12, 357–363. Boon, T., Cerottini, J. C., Van den Eynde, B., van der Bruggen, P., and Van Pel, A. (1994). Tumor antigens recognized by T lymphocytes. Annu. Rev. Immunol. 12, 337–365. Borrello, I., Sotomayor, E. M., Rattis, F. M., Cooke, S. K., Gu, L., and Levitsky, H. I. (2000). Sustaining the graft-versus-tumor effect throughposttransplant immunization with granulocyte-macrophage colony-stimulating factor (GM-CSF)-producing tumor vaccines. Blood 95, 3011–3019. Bourgeois, C., Rocha, B., and Tanchot, C. (2002). A role for CD40 expression on CD8þ T cells in the generation of CD8þ T cell memory. Science 297, 2060–2063. Boyle, J. S., Brady, J. L., and Lew, A. M. (1998). Enhanced responses to a DNA vaccine encoding a fusion antigen that is directed to sites of immune induction. Nature 392, 408–411. Brossart, P., Wirths, S., Stuhler, G., Reichardt, V. L., Kanz, L., and Brugger, W. (2000). Induction of cytotoxic T-lymphocyte responses in vivo after vaccinations with peptide-pulsed dendritic cells. Blood 96, 3102–3108. Burnett, F. M. (1957). Cancer—a biological approach. III. Viruses associated with neoplastic conditions. Br. Med. J. 1, 841–847.
TUMOR VACCINES
87
Busch, D. H., and Pamer, E. G. (1999). T cell affinity maturation by selective expansion during infection. J. Exp. Med. 189, 701–710. Casares, S., Inaba, K., Brumeanu, T. D., Steinman, R. M., and Bona, C. A. (1997). Antigen presentation by dendritic cells after immunization with DNA encoding a major histocompatibility complex class II-restricted viral epitope. J. Exp. Med. 186, 1481–1486. Castiglioni, P., Lu, C., Lo, D., Croft, M., Langlade-Demoyen, P., Zanetti, M., and Gerloni, M. (2003). CD4 T cell priming in dendritic cell-deficient mice. Int. Immunol. 15, 127–136. Celis, E. (2002). Getting peptide vaccines to work: Just a matter of quality control? J. Clin. Invest. 110, 1765–1768. Celluzzi, C. M., Mayordomo, J. I., Storkus, W. J., Lotze, M. T., and Falo, L. D., Jr. (1996). Peptide-pulsed dendritic cells induce antigen-specific CTL-mediated protective tumor immunity. J. Exp. Med. 183, 283–287. Chang, M. H., Chen, C. J., Lai, M. S., Hsu, H. M., Wu, T. C., Kong, M. S., Liang, D. C., Shau, W. Y., and Chen, D. S. (1997). Universal hepatitis B vaccination in Taiwan and the incidence of hepatocellular carcinoma in children. Taiwan Childhood Hepatoma Study Group. N. Engl. J. Med. 336, 1855–1859. Chaplin, P. J., De Rose, R., Boyle, J. S., McWaters, P., Kelly, J., Tennent, J. M., Lew, A. M., and Scheerlinck, J. P. (1999). Targeting improves the efficacy of a DNA vaccine against Corynebacterium pseudotuberculosis in sheep. Infect. Immun. 67, 6434–6438. Chattergoon, M. A., Kim, J. J., Yang, J. S., Robinson, T. M., Lee, D. J., Dentchev, T., Wilson, D. M., Ayyavoo, V., and Weiner, D. B. (2000). Targeted antigen delivery to antigen-presenting cells including dendritic cells by engineered Fas-mediated apoptosis. Nat. Biotechnol. 18, 974–979. Chen, C. H., Wang, T. L., Hung, C. F., Yang, Y., Young, R. A., Pardoll, D. M., and Wu, T. C. (2000). Enhancement of DNA vaccine potency by linkage of antigen gene to an HSP70 gene. Cancer Res. 60, 1035–1042. Chen, W., Syldath, U., Bellmann, K., Burkart, V., and Kolb, H. (1999). Human 60-kDa heat-shock protein: A danger signal to the innate immune system. J. Immunol. 162, 3212–3219. Cheng, W. F., Hung, C. F., Chai, C. Y., Hsu, K. F., He, L., Ling, M., and Wu, T. C. (2001a). Tumorspecific immunity and antiangiogenesis generated by a DNA vaccine encoding calreticulin linked to a tumor antigen. J. Clin. Invest. 108, 669–678. Cheng, W. F., Hung, C. F., Hsu, K. F., Chai, C. Y., He, L., Ling, M., Slater, L. A., Roden, R. B., and Wu, T. C. (2001b). Enhancement of Sindbis virus self-replicating RNA vaccine potency by targeting antigen to endosomal/lysosomal compartments. Hum. Gene Ther. 12, 235–252. Cheng, W. F., Hung, C. F., Hsu, K. F., Chai, C. Y., He, L., Polo, J. M., Slater, L. A., Ling, M., and Wu, T. C. (2002). Cancer immunotherapy using Sindbis virus replicon particles encoding a VP22-antigen fusion. Hum. Gene Ther. 13, 553–568. Ciernik, I. F., Berzofsky, J. A., and Carbone, D. P. (1996). Induction of cytotoxic T lymphocytes and antitumor immunity with DNA vaccines expressing single T cell epitopes. J. Immunol. 156, 2369–2375. Clarke, S. R. (2000). The critical role of CD40/CD40L in the CD4-dependent generation of CD8þ T cell immunity. J. Leukoc. Biol. 67, 607–614. Colmenero, P., Liljestrom, P., and Jondal, M. (1999). Induction of P815 tumor immunity by recombinant Semliki Forest virus expressing the P1A gene. Gene Ther. 6, 1728–1733. Condon, C., Watkins, S. C., Celluzzi, C. M., Thompson, K., and Falo, L. D., Jr. (1996). DNA-based immunization by in vivo transfection of dendritic cells. Nat. Med. 2, 1122–1128. Conry, R. M., LoBuglio, A. F., Wright, M., Sumerel, L., Pike, M. J., Johanning, F., Benjamin, R., Lu, D., and Curiel, D. T. (1995). Characterization of a messenger RNA polynucleotide vaccine vector. Cancer Res. 55, 1397–1400.
88
FREDA K. STEVENSON ET AL.
Cooney, E. L., Collier, A. C., Greenberg, P. D., Coombs, R. W., Zarling, J., Arditti, D. E., Hoffman, M. C., Hu, S. L., and Corey, L. (1991). Safety of and immunological response to a recombinant vaccinia virus vaccine expressing HIV envelope glycoprotein. Lancet 337, 567–572. Corr, M., Lee, D. J., Carson, D. A., and Tighe, H. (1996). Gene vaccination with naked plasmid DNA: Mechanism of CTL priming. J. Exp. Med. 184, 1555–1560. Corr, M., von Damm, A., Lee, D. J., and Tighe, H. (1999). In vivo priming by DNA injection occurs predominantly by antigen transfer. J. Immunol. 163, 4721–4727. Daemen, T., Pries, F., Bungener, L., Kraak, M., Regts, J., and Wilschut, J. (2000). Genetic immunization against cervical carcinoma: Induction of cytotoxic T lymphocyte activity with a recombinant alphavirus vector expressing human papillomavirus type 16 E6 and E7. Gene Ther. 7, 1859–1866. Darnell, R. B., and Posner, J. B. (2003). Observing the invisible: Successful tumor immunity in humans. Nat. Immunol. 4, 201. Deliyannis, G., Boyle, J. S., Brady, J. L., Brown, L. E., and Lew, A. M. (2000). A fusion DNA vaccine that targets antigen-presenting cells increases protection from viral challenge. Proc. Natl. Acad. Sci. USA 97, 6676–6680. Delogu, G., Howard, A., Collins, F. M., and Morris, S. L. (2000). DNA vaccination against tuberculosis: Expression of a ubiquitin-conjugated