SERIES EDITORS
STEPHEN G. WAXMAN Bridget Marie Flaherty Professor of Neurology Neurobiology, and Pharmacology; Director, Center for Neuroscience & Regeneration/Neurorehabilitation Research Yale University School of Medicine New Haven, Connecticut USA
DONALD G. STEIN Asa G. Candler Professor Department of Emergency Medicine Emory University Atlanta, Georgia USA
DICK F. SWAAB Professor of Neurobiology Medical Faculty, University of Amsterdam; Leader Research team Neuropsychiatric Disorders Netherlands Institute for Neuroscience Amsterdam The Netherlands
HOWARD L. FIELDS Professor of Neurology Endowed Chair in Pharmacology of Addiction Director, Wheeler Center for the Neurobiology of Addiction University of California San Francisco, California USA
Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands Linacre House, Jordan Hill, Oxford OX2 8DP, UK 360 Park Avenue South, New York, NY 10010-1710 First edition 2011 Copyright Ó 2011 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (þ44) (0) 1865 843830; fax (þ44) (0) 1865 853333; email:
[email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-444-53825-3 ISSN: 0079-6123
For information on all Elsevier publications visit our website at elsevierdirect.com
Printed and bound in Great Britain 11 12 13 14
10 9 8
7
6 5
4
3
2 1
List of Contributors O. Alluin, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada J.-E. Andujar, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montréal, Québec, Canada L. Avivi-Arber, Department of Prosthodontics, Faculty of Dentistry, University of Toronto, Ontario, Canada H. Barbeau, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada G. Barrière, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada D. Barthélemy, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada M. Bélanger, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada R. Bos, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France P. Boulenguez, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France L. J. Bouyer, Center for Interdisciplinary Research in Rehabilitation and Social Integration (CIRRIS), Department of Rehabilitation, Université Laval and Member of the Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Quebec City, Quebec, Canada H. Bras, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France C. Brocard, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France F. Brocard, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France v
vi
E. Brustein, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada C. Chau, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada M.-P. Côté, Department of Neurobiology and Anatomy, Drexel University College of Medicine, Philadelphia, USA A. Charlesworth, Center for Translational Neuroscience, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA P. Coulon, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France A. Doi, Center for Integrative Brain Research, Seattle Children's Research Institute and Department of Neurological Surgery, University of Washington, Seattle, Washington, USA T. Drew, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montréal, Québec, Canada R. Dubuc, Département de kinanthropologie, Université du Québec à Montréal and Groupe de Recherche sur le Système Nerveux Central, Département de physiologie, Université de Montréal, Montréal, Québec, Canada J. L. Feldman, Department of Neurobiology, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, California, USA A. Frigon, Département de physiologie et biophysique, Université de Sherbrooke, Sherbrooke, Quebec, Canada E. Garcia-Rill, Center for Translational Neuroscience, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA A. J. Garcia III, Center for Integrative Brain Research, Seattle Children’s Research Institute and Department of Neurological Surgery, University of Washington, Seattle, Washington, USA K. Garrison, Center for Translational Neuroscience, University of Arkansas for Medical Sciences, Little Rock and Department of Physical Therapy, University of Central Arkansas, Conway, Arkansas, USA N. Giroux, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada J.-P. Gossard, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montreal, Quebec, Canada S. Grillner, Nobel Institute for Neurophysiology, Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden L. M. Jordan, Department of Physiology, Spinal Cord Research Centre, University of Manitoba, Winnipeg MB, Canada L. Juvin, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montréal, Québec, Canada and Laboratoire Mouvement Adaptation Cognition, Université de Bordeaux, CNRS, Bordeaux, France H. Koch, Center for Integrative Brain Research, Seattle Children's Research Institute and Department of Neurological Surgery, University of Washington, Seattle, Washington, USA K. Lajoie, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montréal, Québec, Canada
vii
C. Langlet, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada D. Le Ray, Laboratoire Mouvement Adaptation Cognition, Université de Bordeaux, CNRS, Bordeaux, France H. Leblond, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montreal, Quebec, Canada J.-C. Lee, Department of Oral Physiology, Faculty of Dentistry, University of Toronto, Ontario, Canada S. Liabeuf, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France J. P. Lundw, Faculty of Dentistry, McGill University, and Groupe de Recherche sur le Système Nerveux Central, Université de Montréal, Montreal, Quebec, Canada J. Marcoux, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada D. S. Marigold, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, Canada M. Martinez, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada A. Ménard, Department of Physiology, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA P. Noué, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montreal, Quebec, Canada E. Pearlstein, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France J. Provencher, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada J.-M. Ramirez, Center for Integrative Brain Research, Seattle Children's Research Institute and Department of Neurological Surgery, University of Washington, Seattle, Washington, USA N. B. Reese, Department of Physical Therapy, University of Central Arkansas, Conway, Arkansas, USA S. Rossignol, Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada D. Ryczko, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montréal, Québec, Canada U. Sławi nska, Laboratory of Neuromuscular Plasticity, Department of Neurophysiology, Nencki Institute of Experimental Biology PAS, Warsaw, Poland
w
Deceased
viii
K. Sadlaoud, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France B. J. Sessle, Faculty of Dentistry, University of Toronto, Toronto, Ontario, Canada J. Sirois, Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montreal, Quebec, Canada A. Stil, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France S. Tazerart, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France J.-C. Viemari, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France L. Vinay, Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France S. Yakovenko, Département de Physiologie, Université de Montréal, Pavillon Paul-G. Desmarais, Montreal, Quebec, Canada C. Yates, Center for Translational Neuroscience, University of Arkansas for Medical Sciences, Little Rock and Department of Physical Therapy, University of Central Arkansas, Conway, Arkansas, USA S. Zanella, Center for Integrative Brain Research, Seattle Children's Research Institute and Department of Neurological Surgery, University of Washington, Seattle, Washington, USA
Preface This book, divided into two volumes of Progress in Brain Research, is largely inspired from presentations made at the 31st International Symposium of Research Group on the Central Nervous System at the University of Montreal held on May 4–5, 2009. The meeting was a special occasion to honor three outstanding neuroscientists, who have been world leaders in their respective fields of motor control, James P. Lund, Serge Rossignol, and Jack L. Feldman. The three honored scientists gave plenary presentations summarizing part of their outstanding accomplishments. In December 2009, James P. Lund prematurely and suddenly deceased. We were stunned and greatly moved by this sad news. We unanimously decided to dedicate this book to his memory. The symposium and this book highlight new findings on the neural control of rhythmic movements with an emphasis on common neuronal mechanisms. The book is divided into four sections from genes and molecules to system physiology. It begins with an historical overview by François Clarac who relates the early research carried out by true pioneers using crude and simple methods to study breathing, walking, and chewing. The first section describes how recent molecular genetics has revolutionized the study of the neuronal networks controlling rhythmic motor behaviors. The second section digs deep into ionic and cellular mechanisms underlying the function of different rhythmic networks. The third section covers broadly the modulation and the plasticity of rhythmic circuits, from chloride homeostasis to spasticity in human subjects. Finally, the 4th and last section, introduced by Dr. Sten Grillner, relates the ideas, contribution, and spirits of the three honorees, Professors Lund, Feldman, and Rossignol. The scope of the knowledge covered here is nothing but breathtaking. From discoveries of the nineteenth century to “in press” material, from zebra fish to humans, with students or senior scientists, rhythmogenesis and its control is explained and discussed, reviewed, and summarized. We thank all the 89 authors who contributed to this book. Jean-Pierre Gossard Réjean Dubuc Arlette Kolta
ix
In Memoriam
James P. Lund (Jim) sadly deceased on December 8, 2009, leaving an enormous hole in the personal and professional lives of many of us. Jim graduated as a dentist in Australia in 1966 from University of Adelaide. With the exceptional intellectual curiosity that characterized him, he gave up private practice after a year to begin a Ph.D. under the supervision of Peter Dellow. The concept of “central pattern generators” (CPGs) was still debated in those years and the “Sherringtonian” view that mastication consisted in a series of alternating jaw opening and jaw closing reflexes prevailed in many laboratories. Jim's pioneering work during his Ph.D. concluded to the existence of a CPG for mastication in the brainstem. After receiving his Ph.D. in 1971, he joined the Medical Research council (MRC) group of the University of Montreal, first as a postdoctoral fellow with Yves Lamarre and later as a faculty member. These were determinant years in Jim's career, and this is when he established himself as an inevitable authority in the field of mastication. His laboratory was recognized as the world's leading center for fundamental studies of orofacial motor function. He asked important questions and touched almost every aspect involved in the neural control of mastication. After studying the cortical influence on the brainstem CPG, he became interested in reflex modulation during movements and showed that reflexes were phasically modulated to avoid perturbation of the movement while maintaining protection of the tissues. With K. Olsson and later K.-G. Westberg, he described the role and connections of brainstem interneurons involved in patterning mastication and modulating jaw reflexes. xi
xii
It had been proposed that one way to modulate reflexes was by presynaptic inhibition and antidromic discharges evoked in sensory afferents. He investigated this issue in the trigeminal system and with his collaborators subsequently discovered that the propagation of antidromic discharges was controlled by GABAergic synapses along the axonal trunk causing a compartmentalization of the axon. Jim was also interested on how sensory inputs interacted with the CPG. Most of his work on this issue, conducted on animal models, eventually led him to formulate a conceptual model about how sensory inputs from nociceptors altered movements. In collaboration with C. Stohler, he developed the pain adaptation model, which shows how pain itself can cause motor and sensory symptoms. This model has been validated by several clinical research groups throughout the world and is prompting changes in clinical practice. Nociceptive inputs are not the only sensory inputs susceptible to alter movements. In an effort at understanding how other types of sensory inputs affect mastication, he worked in close collaboration with J. Feine to document the effects of loss of inputs from teeth and of patient's treatment on the efficiency of mastication and on the consequences on nutrition. More than a highly prolific and respected scientist, Jim was also a tireless visionary who led many battles to improve research in dental faculties and increase knowledge transfer toward clinical applications. For this reason, he generously accepted to serve as Vice Dean of Research at University of Montreal for 8 years and as Dean of Dentistry at McGill University for 13 years. Throughout those years, he founded research networks and built a world leader pain research center at McGill. Jim never failed to consider issues from different perspectives and to think outside the box. He influenced many institutional and health care decisions, but even more importantly, he left an indelible trace in the life of many scientists that he trained or helped recruiting. He deeply cared for his trainees, collaborators, and colleagues. He always managed to provide his trainees with what great emulation needed to foster their growth. He was an outstanding mentor to people at all levels of career development from undergraduate and postgraduate students to junior academic staff and peers in administrative positions. We will miss Jim deeply for his insights, brightness, contagious energy, and genuine caring for others. Jean-Pierre Gossard Réjean Dubuc Arlette Kolta
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 1
Importance of chloride homeostasis in the operation of rhythmic motor networks Jean-Charles Viemari, Rémi Bos, Pascale Boulenguez, Cécile Brocard, Frédéric Brocard, Hélène Bras, Patrice Coulon, Sylvie Liabeuf, Edouard Pearlstein, Karina Sadlaoud, Aurélie Stil, Sabrina Tazerart and Laurent Vinay* Laboratoire Plasticité et Physio-Pathologie de la Motricité (UMR6196), Centre National de la Recherche Scientifique (CNRS) & Aix-Marseille Université, 31 Chemin Joseph Aiguier, Marseille Cedex 20, France
Abstract: GABA and glycine are classically called “inhibitory” amino acids, despite the fact that their action can rapidly switch from inhibition to excitation and vice versa. The postsynaptic action depends on the intracellular concentration of chloride ions ([Cl]i), which is regulated by proteins in the plasma membrane: the Kþ–Cl cotransporter KCC2 and the Naþ–Kþ–Cl cotransporter NKCC1, which extrude and intrude Cl ions, respectively. A high [Cl]i leads to a depolarizing (excitatory) action of GABA and glycine, as observed in mature dorsal root ganglion neurons and in motoneurons both early during development and in several pathological conditions, such as following spinal cord injury. Here, we review some recent data regarding chloride homeostasis in the spinal cord and its contribution to network operation involved in locomotion. Keywords: inhibition; networks; KCC2; chloride; activity.
glycine receptors depends on the intracellular concentration of chloride ions ([Cl]i) in the target cell. In adult healthy neurons, the activation of GABAA and glycine receptors results in an inward flux of Cl and membrane potential hyperpolarization. Therefore, the inhibitory action of glycine and GABA consists in both shunting incoming excitatory currents and moving the membrane potential away from the action potential threshold. This “classical”
Introduction Synaptic inhibition mediated by GABA and glycine strongly modulates mammalian neuronal networks from the early life to adulthood. The postsynaptic action of these neurotransmitters on GABAA and
*Corresponding author. Tel.: (þ33)-4-91-16-40-86; Fax.: (þ33)-4-91-77-50-84 DOI: 10.1016/B978-0-444-53825-3.00006-1
3
4
hyperpolarizing inhibition is not observed in immature neurons; inhibitory postsynaptic potentials (IPSPs) as well as glycine- and GABA-evoked potentials are instead depolarizing and often excitatory (Gao and Ziskind-Conhaim, 1995; Takahashi, 1984; Wu et al., 1992; Ziskind-Conhaim, 1998), because of a high [Cl]i. The [Cl]i is regulated by transporters in the membrane (Delpire and Mount, 2002). Chloride homeostasis and the regulation of those transporters appear as an important emerging mechanism, by which the strength, as well as the polarity, of postsynaptic inhibition can be controlled, even in adult tissue. This review will center on these issues. Differential control of chloride homeostasis in primary afferents and motoneurons The transport of Cl by cation–chloride cotransporters is driven by the concentration gradients of cations (Payne et al., 2003; Fig. 1a). The Naþ gradient generated by the Naþ/KþATPase fuels the inward-directed Cl pump Naþ–Kþ–Cl cotransporter, NKCC1, which is important in active accumulation of intracellular Cl in immature neurons in several brain areas (Dzhala et al., 2005; Ikeda et al., 2003; Plotkin et al., 1997; Sun and Murali, 1999; Vardi et al., 2000). It is down-regulated with development except in a few neurons such as dorsal root ganglion (DRG) neurons (Sung et al., 2000). As shown in Fig. 1b, the transporter is predominantly located at the plasma membrane of these cells. By contrast, DRG neurons from a NKCC1 knockout animal demonstrate the complete absence of NKCC1 expression (Fig. 1c). Gramicidin perforated-patch recordings from DRG neurons isolated from these mice reveal that EGABA is consistently more depolarized in DRG neurons from wild type compared with neurons from NKCC1 knock-out mice ( 37 and 53 mV, respectively; Fig. 1d). This corresponds to intracellular Cl concentrations of 46 and 24 mM, respectively (Sung et al., 2000).
Inhibitory amino acid transmission to motoneurons undergoes marked changes during development. There is a developmental switch from predominantly long-duration GABAergic to short-duration glycinergic currents (Gao et al., 2001) and an increase in the density of the glycine currents (Gao and Ziskind-Conhaim, 1995). These physiological observations are correlated with a developmental down-regulation of GABAA receptors and a concomitant up-regulation of glycine receptors (Sadlaoud et al., 2010). A striking observation is related to chloride homeostasis. Glycine- and GABA-evoked potentials are depolarizing and often excitatory (Gao and Ziskind-Conhaim, 1995; Takahashi, 1984; Wu et al., 1992; Ziskind-Conhaim, 1998) early during development, because of a high [Cl]i that favors Cl efflux through GABAA- or glycineoperated Cl channels. For instance, a brief application of glycine onto the in vitro spinal cord isolated from fetal rats, at embryonic day (E) 15.5 (i.e., one week prior to birth), evokes excitatory responses that are abolished by strychnine (Nishimaru et al., 1996). The cation–chloride cotransporter NKCC1 (Delpy et al., 2008) and the anion exchanger AE3 (Gonzalez-Islas et al., 2009) play a significant role in accumulating chloride in immature motoneurons. With development, [Cl]i decreases, leading to a shift of the chloride equilibrium potential toward further negative values, and thereby to a change in glycine and GABA-evoked potentials from depolarization to hyperpolarization (Gao and Ziskind-Conhaim, 1995; Jean-Xavier et al., 2006; Stil et al., 2009; Takahashi, 1984; Vinay and Jean-Xavier, 2008). The up-regulation of the outward-directed Cl pump, the neuron specific Kþ–Cl cotransporter KCC2, is widely accepted to underlie the shift from GABA/glycine-induced depolarization to hyperpolarization in several regions of the central nervous system (Delpire and Mount, 2002; Payne et al., 2003; Rivera et al., 1999, 2004). In contrast to DRG neurons, mature motoneurons highly express this transporter (Fig. 1e; Boulenguez et al., 2010; Delpy et al., 2008; Hübner et al., 2001; Jean-
5 (a)
Cl–
2K+
Cl– K+ Na+ GABAAR & GlyR
GABAAR & GlyR
Out
Out
Na+/K+ ATPase
NKCC
KCC2
In
In ATP Cl– Depolarizing current
3Na+
(c)
(b)
NKCC1+/+ (d)
Hyperpolarizing current
(e)
NKCC1–/– (f)
0 NKCC1+/+
NKCC1–/–
–20
–40
0 –20
EC1 (mV)
EC1 (mV)
K+ Cl–
KCC2+/+
KCC2–/–
–40 –60
–60
–80 ∗∗∗
∗∗
Fig. 1. Different mechanisms underlie chloride homeostasis in dorsal root ganglion (DRG) neurons and motoneurons. (a) Schematic illustration (adapted from Payne et al., 2003) of the contribution of cation–chloride cotransporters to the regulation of [Cl]i. Under physiological conditions, NKCC1 and KCC2 cotransporters intrude and extrude Cl, respectively. Arrows indicate the direction of net transports, which are driven by the concentration gradients of cations, generated by the Naþ/Kþ-ATPase. As a result of NKCC1 activity in DRG neurons, activation of GABAA receptors generates a depolarizing current whereas the activity of KCC2 cotransporters in adult motoneurons underlies hyperpolarizing currents across anion-permeable channels. (b–d) Adapted from Sung et al. (2000). Control mouse DRG neurons highly express the NKCC1 protein (b). The signal is predominantly located at the cell membrane (arrows). Note the presence of a few neurons with a minimal amount of cotransporter expression (arrowhead). Immunofluorescence staining shows the absence of NKCC1 expression in DRG neurons from homozygote mutant mice (c). Scale bar, 20 mm. As a result of the lack of NKCC1, ECl is significantly more negative in NKCC1 knock-out mice, compared to controls (d). (e) Strong KCC2 expression in ventral horn of adult rats. Note the labeling of the plasma membrane (arrows) of motoneurons. (f) Application of a GABAA- and glycine-receptor agonists in the recording chamber induce a larger depolarization in lumbar motoneurons in KCC2 knock-out mice than in wild-type animals (Adpated from Hübner et al., 2001).
6
Xavier et al., 2006; Stein et al., 2004; Stil et al., 2009; Vinay and Jean-Xavier, 2008). Motoneurons are significantly more depolarized (> 30 mV) by GABA or glycine in KCC2 knock-out mice at E18 than in wild type ( 10 mV; Fig. 1f; Hübner et al., 2001). Altogether, these data demonstrate that NKCC1 and KCC2 cotransporters are important for maintaining high and low chloride concentrations in mature DRG neurons and motoneurons, respectively. Contribution of chloride homeostasis to cell excitability According to the classical view, postsynaptic inhibition induced by the activation of GABAA and glycine receptors consists in two mechanisms: shunting incoming excitatory currents and moving the membrane potential away from the action potential threshold. As already mentioned, this hyperpolarization from rest is not observed in immature spinal neurons (Gao and ZiskindConhaim, 1995; Takahashi, 1984; Wu et al., 1992; Ziskind-Conhaim, 1998), thereby raising the question of the effect of subthreshold depolarizing IPSPs on neuronal excitability. It is widely accepted that despite their depolarizing action, these potentials are inhibitory in immature spinal motoneurons because of the shunting mechanism. On the basis of two reports on the cortex (Gulledge and Stuart, 2003) and the hypothalamus (Gao et al., 1998), we challenged the idea that depolarizing IPSPs may interact with excitatory inputs and increase the excitability of motoneurons (Jean-Xavier et al., 2007). We demonstrated that depolarizing IPSPs have an excitatory action by facilitating action potential generation when paired with subthreshold excitatory inputs (Fig. 2a). The effect of depolarizing IPSPs on excitability therefore depends on the relative weights of the inhibitory action of the shunting mechanisms and the excitatory action of the depolarization. The value of ECl affects the strength of inhibitory connections within
spinal cord locomotor networks; the more positive ECl, the lower the efficacy of inhibition. A first key point is the timing between inhibitory and excitatory inputs. The inhibitory action of shunting is exerted over a rather short period after IPSP onset (Fig. 2a, blockade of action potentials evoked by current pulses). In contrast, the time course of the depolarization is much longer, because of the time constant of the membrane, such that the excitations occurring in the late phase of the depolarizing IPSP sum up with the depolarization whereas the conductance has returned to baseline (Fig. 2a; bottom trace showing a subthreshold current pulse triggering an action potential when occurring in the decay phase of the IPSP). A second key point, revealed by the use of a model (Fig. 2b), is the distance between inhibitory and excitatory synapses. The inhibitory action of conductance changes is local whereas the depolarization spreads electrotonically along the dendrites. Consequently, GABA/glycine synapses exert a strong inhibition on proximal excitatory inputs (Fig. 2c; somatic excitations and inhibitory synapses on the soma and on proximal dendrites) and a pure facilitation (whatever the timing) on distal excitation (Fig. 2c, bottom graph with inhibitory inputs 200 mm away from excitatory synapses). Contribution of chloride dynamics to network activity Maturation of chloride homeostasis affects network activity. There is a switch in the contribution of chloride-mediated conductances to spontaneous activity from excitation to inhibition during late gestation. A rhythmic spontaneous activity can be recorded in vitro very early ( E12–E14 in rodents), when many lumbar motoneurons are still migrating and extending their peripheral projections (Hanson and Landmesser, 2003). Electrical transmission plays a significant role in the generation of episodes since blockade of gap junction coupling by carbenoxolone abolishes
7 (a)
(b)
Inhibition
Exc.
IPSP (EIPSP = –56 mV)
soma
25 μm Inhibitory inputs
IPSP + current-evoked action potentials
200 μm
EC1 (mV)
(c) –60 soma
–70 –80
Facilitation
EC1 (mV)
0
50
100
150
–60 25 μm
–70 –80
IPSP + sub-threshold depolarizing pulses EC1 (mV)
0
20 mV 100 ms
50
100
150
Facilitation
–60
200 μm
–70 Inhibition –80 0
50
100
150
Time (ms)
Fig. 2. Inhibitory and excitatory actions of depolarizing inhibitory postsynaptic potentials (IPSPs). (a) Depolarizing IPSP evoked in a lumbar motoneuron (VREST: 68 mV) by electrical stimulation of the ventral funiculus of the spinal cord isolated from neonatal rat (top trace). Recordings were performed after blocking excitatory amino acid transmission. Suprathreshold current pulses were injected to test the ability of the inhibitory input to block action potentials (middle trace: 50 sweeps). (b) Model used in the simulations to test the interaction between depolarizing IPSPs and subthreshold excitatory inputs set on the soma. (c) Global pictures of inhibitory (inhibition of suprathreshold EPSPs) and excitatory (facilitation of subthreshold EPSPs) effects depending on both ECl and the location of inhibitory inputs. Note that IPSPs generated on the soma, close to the excitations are inhibitory, whatever the value of ECl, because of shunting mechanisms. For inhibitory inputs on proximal segments of dendrites (25 mm), the timing is important when ECl is set at depolarized values (> 60 mV); the initial part of the IPSP is inhibitory whereas the late phase can facilitate subthreshold excitations. The action of distal inhibitory inputs (200 mm) can switch from inhibition to excitation within a narrow range of ECl values, close to the resting membrane potential. (a–c) Adapted from Jean-Xavier et al. (2007).
the rhythmic episodes. The relative contribution and the role of the different neurotransmitters change prior to birth (Hanson and Landmesser, 2003; Myers et al., 2005; Ren and Greer, 2003). At the earliest stages, the spinal networks comprising cholinergic and glycinergic synaptic interconnections are capable of generating rhythmic activity, while GABAergic synapses play a role in supporting this activity. Stimulation of motor neurons at this stage elicits episodes of activity that propagate through the lumbar spinal cord suggesting that motoneurons make excitatory
connections on each other and on glycine/ GABAergic interneurons via nicotinic receptors. At a later stage (E16.5–E17.5), the spontaneous activity results from the combined action of nonNMDA (glutamatergic), nicotinic (acetylcholine), glycine, and GABAA receptors. Closer to birth (E18.5–E21.5), glutamate drive acting via nonNMDA receptors is primarily responsible for the rhythmic activity. Application of strychnine to block glycine receptors markedly reduces and increases the bursting frequency in early and late embryos, respectively, suggesting that glycine
8
switches from an excitatory to an inhibitory contribution to spontaneous activity. This transition occurs at about the same time as the switch from cholinergic to glutamatergic transmission. The fact that the classical inhibitory neurotransmitters can be functionally excitatory moves the balance between excitatory and inhibitory drives toward excitation. This renders developing networks hyperexcitable. Intracellular chloride concentration undergoes significant changes during spontaneous activity in the chick spinal cord (Chub and O'Donovan, 2001). After an episode, the chloride equilibrium potential falls by 10 mV, corresponding to a decline of intracellular chloride of 15 mM. This reduces the excitatory contribution of GABA and glycine and hence network excitability. The intracellular chloride concentration is restored during the interepisode interval, as a result presumably of inward chloride pumping via cation–chloride cotransporters, leading to a progressive increase in network excitability. The importance of chloride dynamics in the genesis of spontaneous activity in the in vitro chick spinal cord between E9 and E11 was confirmed by the use of a model (Marchetti et al., 2005). Application of serotonin on the spinal cord at birth, to activate the CPGs, evokes a fictive locomotor pattern consisting of alternation between the motor bursts on the left and right sides of the spinal cord, as well as alternation between flexor and extensor bursts on the same side. The same kind of experiments made five days prior to birth reveals a motor pattern with all bursts in phase (Iizuka et al., 1997; Nishimaru and Kudo, 2000). The left–right alternation appears two days later but rhythmic bursts still occur synchronously in flexors and extensors. The transition from synchrony to alternation in the left and right ventral roots is likely due to the maturation of chloride homeostasis. In the respiratory network, although GABA/ glycine-mediated inhibition is not essential for rhythm generation in neonatal rodent, it is well known that synaptic inhibition is important for
setting the different phases of the respiratory rhythm and regulates both the excitability of glutamatergic connections and the intrinsic membrane properties (Busselberg et al., 2001a,b; Ramirez et al., 1997; Shao and Feldman, 1997). As described in the case of spinal cord, the [Cl]i determines the functional role of inhibitory neurotransmitters. The transition from an excitatory to an inhibitory effect occurs earlier on inspiratory neurons than in the lumbar spinal cord ( E19; Ren and Greer, 2006). From this time onward, GABA and glycine hyperpolarize inspiratory neurons and suppress respiratory frequency. It is important to note that KCC2 cotransporters are sensitive to subtle changes in the extracellular concentration of potassium ions ([Kþ]o). A raise of [Kþ]o from 2–3 to 9–10 mM is sufficient to reverse the driving force for the net K–Cl cotransport, thus allowing KCC2 to operate in a reverse mode as a net influx pathway (Payne et al., 2003). As a consequence, such a raise induces a 10- to 20-mV depolarizing shift of ECl (Ren and Greer, 2006; Vinay and JeanXavier, 2008) in medullary slices and lumbar spinal cord preparations, respectively). Raising [Kþ]o is commonly used as a methodological tool to increase the excitability of in vitro preparations. This may, in addition, modify the strength of postsynaptic inhibition within networks and explain some inconsistency among results obtained in different experimental conditions. The effects of GABA on the respiratory network highly depend on the expression and the functionality of cation–chloride cotransporters. During the early life, NKCC1 is expressed at high levels and perturbations of its functionality by removing [Naþ]o or by application of bumetanide shifts ECl to a more hyperpolarized value and reverses the excitatory effect of GABAA receptor agonists (muscimol), whereas blockade of KCC2 by furosemide has no effect (Ren and Greer, 2006). In contrast, in neonatal preparations, furosemide depolarizes ECl and blocks the inhibitory effect of muscimol on respiratory frequency. At that time,
9
blockade of NKCC1 has no significant effect. These results are consistent with a developmental increase of KCC2 expression in the respiratory network, as in the locomotor system. Primary afferent depolarizations and antidromic discharges as part of the motor network It is well established that one form of presynaptic inhibition in the vertebrate spinal cord is associated with primary afferent depolarization (PAD, Alvarez-Leefmans et al., 1998; Rudomin, 1990; Rudomin et al., 1993) and that GABA, through the activation of GABAA receptors, plays a major role in the generation of PAD. Axo–axonic interactions between GABAergic terminals and primary afferents have been demonstrated (see Alvarez, 1998, for review). PADs are reduced by GABAA receptor antagonists (Curtis and Lodge, 1982; Eccles et al., 1963; Levy, 1975; Rudomin et al., 1981). When PADs are large enough to reach firing threshold, they trigger discharges that are antidromically conducted into peripheral nerves. This was first demonstrated by recording the reflex response elicited in dorsal roots, by electrical stimulation of an adjacent dorsal root (“dorsal root reflex,” Barron and Matthews, 1938; Toennies, 1938, see Kerkut and Bagust, 1995, for review). Antidromic discharges of primary afferents have been observed during locomotion in the cat (fictive locomotion: Beloozerova and Rossignol, 1999; Dubuc et al., 1988; Duenas et al., 1990; Gossard et al., 1991; treadmill locomotion: Beloozerova and Rossignol, 2004; Dubuc et al., 1985; Rossignol et al., 1998) and in the rat (Pilyavskii et al., 1988). A spontaneous antidromic activity in the dorsal roots has been described in the in vitro spinal cord isolated from adult (Bagust et al., 1989; Chen et al., 1993) or young hamsters (Abdul-Razzak et al., 1994; see Kerkut and Bagust, 1995, for review). Antidromic discharges are also recorded from lumbar dorsal
roots in the neonatal rat spinal cord in vitro (Fellippa-Marques et al., 2000; Kremer and Lev-Tov, 1998; Vinay and Clarac, 1999; Vinay et al., 1999). Most of these action potentials are blocked by bath application of bicuculline or picrotoxin, two GABAA receptor antagonists. Some activity occurs spontaneously, consisting of rhythmic bursts, and is therefore likely triggered by a centrally generated rhythm. A motor activity is sometimes recorded from ventral roots, occurring in phase with the dorsal root burst (Fig. 3a). Intracellular recordings from motoneurons show subthreshold depolarizations, confirming the existence of neuronal connections that coactivate the motoneurons and the primary afferent terminals. Alternatively, antidromic discharges in dorsal roots may have postsynaptic effects in lumbar motoneurons and first-order interneurons, as shown in the trigeminal system. Lund and coworkers indeed demonstrated the existence of such antidromic discharges during fictive mastication (Kolta et al., 1995; Verdier et al., 2003). They proposed the very elegant hypothesis of a functional compartmentalization of muscle spindle afferents (Fig. 3b). In contrast to other primary afferents, the cell bodies of primary afferents that innervate the spindles of jaw-closing muscles are located in the trigeminal mesencephalic nucleus (Nvmes) and not in DRG. During fictive mastication, the firing patterns recorded from the soma differ markedly from those recorded from the caudal compartment of the central axon of these afferents. In 65% of cells, activation of the motor circuits during fictive mastication does not alter tonic orthodromic activity induced by stretch of the jaw-closing muscles and in one-third of cases only, a phasic inhibition of this activity is observed, coincident with jaw-opening phase of the cycle. In contrast, phasic inhibition is seen in the great majority (83%) of recordings from the caudal compartment of the central axon. In addition, these inhibitions alternate with phasic excitation occurring in the jaw-closing phase. The
10 b
(a)
hypothetical synapses on the axon that stops antidromic potentials from reaching the cell body. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this chapter.)
L5 DR
L5 VR
L5 MN –69 mV 10 mV 1s (b) Soma
NV mes
65% 35%
Rostral compartment
JO JC
Masseter muscle spindle
A NV mt
Central axon 17 %
Caudal compartment
83 % JO JC
PAD
INs
Fig. 3. Antidromically propagated discharges triggered by primary afferent depolarizations. (a) Spontaneous bursting activity recorded from lumbar dorsal and ventral roots in the in vitro spinal cord preparation isolated from neonatal rats. (b) Functional compartmentalization of jaw-closing muscle spindle afferents, redrawn from Verdier et al. (2003). The cell bodies of the afferents are located in the trigeminal mesencephalic nucleus (NV mes) and have a long descending axon that gives off branches at many levels. Axonal branches terminate on jaw-closing motoneurons in the trigeminal motor nucleus (NV mt) and on interneurons (INs). During fictive mastication, stretch-induced tonic activity is unaltered in 65% of somatic recordings and is phasically inhibited during the jaw-opening (JO) phase in 35% of cases. In contrast, phasic inhibition during the JO phase alternates with a phasic excitation during the jaw-closing (JC) phase in the majority of recordings from caudal axons. Antidromic action potentials generated by PAD (in red) of the central terminals do not reach the soma but do reach trigeminal motoneurons and interneurons via collaterals. (a) represents
extraspikes appearing in the latter phase have been attributed to antidromic action potentials generated by PAD of the central terminals. These spikes do not reach the soma but do reach trigeminal motoneurons via collaterals. They provide additional excitation to trigeminal MNs. To summarize, the central axon plays the role of a premotor interneuron carrying signals from the CPG via a set of presynaptic terminals to motoneurons whereas the rostral portion of the neuron provides feedback from its receptors, although even this sensory feedback is phasically gated by the CPG. The possibility that a similar mechanism also applies to the spinal cord is under investigation.
Dysfunction of chloride homeostasis in pathological conditions The hyperpolarizing shift of ECl from above to below the resting membrane potential in rodent motoneurons occurs during perinatal development, a time window during which pathways descending from the brainstem arrive in the lumbar enlargement (Brocard et al., 1999; Vinay et al., 2000, 2002). A complete spinal cord transection was performed on the day of birth to investigate the contribution of descending pathways to the maturation of chloride homeostasis (Jean-Xavier et al., 2006). This early removal of supraspinal influences prevented both the hyperpolarizing shift of ECl and the up-regulation of KCC2 that normally occur during the first postnatal week. These results may account for the disorganization of the locomotor pattern that is observed in these animals following neonatal spinal cord transaction (Norreel et al., 2003).
11
It is well known that several inhibitory reflex mechanisms are decreased in patients after SCI (Boorman et al., 1996; Mazzocchio and Rossi, 1997; Morita et al., 2001). The mechanisms responsible for this reduced inhibition were unknown until recently. Based on the above-mentioned results showing that descending pathways modulate chloride homeostasis, we hypothesized that a spinal cord injury in adults may affect the expression of KCC2 in the lumbar spinal cord and hence the strength of postsynaptic inhibition. We showed that the expression of KCC2 in the plasma membrane of motoneurons is reduced after spinal cord injury (Boulenguez et al., 2010). We demonstrated that a blockade of these cotransporters in intact animals reproduces some of the observations made in spastic paraplegic rats, suggesting that a down-regulation of KCC2 can contribute to the hyperexcitability of spinal networks, which is the hallmark of spasticity following spinal cord injury (Boulenguez et al., 2010).
Conclusion To conclude, primary afferent terminal and motoneurons exhibit opposite mechanisms for chloride homeostasis. A high [Cl]i maintained by NKCC1 cotransporters is responsible for PADs. The action potentials that are generated by PADs reaching firing threshold have long been considered as an epiphenomenon or an artifact due to the experimental conditions. The ubiquity of their observation in different motor networks underscores the need to investigate their role further. A low [Cl]i is maintained in healthy motoneurons by KCC2 cotransporters; this is a requirement for a strong postsynaptic inhibition. The excitatory effects of GABA and glycine used to be considered only in a developmental perspective. The demonstration that KCC2 transporters and chloride homeostasis are affected in pathological conditions sheds new light on these mechanisms. Restoring chloride
homeostasis, by up-regulating the functionality of KCC2 transporters, may pave the way for new treatments aimed at reducing spasticity following spinal cord injury. The demonstration that depolarizing IPSPs can facilitate subthreshold excitations also warrants further studies to investigate how these interactions contribute to synaptic integration and plasticity. That [Kþ]o and ECl are critically dependent upon the activity raises the possibility that the role of GABA/glycinergic inputs may reversibly switch from inhibition to excitation, depending on the recent experience of the network. Acknowledgments Our study on the plasticity of inhibitory synaptic transmission in the spinal cord is supported by grants (to L.V.) from the French Agence Nationale pour la Recherche, the French Institut pour la Recherche sur la Moelle épinière et l'Encéphale (to L.V.) and the Christopher and Dana Reeve Foundation (VB1-0502-2 and VB2-08012). K.S. received a grant from the Association Française contre les Myopathies (Grant 13912). S.T. received a grant from the Fondation pour la Recherche Médicale (Grant FDT20081213783). References Abdul-Razzak, R., Bagust, J., & Kerkut, G. A. (1994). Postnatal changes in the dorsal root reflex in the isolated spinal cord of the hamster Mesocricetus auratus. Comparative Biochemistry and Physiology, 107C, 195–204. Alvarez, F. J. (1998). Anatomical basis for presynaptic inhibition of primary sensory fibers. In P. Rudomin, L. M. Romo & S. Mendell (Eds.), Presynaptic inhibition and neural control (pp. 13–49). Oxford: Oxford University Press. Alvarez-Leefmans, F. J., Nani, A., & Marquez, S. (1998). Chloride transport, osmotic balance, and presynaptic inhibition. In P. Rudomin, R. Romo & L. M. Mendell (Eds.), Presynaptic inhibition and neural control (pp. 50–79). Oxford: Oxford University Press. Bagust, J., Kerkut, G. A., & Rakkah, N. I. A. (1989). The dorsal root reflex in isolated mammalian spinal cord. Comparative Biochemistry and Physiology, 93A, 151–160.
12 Barron, D. H., & Matthews, B. H. C. (1938). Dorsal root reflexes. Journal of Physiology (London), 94, 26P–27P. Beloozerova, I., & Rossignol, S. (1999). Antidromic discharges in dorsal roots of decerebrate cats. I: Studies at rest and during fictive locomotion. Brain Research, 846, 87–105. Beloozerova, I. N., & Rossignol, S. (2004). Antidromic discharges in dorsal roots of decerebrate cats. II: Studies during treadmill locomotion. Brain Research, 996, 227–236. Boorman, G. I., Lee, R. G., Becker, W. J., & Windhorst, U. R. (1996). Impaired “natural reciprocal inhibition” in patients with spasticity due to incomplete spinal cord injury. Electroencephalography and Clinical Neurophysiology, 101, 84–92. Boulenguez, P., Liabeuf, S., Bos, R., Bras, H., Jean-Xavier, C., Brocard, C., et al. (2010). Down-regulation of the potassium-chloride cotransporter KCC2 contributes to spasticity after spinal cord injury. Nature Medicine, 16, 302–307. Brocard, F., Vinay, L., & Clarac, F. (1999). Gradual development of the ventral funiculus input to lumbar motoneurons in the neonatal rat. Neuroscience, 90, 1543–1554. Busselberg, D., Bischoff, A. M., Becker, K., Becker, C. M., & Richter, D. W. (2001). The respiratory rhythm in mutant oscillator mice. Neuroscience Letters, 316, 99–102. Busselberg, D., Bischoff, A. M., Paton, J. F., & Richter, D. W. (2001). Reorganisation of respiratory network activity after loss of glycinergic inhibition. Pflügers Archiv, 441, 444–449. Chen, Y., Bagust, J., Kerkut, G. A., & Tyler, A. W. (1993). Correlation between spontaneous bursts of activity recorded from the dorsal roots in an isolated hamster spinal cord. Experimental Physiology, 78, 811–824. Chub, N., & O'Donovan, M. J. (2001). Post-episode depression of GABAergic transmission in spinal neurons of the chick embryo. Journal of Neurophysiology, 85, 2166–2176. Curtis, D. R., & Lodge, D. (1982). The depolarization of feline ventral horn group Ia spinal afferent terminations by GABA. Experimental Brain Research, 46, 215–233. Delpire, E., & Mount, D. B. (2002). Human and murine phenotypes associated with defects in cation–chloride cotransport. Annual Review of Physiology, 64, 803–843. Delpy, A., Allain, A. E., Meyrand, P., & Branchereau, P. (2008). NKCC1 cotransporter inactivation underlies embryonic development of chloride-mediated inhibition in mouse spinal motoneuron. Journal of Physiology, 586, 1059–1075. Dubuc, R., Cabelguen, J. M., & Rossignol, S. (1985). Rhythmic antidromic discharges of single primary afferents recorded in cut dorsal root filaments during locomotion in the cat. Brain Research, 359, 375–378. Dubuc, R., Cabelguen, J.-M., & Rossignol, S. (1988). Rhythmic fluctuations of dorsal root potentials and antidromic discharges of primary afferents during fictive locomotion in the cat. Journal of Neurophysiology, 60, 2014–2036.
Duenas, S. H., Loeb, G. E., & Marks, W. B. (1990). Monosynaptic and dorsal root reflexes during locomotion in normal and thalamic cats. Journal of Neurophysiology, 63, 1467–1476. Dzhala, V. I., Talos, D. M., Sdrulla, D. A., Brumback, A. C., Mathews, G. C., Benke, T. A., et al. (2005). NKCC1 transporter facilitates seizures in the developing brain. Nature Medicine, 11, 1205–1213. Eccles, J. C., Schmidt, R. F., & Willis, W. D. (1963). Pharmacological studies on presynaptic inhibition. Journal of Physiology (London), 168, 500–530. Fellippa-Marques, S., Vinay, L., & Clarac, F. (2000). Spontaneous and locomotor-related GABAergic input onto primary afferents in the neonatal rat. European Journal of Neuroscience, 12, 155–164. Gao, X. B., Chen, G., & van den Pol, A. N. (1998). GABAdependent firing of glutamate-evoked action potentials at AMPA/kainate receptors in developing hypothalamic neurons. Journal of Neurophysiology, 79, 716–726. Gao, B. X., Stricker, C., & Ziskind-Conhaim, L. (2001). Transition from GABAergic to glycinergic synaptic transmission in newly formed spinal networks. Journal of Neurophysiology, 86, 492–502. Gao, B.-X., & Ziskind-Conhaim, L. (1995). Development of glycine- and GABA-gated currents in rat spinal motoneurons. Journal of Neurophysiology, 74, 113–121. Gonzalez-Islas, C. E., Chub, N. L., & Wenner, P. (2009). NKCC1 and AE3 appear to accumulate chloride in embryonic motoneurons. Journal of Neurophysiology, 101, 507–518. Gossard, J. P., Cabelguen, J.-M., & Rossignol, S. (1991). An intracellular study of muscle primary afferents during fictive locomotion in the cat. Journal of Neurophysiology, 65, 914–926. Gulledge, A. T., & Stuart, G. J. (2003). Excitatory actions of GABA in the cortex. Neuron, 37, 299–309. Hanson, M. G., & Landmesser, L. T. (2003). Characterization of the circuits that generate spontaneous episodes of activity in the early embryonic mouse spinal cord. Journal of Neuroscience, 23, 587–600. Hübner, C. A., Stein, V., Hermans-Borgmeyer, I., Meyer, T., Ballanyi, K., & Jentsch, T. J. (2001). Disruption of KCC2 reveals an essential role of K–Cl cotransport already in early synaptic inhibition. Neuron, 30, 515–524. Iizuka, M., Kiehn, O., & Kudo, N. (1997). Development in neonatal rats of the sensory resetting of the locomotor rhythm induced by NMDA and 5-HT. Experimental Brain Research, 114, 193–204. Ikeda, M., Toyoda, H., Yamada, J., Okabe, A., Sato, K., Hotta, Y., et al. (2003). Differential development of cation–chloride cotransporters and Cl homeostasis contributes to differential GABAergic actions between developing rat visual cortex and dorsal lateral geniculate nucleus. Brain Research, 984, 149–159.
13 Jean-Xavier, C., Mentis, G. Z., O'Donovan, M., Cattaert, D., & Vinay, L. (2007). Dual personality of GABA/glycinemediated depolarizations in the immature spinal cord. Proceedings of the National Academy of Sciences of the United States of America, 104, 11477–11482. Jean-Xavier, C., Pflieger, J.-F., Liabeuf, S., & Vinay, L. (2006). Inhibitory post-synaptic potentials in lumbar motoneurons remain depolarizing after neonatal spinal cord transection in the rat. Journal of Neurophysiology, 96, 2274–2281. Kerkut, G. A., & Bagust, J. (1995). The isolated mammalian spinal cord. Progress in Neurobiology, 46, 1–48. Kolta, A., Lund, J. P., Westberg, K. G., & Clavelou, P. (1995). Do muscle-spindle afferents act as interneurons during mastication? (letter; comment). Trends in Neurosciences, 18, 441. Kremer, E., & Lev-Tov, A. (1998). GABA-Receptor-independent dorsal root afferents depolarization in the neonatal rat spinal cord. Journal of Neurophysiology, 79, 2581–2592. Levy, R. A. (1975). The effect of intravenously administered gamma-aminobutyric acid on afferent fiber polarization. Brain Research, 92, 21–34. Marchetti, C., Tabak, J., Chub, N., O'Donovan, M. J., & Rinzel, J. (2005). Modeling spontaneous activity in the developing spinal cord using activity-dependent variations of intracellular chloride. Journal of Neuroscience, 25, 3601–3612. Mazzocchio, R., & Rossi, A. (1997). Involvement of spinal recurrent inhibition in spasticity. Further insight into the regulation of Renshaw cell activity. Brain, 120(Pt 6), 991–1003. Morita, H., Crone, C., Christenhuis, D., Petersen, N. T., & Nielsen, J. B. (2001). Modulation of presynaptic inhibition and disynaptic reciprocal Ia inhibition during voluntary movement in spasticity. Brain, 124, 826–837. Myers, C. P., Lewcock, J. W., Hanson, M. G., Gosgnach, S., Aimone, J. B., Gage, F. H., et al. (2005). Cholinergic input is required during embryonic development to mediate proper assembly of spinal locomotor circuits. Neuron, 46, 37–49. Nishimaru, H., Iizuka, M., Ozaki, S., & Kudo, N. (1996). Spontaneous motoneuronal activity mediated by glycine and GABA in the spinal cord of rat fetuses in vitro. Journal of Physiology, 497, 131–143. Nishimaru, H., & Kudo, N. (2000). Formation of the central pattern generator for locomotion in the rat and mouse. Brain Research Bulletin, 53, 661–669. Norreel, J.-C., Pflieger, J.-F., Pearlstein, E., Simeoni-Alias, J., Clarac, F., & Vinay, L. (2003). Reversible disorganization of the locomotor pattern after neonatal spinal cord transection in the rat. Journal of Neuroscience, 23, 1924–1932. Payne, J. A., Rivera, C., Voipio, J., & Kaila, K. (2003). Cation–chloride co-transporters in neuronal communication, development and trauma. Trends in Neurosciences, 26, 199–206.
Pilyavskii, A. I., Yakhnitsa, V. A., & Bulgakova, N. V. (1988). Antidromic dorsal root impulses during naturally occurring locomotion in rats. Neurophysiology, 20, 417–422. Plotkin, M. D., Snyder, E. Y., Hebert, S. C., & Delpire, E. (1997). Expression of the Na–K–2Cl cotransporter is developmentally regulated in postnatal rat brains: A possible mechanism underlying GABA's excitatory role in immature brain. Journal of Neurobiology, 33, 781–795. Ramirez, J. M., Telgkamp, P., Elsen, F. P., Quellmalz, U. J., & Richter, D. W. (1997). Respiratory rhythm generation in mammals: Synaptic and membrane properties. Respiratory Physiology, 110, 71–85. Ren, J., & Greer, J. J. (2003). Ontogeny of rhythmic motor patterns generated in the embryonic rat spinal cord. Journal of Neurophysiology, 89, 1187–1195. Ren, J., & Greer, J. J. (2006). Modulation of respiratory rhythmogenesis by chloride-mediated conductances during the perinatal period. Journal of Neuroscience, 26, 3721–3730. Rivera, C., Voipio, J., & Kaila, K. (2004). Two developmental switches in GABAergic signalling: The K–Cl cotransporter KCC2, and carbonic anhydrase CAVII. Journal of Physiology, 562, 27–36. Rivera, C., Voipio, J., Payne, J. A., Ruusuvuori, E., Lahtinen, H., Lamsa, K., et al. (1999). The Kþ/Cl co-transporter KCC2 renders GABA hyperpolarizing during neuronal maturation. Nature, 397, 251–255. Rossignol, S., Beloozerova, I. N., Gossard, J. P., & Dubuc, R. (1998). Presynaptic mechanisms during locomotion. In P. Rudomin, R. Romo & L. M. Mendell (Eds.), Presynaptic inhibition and neural control (pp. 385–397). Oxford: Oxford University Press. Rudomin, P. (1990). Presynaptic inhibition of muscle spindle and tendon organ afferents in the mammalian spinal cord. Trends in Neurosciences, 13, 499–505. Rudomin, P., Engberg, I., & Jimenez, I. (1981). Mechanisms involved in presynaptic depolarization of group I and rubrospinal fibers in cat spinal cord. Journal of Neurophysiology, 46, 532–548. Rudomin, P., Quevedo, J., & Eguibar, J. R. (1993). Presynaptic modulation of spinal reflexes. Current Opinion in Neurobiology, 3, 997–1004. Sadlaoud, K., Tazerart, S., Brocard, C., Jean-Xavier, C., Portalier, P., Brocard, F., et al. (2010). Differential plasticity of the GABAergic and glycinergic synaptic transmission to rat lumbar motoneurons after spinal cord injury. Journal of Neuroscience, 30, 3358–3369. Shao, X. M., & Feldman, J. L. (1997). Respiratory rhythm generation and synaptic inhibition of expiratory neurons in preBotzinger complex: Differential roles of glycinergic and GABAergic neural transmission. Journal of Neurophysiology, 77, 1853–1860.
14 Stein, V., Hermans-Borgmeyer, I., Jentsch, T. J., & Hubner, C. A. (2004). Expression of the KCl cotransporter KCC2 parallels neuronal maturation and the emergence of low intracellular chloride. Journal of Comparative Neurology, 468, 57–64. Stil, A., Liabeuf, S., Jean-Xavier, C., Brocard, C., Viemari, J. C., & Vinay, L. (2009). Developmental up-regulation of the potassium–chloride cotransporter type 2 in the rat lumbar spinal cord. Neuroscience, 164, 809–821. Sun, D., & Murali, S. G. (1999). Naþ–Kþ–2Cl cotransporter in immature cortical neurons: A role in intracellular Cl regulation. Journal of Neurophysiology, 81, 1939–1948. Sung, K. W., Kirby, M., McDonald, M. P., Lovinger, D. M., & Delpire, E. (2000). Abnormal GABAA receptor-mediated currents in dorsal root ganglion neurons isolated from Na–K–2Cl cotransporter null mice. Journal of Neuroscience, 20, 7531–7538. Takahashi, T. (1984). Inhibitory miniature synaptic potentials in rat motoneurons. Proceedings of Royal Society of London B Biological Sciences, 221, 103–109. Toennies, J. F. (1938). Reflex discharge from the spinal cord over the dorsal roots. Journal of Neurophysiology, 1, 378–390. Vardi, N., Zhang, L. L., Payne, J. A., & Sterling, P. (2000). Evidence that different cation chloride cotransporters in retinal neurons allow opposite responses to GABA. Journal of Neuroscience, 20, 7657–7663. Verdier, D., Lund, J. P., & Kolta, A. (2003). GABAergic control of action potential propagation along axonal branches
of mammalian sensory neurons. Journal of Neuroscience, 23, 2002–2007. Vinay, L., Brocard, F., Clarac, F., Norreel, J. C., Pearlstein, E., & Pflieger, J. F. (2002). Development of posture and locomotion: An interplay of endogenously generated activities and neurotrophic actions by descending pathways. Brain Research Review, 40, 118–129. Vinay, L., Brocard, F., Fellippa-Marques, S., & Clarac, F. (1999). Antidromic discharges of dorsal root afferents in the neonatal rat. Journal of Physiology Paris, 93, 359–367. Vinay, L., Brocard, F., Pflieger, J. F., Simeoni-Alias, J., & Clarac, F. (2000). Perinatal development of lumbar motoneurons and their inputs in the rat. Brain Research Bulletin, 53, 635–647. Vinay, L., & Clarac, F. (1999). Antidromic discharges of dorsal root afferents and inhibition of the lumbar monosynaptic reflex in the neonatal rat. Neuroscience, 90, 165–176. Vinay, L., & Jean-Xavier, C. (2008). Plasticity of spinal cord locomotor networks and contribution of cation–chloride cotransporters. Brain Research Review, 57, 103–110. Wu, W.-L., Ziskind-Conhaim, L., & Sweet, M. A. (1992). Early development of glycine- and GABA-mediated synapses in rat spinal cord. Journal of Neuroscience, 12, 3935–3945. Ziskind-Conhaim, L. (1998). Physiological functions of GABA-induced depolarizations in the developing rat spinal cord. Perspectives on Developemental Neurobiology, 5, 279–287.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 2
The spinal generation of phases and cycle duration Jean-Pierre Gossard{,*, Jennifer Sirois{, Patrick Noué{, Marie-Pascale Côté{, Ariane Ménard}, Hugues Leblond{ and Alain Frigonk {
Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montreal, Quebec, Canada { Department of Neurobiology and Anatomy, Drexel University College of Medicine, Philadelphia, USA } Department of Physiology, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA k Département de physiologie et biophysique, Université de Sherbrooke, Sherbrooke, Quebec, Canada
Abstract: During walking, an increase in speed is accompanied by a decrease in the stance phase duration while the swing phase remains relatively invariant. By definition, the rhythm generator in the lumbar spinal cord controls cycle period, phase durations, and phase transitions. Our first aim was to determine if this asymmetry in the control of locomotor cycles is an inherent property of the central pattern generator (CPG). We recorded episodes of fictive locomotion, that is, locomotor patterns in absence of reafference, in decerebrate cats with or without a complete spinal transection (acute or chronic). In fictive locomotion, stance and swing phases typically correspond to extension and flexion, respectively. In the vast majority of locomotor episodes, cycle period varied more with extensor phase duration. This could be observed without phasic sensory feedback or supraspinal structures or pharmacology. In a few experiments, we stimulated the mesencephalic locomotor region or selected peripheral nerves during fictive locomotion and both could alter the phase/cycle period relationship. We conclude that there is a built-in asymmetry within the spinal rhythm generator for locomotion, which can be modified by extraneous factors. Locomotor and scratching rhythms are characterized by alternation of flexion and extension phases within one hindlimb, which are mediated by rhythmgenerating circuitry within the spinal cord. Our second aim was to determine if rhythm generators for locomotion and scratch have similar control mechanisms in adult decerebrate cats. The regulation of cycle period during fictive scratching was evaluated, as were the effects of specific sensory inputs on phase durations and transitions during pinna-evoked fictive scratching. Results show that cycle period during fictive scratching varied predominantly with flexion phase duration, contrary to spontaneous fictive locomotion. Ankle dorsiflexion greatly increased extension phase duration and cycle period during fictive locomotion but did not alter cycle period during scratching. These data indicate that cycle period, phase durations, and phase transitions are not regulated similarly during fictive *Corresponding author. Tel.: þ1-514-343-5879 DOI: 10.1016/B978-0-444-53825-3.00007-3
15
16
locomotion and scratching, with or without sensory inputs, providing evidence for the existence of distinct interneuronal components of rhythm generation within the mammalian spinal cord. Keywords: CPG; locomotion; scratch; sensory control; step cycle; spinal cord.
Speed and stance Umnitsa the dog ran on a treadmill at different speeds for experiments led by Arshavsky and colleagues in 1965 for their study of the biomechanics of running (Arshavsky et al., 1965). When the speed increased from 3 to 8 km/h, they reported that “cycle period fell from 780 to 50 ms and this reduction was essentially due to shortening of the support phase from 520 to 240 ms, that is, 2.2 times. On the other hand, the duration of the transfer phase (i.e., swing) was almost constant for all values of speed and was approximately 260 ms.” Later, Goslow and colleagues (Goslow et al., 1973) studied the duration of subcomponents of the step cycle as a function of speed from in nine cats. They found that “an increase in forward speed from 1 to 16 miles/h involved only a 5% reduction in the duration of the swing phase but an 82% reduction in the stance phase.” From analysis of joint angles and electromyograms (EMG), Halbertsma studied how locomotor movements of intact cats changed with speed and stride duration (Halbertsma, 1983). He first showed that the support phase duration decreased much more than the swing duration with velocity of walking. He further showed that there was a linear relationship between cycle duration and both the swing and support phase durations but that the duration of support phases changed proportionally more than those of swing with increasing velocity. The duration of the phases and joint movements change with speed not only for walking but for running as well (Arshavsky et al., 1965; Goslow et al., 1973; Grillner, 1981; Grillner et al., 1979; Halbertsma, 1983; Musselman and Yang, 2007; Wetzel and Stuart, 1976; Yakovenko et al.,
2005). Approximately linear relationship between flexion and extension and step cycle duration occurs throughout the animal kingdom from cockroach (Pearson and Duysens, 1976) to man (Grillner et al., 1979) and the slope of the extension phase relative to step cycle duration is always much steeper than that of flexion. During stance, the contact with the ground is used to accelerate the pace by thrusting the body forward while during swing the limb travels through air while the contralateral limb is pushing on the ground. It, thus, makes sense that a change in speed is accomplished by changes in stance duration and related activities. So, where is the asymmetry in phase duration modulation coming from? It was suggested that the asymmetry in the control of cycle period is accomplished within the spinal network (Grillner and Dubuc, 1988), or that it is induced by descending supraspinal signals (Armstrong, 1986) and/or phasic sensory inputs from peripheral receptors (Juvin et al., 2007; Musselman and Yang, 2007; Yakovenko et al., 2005). All agree that supraspinal and sensory inputs can modify phase duration within the cycle and that, under normal circumstances, modification results from the concerted actions of all of these control levels. However, there is no easy way to dissect the contribution of each subsystem to the timing of phases in locomotion. Fictive locomotion can be used to investigate the centrally generated rhythm, as phasic sensory feedback from the moving limbs, or reafference, is absent. In this case, activity is recorded from peripheral nerves (electroneurography, ENG), or ventral roots, and a pattern of alternating extension and flexion phases, which roughly corresponds to stance and swing phases of walking, can be observed. Cycle
17
period, defined as the time between two successive extensor or flexor bursts of activity, is composed of a flexion and an extension phase, and can change by varying the duration of either phase, or both concurrently (“covariance”). A cycle period that changes as a function of the extension phase is termed an “extensordominated” locomotion, whereas a cycle period that changes as a function of the flexion phase is considered “flexor-dominated” (Fleshman et al., 1984; Yakovenko et al., 2005). Grillner and Zangger (1979) recorded ENG activity in spinal cats injected with l-dihydroxyphenylalanine (l-DOPA) and nialamide to test whether the spinal locomotor networks could generate the asymmetry reported in intact walking cats. They analyzed 14 sequences of fictive stepping from 10 deafferented and curarized cats and found that the duration of flexor discharges varied less with cycle duration than extensor discharges and concluded that “this asymmetry was thus not due to peripheral feedback, which may have seemed likely.” These authors reached the same conclusion from experiments comparing the locomotor patterns and cycle/phase durations in cats walking on a treadmill induced by mesencephalic stimulation before and after bilateral dorsal root transection (Grillner and Zangger, 1984). During fictive locomotion in paralyzed thalamic cats (Grillner and Dubuc, 1988), there was a significant linear relationship between changes in cycle duration variations in the duration of extensor bursts (r ¼ 0.99) but not to variations in flexor bursts duration (r ¼ 0.27). Because sensory feedback from the limbs is absent in these experiments, central networks alone are able to change cycle period by changing the duration of extension. However, recent studies suggested that extensor-dominated fictive locomotion is heavily dependent on phasic sensory feedback from the moving limbs as in intact walking. For example, phase/cycle period relationships were investigated during fictive locomotion induced by drugs or brainstem electrical stimulation in neonatal rats
(Juvin et al., 2007). It was shown that extension and flexion phases, approximated from ventral root burst durations, varied similarly with cycle period. However, when the hindlimbs were left attached to the cord, reafference induced an extensor-dominated pattern. In another study, Yakovenko et al. (2005) showed that fictive locomotion evoked by stimulating the mesencephalic locomotor region (MLR) was primarily flexordominated in decerebrate curarized cats. They proposed a symmetric half-centers oscillator model where variations in background drive could produce large variations in phase duration and determine which half-center was dominant. By varying the background drive and the individual gain of the oscillators, they could reproduce experimental data and concluded that the CPG oscillator behaves like reciprocally coupled integrators. The above studies concluded that the absence of phasic sensory feedback during fictive locomotion was responsible for the rarity of extensor-dominated locomotion. Another rhythmic movement that has been reported to be flexor-dominated is scratching in the cat. Most of the early work, however, was done in decerebrate cats during air-scratching or paralyzed animals with no movement (Berkinblit et al., 1978a,b; Deliagina et al., 1981; Deliagina et al., 1975), so in both cases, there was no contact with the paw. During air-scratching, the flexion period likely encounters more resistance as the limb flexes against gravity, compared with the extension period that would be assisted by gravity. In intact cats, scratching involves contact with the body which stimulates skin and stretch sensitive receptors in some muscles (i.e., increased load) during the extension period of the movement. During normal scratching, the period of extension is sometimes lengthened compared with air-scratching, resulting in an approximately equal period of flexion and extension (Kuhta and Smith, 1990). Rhythmic alternation between antagonistic muscles is also seen in respiration and mastication. A certain parallel in the control of phase
18
durations can be drawn between these rhythmic movements. Stance in stepping, or powerstroke, corresponds to the expiratory phase of breathing and to the closing phase in mastication while the swing phase of locomotion, or returnstroke, corresponds to inspiration during breathing and mandible opening during mastication. In the respiratory system, there is now convincing evidence that inspiration and expiration are generated by coupled, anatomically separate rhythm generators, one generating active expiration and the other generating inspiration (Del Negro et al., 2009; Feldman and Del Negro, 2006; Janczewski and Feldman, 2006). This work involved the contribution of Dr Jack Feldman, one of the three honorees of this volume. Unfortunately, there is no comparable evidence indicating whether there are separate oscillators for each phase of locomotion or mastication. Evidence from other species also indicates that different locomotor behaviors may be produced by shared and specialized networks (cf. Frigon, 2009). If there is only one common rhythm generator that can produce multiple rhythmic motor behaviors, there should be similar control mechanisms from one behavior to another. Alternatively, if different rhythmic motor behaviors have unique control mechanisms, it would provide evidence for different configurations of the rhythm-generating networks. Consequently, systematically investigating similarities and differences in the control of different rhythms could provide important clues regarding the functional organization and reorganization of CPGs. In a series of experiments summarized in the next section, we first investigated whether fictive locomotion is primarily flexor- or extensordominated without phasic sensory feedback by studying episodes of rhythm in the decerebrate cat and in complete spinal cats. We also evaluated the regulation of phases and cycle period during fictive scratching in the adult decerebrate cat. For this, we analyzed all episodes of fictive locomotion and scratch occurring without any nerve stimulation from different preparations of the
cat recorded over the last 15 years in my laboratory. We also compared how phases and cycle period are modified by specific sensory inputs during fictive locomotion and scratch. By applying stimuli at different times in the cycle, and by studying changes in phase duration, it is possible to gain some insight into the structure of the phases (Duysens, 1977). In particular, inputs from hip and ankle muscle afferents exert potent effect on the locomotor rhythm (Conway et al., 1987; Frigon and Gossard, 2010; Guertin et al., 1995; Perreault et al., 1995; Stecina et al., 2005). The ability of such inputs to reset and entrain the locomotor rhythm is generally taken as evidence that they are directly accessing the rhythm-generating circuitry (Conway et al., 1987; Gossard et al., 1994; Hultborn et al., 1998; Kriellaars et al., 1994; Andersson and Grilner, 1983; McCrea, 1998; Pearson et al., 1998; Schomburg et al., 1998). Also, as mentioned above, load and contact on the paw can prolong the extension phase during scratch. We, thus, studied the effects of ankle dorsiflexion, manually applied for several seconds, that stretches the ankle extensors (group IA and IB afferents) on the rhythm of fictive locomotion and scratch. A difference in the regulation of cycle period and in the ability to reset the cycle during fictive locomotion and scratching would provide evidence that both rhythms are produced by distinct rhythm generators.
Phase–cycle relationship Following a precollicular/premammillary decerebration, spontaneous fictive locomotion can be generated after anaesthesia is discontinued. We analyzed 35 episodes of spontaneous fictive locomotion in 28 cats (Frigon and Gossard, 2009). Three patterns emerged: extensor-dominated, covariance, and flexor-dominated. Across the group, extensor-dominated locomotion accounted for the majority of spontaneous fictive locomotion episodes (24/35 episodes), whereas covariance
19
(6/35), invariance (2/35), and flexor-dominated (3/35) were less frequent. It was shown that “speed” or pace of the fictive locomotor rhythm did not correlate with the above patterns. Also, the proportion, or percent time, occupied by extensor and flexor bursts relative to cycle period was usually involved in determining the dominance. However, it is important to note that only the percent time of flexion relative to cycle period correlated significantly with cycle period. In other words, in some cases, variations in the “longer” phase were not determining the variations in the cycle period. So, extensor dominance is not dependent on reafference but is it due to a particular descending drive resulting from the decerebration or is it due to spinal networks? In cats acutely spinalized, fictive locomotion can be induced by intravenous injection of nialamide and l-DOPA. We analyzed six such episodes in five cats and all episodes were extensor-dominated. The conclusion is exactly the same as reported in previous work on deafferented spinal cats (Grillner and Zangger, 1979) though we also tested whether there was covariance and used statistical analysis at every step. Together with the data of Grillner and Zangger (1979), there were 20 episodes of fictive locomotion from 15 acute spinal cats and in all cases the fictive locomotion was extensordominated. So, is it due to the fact that the spinal cord is acutely lesioned or to the injection of nialamide and L-DOPA? To determine if there is a particular dominance following a long-term spinal section, we studied cats completely spinalized 1 or 3 months before the acute experiment. Fictive locomotion (26 episodes in 16 cats) was induced by intravenous clonidine injection combined with perineal stimulation (Côté and Gossard, 2004; Côté et al, 2003). Extensordominated locomotion accounted for 24/26 episodes, one did not vary and the other covaried. Thus, fictive locomotion in cats after a complete spinal transection (acute or chronic) is primarily extensor-dominated. In such conditions, the
dominance is not determined by phasic sensory feedback or descending inputs but by the activity of isolated spinal cord networks. So, is extensor dominance a genuine property of the CPG networks devoid of other influences? One could argue that the choice of receptor agonists to induce rhythm in spinal cats could still be responsible for the bias in many of these episodes. However, in three exceptional cases (from three chronic spinal cats) fictive stepping could be evoked without drugs or manipulation. Following 5–6 trains of stimulation to the sciatic nerve, the temporary increase in spinal excitability evoked short episodes of well-organized fictive locomotion. In such conditions, the stepping rhythm occurred with no phasic sensory feedback, descending influence, pharmacology, or peripheral manipulation and these episodes were all extensor-dominated (Frigon and Gossard, 2009). So, we may conclude that, in the adult spinal cord, the rhythm generator is producing primarily an extensor-dominated stepping rhythm. Is dominance task-dependent? Or are different rhythmic behaviors generated by a single common CPG with the same dominance? Scratching is another rhythmic movement of the hindlimb in the cat involving alternating activities between flexors and extensors. To exclude the contribution of reafference, we again studied the regulation of cycle period during fictive scratch (28 episodes in 15 cats) (Frigon and Gossard, 2010). In some cases, we succeeded in recording episodes of spontaneous fictive locomotion and scratch in the same cats. During fictive scratch, the ENG profiles show a brisk extension phase and a longer flexion phase. First, our measurements revealed that the duration of bursts in extensors did not equal the time between two successive flexor bursts as in locomotion, so we could not use the extensor burst duration as a measure of extension phase as in locomotion. Consequently, we measured the “silent” interval between two flexor bursts that we called “flexor-interval.” Three patterns emerged from our analysis: flexor-dominated
20 (a)
3000
Extensor burst duration (ms)
r = 0.90 2500 2000 1500 1000 r = 0.16
500 0 –500 0
(b)
500
1000 1500 2000 2500 Extensor cycle period (ms)
3000
3500
3000 r = 0.86 2500
Flexor burst duration (ms)
episodes that accounted for the majority of fictive scratching episodes (20/28 episodes), flexor interval-dominated (7/28), and invariance (1/28). There was no significant difference in cycle period and phase durations between flexordominated and nonflexor-dominated episodes. As expected, the flexion–interval phase (but not the extension phase) occupied a significantly greater percentage of cycle periods in flexor interval-dominated episodes but not the extension phase. Thus, although there is a clear flexordominated bias in the regulation of cycle period during fictive scratch, this bias can change if the percentage of the flexion–interval phase relative to cycle period increases. In Fig. 1, the two plots summarize all the data from fictive locomotion and scratch. The relationships between extension (Fig. 1a), flexion (Fig. 1b), and cycle period are illustrated for each episode for both fictive locomotion (black dots, n ¼ 35 episodes) and scratch (grey dots, n ¼ 28 episodes). In Fig. 1a, there is a significant linear regression between the extensor burst duration and the cycle duration variations for locomotion (r ¼ 0.90) but not for scratch (r ¼ 0.16). On the contrary, in Fig. 2b, there is no significant relationship between the flexor burst duration and cycle period changes during locomotion (r ¼ 0.21), but there is one for scratch (r ¼ 0.86). These plots represent pooled averaged values (as in Juvin et al., 2007) and corresponds to what was found for the majority of episodes analyzed individually (Frigon and Gossard, 2009, 2010). We could not find similar studies for respiration and mastication. However, from the literature about fictive mastication we chose the two longest episodes we could find (Donga and Lund, 1991; Tsuboi et al., 2003) both involving Dr Jim Lund, one of three honorees of this volume. We then estimated the phase–cycle relationship by measuring ENG bursts duration. In both cases, there was a linear relationship between cycle period and the duration of the closing phase (the powerstroke) but not with the opening phase whether it occupied a smaller or larger proportion
2000 1500 1000 r = 0.21 500 0 0
500
1000 1500 2000 2500 Flexor cycle period (ms) Locomotion
3000
3500
Scratch
Fig. 1. Phase–cycle relationship during fictive locomotion and scratch. Plots of burst duration of extension (a) and flexion (b) versus cycle period for each episode of both fictive locomotion (black dots, n ¼ 35 episodes) and scratch (grey dots, n ¼ 28 episodes). In (a), there is a significant linear regression between the extensor burst duration and the cycle duration variations for locomotion (r ¼ 0.90) but not for scratch (r ¼ 0.16). In (b), there is no significant relationship between the flexor burst duration and cycle duration changes during locomotion (r ¼ 0.21) but there is one for scratch (r ¼ 0.86).
of the cycle (unpublished results). A simple analysis of already existing data from laboratories studying mastication could confirm whether cycle period duration is always dependent on variations of the powerstroke.
21
MLR TA
SmAB
1s 900
1600 1400
re = 0.92, P < 0.001 rf = 0.06, P < 0.23 Pef < 0.001
700 600 500 400 300 200 200
Burst duration (ms)
Burst duration (ms)
800
re = 0.18, P = 0.025 rf = 0.99, P < 0.001 Pef < 0.001
1200 1000 800 600 400 200 0
400
600
800
1000
1200
200 400 600 800 1000 1200 1400 1600 1800
1400
Step cycle duration (ms)
Step cycle duration (ms) TA burst
SmAB burst
Fig. 2. From spontaneous to MLR-induced fictive locomotion. The top two traces show alternating ENG bursts of an ankle flexor (Tibialis Anterior, TA) and of a hip extensor (Semi-membranosus-Anterior Biceps, SmAB) during two episodes of fictive locomotion. At the beginning, the rhythm is spontaneous and, when MLR is stimulated (as indicated), the rhythm stops with irregular TA activity. After a few seconds, a new robust and rapid rhythm is evoked. Under these episodes are plots of the burst duration of extension (grey diamonds) and flexion (black circles) versus the step cycle duration calculated from the entire episodes of spontaneous rhythm (on the left) and MLR-induced rhythm (on the right). Also shown for each plot are the coefficient of linear regression (re or rf) and level of significance (P) and whether there is a significant difference between the two slopes (Pef).
Modulation by descending or sensory inputs In a book about the neural control of rhythmic movements published in 1988, Serge Rossignol, another distinguished honoree of this volume, together with Drs Lund and Drew, wrote a remarkable review comparing the sensory control of locomotion, respiration, and mastication
(Rossignol et al., 1988). They proposed that the frequency and structure of the rhythmic pattern may be changed by the additive interaction of descending and sensory inputs acting on a common generating network, the CPG. In the three movements, primary afferents display either a tonic or phasic firing pattern that can be used to control the duration and amplitude of phase and
22
EMG bursts. Also common to all is the existence of sensory feedback that changes only the amplitude of particular EMG bursts and their phase-dependency modulation while other sensory inputs can also alter cycle duration. During rhythmic movements, sensory feedback is important in shaping the transitions between phases and two particularly powerful signals are hip joint position and load on ankle extensors (Duysens and Pearson, 1980; Duysens et al., 2000; Grillner and Rossignol, 1978; Harkema et al., 1997; Rossignol et al., 2006). We were, thus, curious to compare the effects of loading ankle extensors on the rhythm of locomotion and scratch. A slight and maintained dorsiflexion of the left ankle, which stretches ankle extensors, was applied during episodes of fictive locomotion (13 episodes in 4 cats) and scratch (11 episodes in 4 cats) (Frigon and Gossard, 2010). Ankle dorsiflexion had huge effects on the extension phase during fictive stepping but very modest effects during scratch. During spontaneous fictive locomotion, ankle dorsiflexion significantly increased cycle period by increasing the duration of the extension phase, while the duration of the flexion phase decreased. It was also possible to switch a locomotor episode that was primarily flexor-dominated into extensor-dominated one. With dorsiflexion, 100% of episodes of fictive locomotion were extensor-dominated. During fictive scratch, ankle dorsiflexion had no effect on cycle period, because the modest reduction in the flexion phase was compensated by an increase in the extension and/or flexor-interval phases. During dorsiflexion, 64% scratch episodes remained flexor-dominated. Thus, during fictive locomotion, ankle dorsiflexion increases cycle period and strengthens extensor-dominance by increasing the duration of the extension phase. During fictive scratch, however, dorsiflexion does not change cycle period but it weakens flexordominance by increasing the duration of the extension and flexion–interval phases. The effects of dorsiflexion are most probably due to sensory inputs from group I afferents of
the stretched ankle extensors. The action of such afferents and the underlying pathways have been the subject of many reports and reviews in locomotion (Conway et al., 1987; Dietz and Duysens, 2000; Duysens and Pearson, 1980; Gossard et al., 1994; Guertin et al., 1995; Hiebert and Pearson, 1999; Pearson, 2007; Pearson et al., 1992; Rossignol et al., 2006; Sinkjaer et al., 2000; Stein et al., 2000; Whelan et al., 1995). Activation of these receptors signal load on the limbs and through mono-, di-, and polysynaptic pathways, evokes excitation of hindlimb extensors to increase weight support and propulsion during stance. The effects of dorsiflexion on fictive locomotion can easily be explained by such pathways. In another project (Frigon and Gossard, 2010), group I afferents from ankle extensors (Plantaris or Lateral Gastrocnemius-Soleus) were stimulated with short trains (25 pulses 1.8T) at different moments of the cycle to compare their ability to reset the rhythm during fictive locomotion and scratch. As expected and described before (Conway et al., 1987; Gossard et al., 1994; Guertin et al. 1995), the stimuli could stop the on-going flexion phase, trigger a new extension phase and advance all of the following cycles, that is, a reset of the stepping rhythm. During the extension phase, the stimuli would prolong the phase and thus delay all following cycles. Remarkably, such resets occurred when the stimuli were given at any moment within the step cycle. This result indicates that such inputs can reboot the clock function of the locomotor generator at any time. The same nerve stimulation during fictive scratch, given anytime in extension, was able to prolong that phase, even though the magnitude of the increase was modest as compared to locomotion and produced modest changes in cycle period. The same stimulation could decrease flexion phase duration when given in mid- or late-flexion but, remarkably, could not trigger a new extensor phase and reset the rhythm. This result indicates that such inputs have a restricted access to the clock function of the scratch generator.
23
Supraspinal structures are obviously able to control rhythm generation in many ways to comply with specific demands of voluntary movements. The cortex, reticular formation, red nucleus, the lateral vestibular nucleus among others have been shown to adjust the rhythm of locomotion (Armstrong and Drew, 1985; Drew and Rossignol, 1984; Orlovsky, 1972; Perreault et al., 1994; Rho et al., 1999; Russel and Zajac, 1979; Shik et al., 1966). Inputs from some of these descending tracts converge with sensory inputs from group I afferents of extensors on common spinal pathways to boost extensor burst duration (Leblond et al., 2000, 2001). The relative contribution of each supraspinal structure in determining and adjusting the duration of phases and of the cycle of rhythmic movements is largely unknown. A relevant structure to investigate is the MLR because of its importance in triggering stepping (Jordan et al., 2008; Shik et al., 1966). We were curious to see what effects adding electrical stimulation of the MLR would have on episodes of spontaneous fictive locomotion. Figure 2 shows an episode of spontaneous fictive locomotion followed by an episode of MLRevoked locomotion (see also Frigon and Gossard, 2009). The spontaneous fictive locomotor episode (mean cycle period ¼ 0.71 s) was extensordominated. Once MLR stimulation began, the spontaneous rhythm stopped and was followed a few seconds later by a faster and more vigorous rhythm (mean cycle period ¼ 0.68 s) that was flexor-dominated. In seven trials in two cats, MLR stimulation did not appear to modify the ongoing spontaneous rhythm but replaced it with another rhythm. During the spontaneous locomotor episode shown in Fig. 2, extension and flexion, respectively occupied 68.0% (extensor burst / extensor cycle) and 38.1% (flexor burst / flexor cycle) of cycle period, whereas during MLRevoked locomotion this changed to 35.9% and 84.1%. Thus, under identical experimental conditions, the control of cycle period can be switched from extensor- to flexor-dominance, and the percent time occupied by extension and
flexion relative to cycle period can also change from a greater extension proportion during spontaneous fictive locomotion to a greater flexion percentage during MLR-evoked locomotion. The bias toward either flexor- or extensor-dominance seen with MLR stimulation could be due to the level of decerebration and/or to changes in excitability in spinal networks and/or reticular formation. Indeed, in the majority of cases using MLR stimulation, the decerebration consists of a precollicular, postmammillary lesion. This type of decerebration leads to silent background activity in hindlimb nerves and MLR stimulation is required to induce fictive locomotor rhythm, which is primarily flexor-dominated. However, the decerebration used in our study involves a precollicular, premammillary section, sparing the subthalamic locomotor region (SLR), which is important for the initiation of goal-directed locomotion (Orlovsky, 1969; Whelan, 1996). The resulting reticulospinal commands, coupled with the intrinsic excitability of spinal locomotor networks, are able to activate spontaneously the rhythm generators and fictive locomotion that is primarily extensor-dominated (Frigon and Gossard, 2009). Another possible explanation and our current hypothesis is that the MLR recruits reticulospinal cells with a particular emphasis on spinal networks to flexors. Indeed, stimuli applied within the reticular formation usually favor activities in flexors at rest or during locomotion (Drew, 1991; Drew and Rossignol, 1984; Orlovsky, 1972; Perreault et al., 1994; Russel and Zajac, 1979). However, it is possible to activate reticulospinal axons in the medial longitudinal fasciculus (MLF) that can reset the locomotor rhythm by stopping the ongoing flexion and triggering a new extension phase, just as group I afferents from ankle extensors (Leblond et al., 2000). It is, therefore, possible that different loci in and around the MLR and/or excitability changes within the reticular formation could result in either dominance. Unfortunately the identity and location of reticular cells leading to one dominance or the other are unknown.
24
It is also possible that the use of particular chemicals imitating descending neuromodulators could induce specific phase dominance in rhythm generation. In neonatal rats, fictive rhythms induced by different pharmacological agents can differ from one another (Cowley and Schmidt, 1997; Kiehn and Kjaerulff, 1996). For instance, the timing and magnitude of flexor and extensor bursts, differed in fictive rhythms induced by dopamine or serotonin (Kiehn and Kjaerulff, 1996). In Juvin et al. (2007), the fictive locomotion in neonatal rats induced by a mix of N-methyl-D,L-aspartate, serotonin, and dopamine showed covariance. On the other hand, the same result was observed while stimulating the brainstem attached to the isolated spinal cord. Therefore, it is tempting to suggest that immaturity in neonatal rats or human infants (Musselman and Yang, 2007) is responsible for the absence of either dominance. Our current hypothesis is that mammals are born with symmetrically coupled oscillators (covariance) but that sensory feedback, elicited by a lifetime of walking, induces plastic changes within the rhythm generator to produce an inherent bias for extensor dominance in the adult (Frigon and Gossard, 2009). It is believed that, under normal circumstances, the control of phase generation should result from integrating descending and sensory inputs. For example, we already described a shift from flexor- to extensor-dominance of fictive locomotor rhythm by adding dorsiflexion (Frigon and Gossard, 2009). We could also shift an extensordominance during fictive stepping in spinal cat (Nialamide and L-DOPA) to flexor-dominance by stimulating a cutaneous nerve (superficial peroneus, SP) with a single pulse twice per second (unpublished result). In 1984, Burke and colleagues reported that stimulating a cutaneous nerve (sural) with an increasing intensity during an MLR-induced rhythm shifted the balance from complete flexor dominance toward predominant extensor activation (Fleshman et al., 1984). Also, Jordan and colleagues reported that when the MLR-induced rhythm was entrained by imposed
sinusoidal hip movements, the match of hip cycle duration and step cycle duration was accomplished by a variation in extensor burst duration (Kriellaars et al., 1994). Pearson and Rossignol (1991) showed that simply changing the posture of the limb, from flexion to extension, in chronic spinal cats progressively increased flexor burst duration and decreased cycle period during fictive locomotion. From these observations, it is obvious that the integration of sensory cues and descending drive can produce variations in either phase to enable the desired change in step cycle period.
Functional organization of rhythm generator In 1985, Lennard (1985; Lennard and Hermanson, 1985) proposed a separation of the central pattern generator into two distinct units: the central timing network (CTN) and the central intracycle pattern generator (CIPG). His work on the sensory modulation of the monodopal swimming cycle in the turtle showed that proprioceptive (mostly spindle afferents) inputs could produce a permanent phase shift of the locomotor rhythm with earlier or later onset than expected of all poststimulus swimming cycles (a “reset” of the locomotor rhythm) while cutaneous inputs only produced a temporary phase shift. It was proposed that muscle afferents have access to the part of the CPG which organizes the cycle-by-cycle repetition (CTN) while cutaneous information modulates the coupling between CTN and CIPG. This idea of having a rhythm generator separated form a pattern generator was picked up into many models of different rhythmic movements. For a complete discussion of CPG models, one should refer to an excellent review from McCrea and Rybak (McCrea and Rybak, 2008). With respect to phase and cycle generation, it is conceptually useful to consider the “central timing unit” or the “clock function” of the CPG as a simple set of coupled oscillators, one for
25
flexion and one for extension, with reciprocal connections (Kiehn, 2006; Koshland and Smith, 1989; Orsal et al., 1990; Stein and Smith, 1997). Dominance of a particular phase in determining cycle period is explained by an increased excitability in either oscillator (Yakovenko et al., 2005). The commands from the rhythm generator are then transmitted to the pattern generator to coordinate and distribute inputs to specific motor pools. On top of this core generator are the modulatory systems mediated by sensory and descending pathways. We have seen that the control of phase duration can arise from all of these systems. Our data indicate that there is an inherent bias within the core rhythm generator for locomotion to change the duration of the extension phase when varying the cycle period as seen in intact stepping. We have seen that reafference is not necessary to observe this bias in thalamic cats (Grillner and Dubuc, 1988), decerebrate cats (Frigon and Gossard, 2009; Yakovenko et al., 2005), acute spinal cats (Frigon and Gossard, 2009; Grillner and Zangger, 1979), and chronic spinal cats (Frigon and Gossard, 2009). Supraspinal signals are also not required because extensor dominance is almost always observed during fictive locomotion in the spinal cat with acute or chronic complete lesion which also does not depend on the choice of neuromodulators to induce stepping. Fictive locomotion was used in an attempt to differentiate the contribution of spinal networks and of modulatory systems in determining the regulation of cycle period by changes in phase duration. We used the same approach to extract the contribution of central networks during fictive scratch. We found that, contrary to locomotion, that the scratch cycle was modified mostly by changing the duration of flexion. A flexordominated pattern is also found during scratch in intact cats (Kuhta and Smith, 1990), which indicates that the regulation of cycle period does not depend on phasic sensory feedback from the moving limbs as in locomotion. Moreover, sensory inputs from load- and/or stretch-sensitive
receptors from ankle extensors exerted markedly different effects on cycle period, phase durations, and phase transitions during fictive locomotion and scratching. The distinct phase control in locomotion and scratch was a surprise because the dogma was that locomotor and scratching rhythms are produced by the same CPG (Baev, 1978; Berkinblit et al., 1978a,b; Deliagina et al., 1975, 1981; Gelfand et al., 1988; Perreault et al., 1999). The arguments for and against a single rhythm generator are discussed in Frigon and Gossard (2010). From our perspective, the most salient difference between the two behaviors is their specific sensory control. As described above, our results showed that the access to the rhythm generator from group I afferents of ankle extensors is quite different in the two tasks. Others groups have found that motoneuronal responses evoked by stimulating cutaneous nerves were prominent during fictive locomotion and showed strong phasic modulation, whereas during fictive scratching the responses were suppressed with little phase modulation (Degtyarenko et al., 1998). On the other hand, motoneuronal responses evoked by muscle nerve stimulation were strongly modulated during fictive locomotion and scratching but the modulation pattern was often reversed. Stimulating the MLR also produced phase-dependent responses in motoneurons during fictive locomotion (Shefchyk and Jordan, 1985) but these responses were abolished during fictive scratch (Degtyarenko et al., 1998). Thus, the processing of inputs from the periphery and from supraspinal structures is strikingly different during fictive locomotion and scratch, which is also paralleled by differences in presynaptic inhibition (Côté and Gossard, 2003). The differences in accessing the rhythm generator during locomotion and scratch are most easily explained by different sensory processing at the interneuronal level. Our current hypothesis is that specific configurations of the rhythm-generating interneuronal circuitry are required to produce unique features of locomotion and scratching.
26
How different must interneuronal configurations have to be for CPGs to be considered distinct? How much shared interneurons are enough to consider them as one rhythm generator? This can be debated endlessly because there is not enough identified CPG neurons in mammals. Indeed, the CPG for locomotion in the mammalian spinal cord is still pretty much a black box. Fortunately, research using transgenic mice with genetically identified cells (Chapters in previous volume of this book) has opened a new era that promises to soon provide data about the location, projections, and function of identified spinal interneurons. We will then have a rigorous base to construct realistic models of rhythm generator that will further our understanding of how phases and cycle duration are controlled and how different rhythms can be generated within the lumbar spinal cord. References Andersson, O., & Grillner, S. (1983). Peripheral control of the cat's step cycle II. Entrainment of the central pattern generators for locomotion by sinusoidal hip movements during “fictive locomotion” Acta Physiologica Scandinavica, 118, 229–239. Armstrong, D. M. (1986). The supraspinal control of mammalian locomotion. Journal of Physiology, 405, 1–37. Armstrong, D. M., & Drew, T. (1985). Forelimb elctromyographic response to motor cortex stimulation during locomotion in the cat. Journal of Physiology, 367, 327–351. Arshavsky, Y. I., Kots, Y. M., Orlovsky, G. N., Rodionov, I. M., & Shik, M. L. (1965). Investigation of the biomechanics of running by the dog. Biofizika, 10, 665–672. Baev, K. V. (1978). Central locomotor program for the cat's hindlimb. Neuroscience, 3, 1081–1092. Berkinblit, M. B., Deliagina, T. G., Feldman, A. G., Gelfand, I. M., & Orlovsky, G. N. (1978). Generation of scratching. II. Nonregular regimes of generation. Journal of Neurophysiology, 41, 1058–1069. Berkinblit, M. B., Deliagina, T. G., Feldman, A. G., & Orlovsky, G. N. (1978). Generation of scratching. I. Activity of spinal interneurons during scratching. Journal of Neurophysiology, 41, 1040–1057. Conway, B. A., Hultborn, H., & Kiehn, O. (1987). Proprioceptive input resets central locomotor rhythm in the spinal cat. Experimental Brain Research, 8, 643–656.
Côté, M.-P., & Gossard, J.-P. (2003). Task-dependent presynaptic inhibition. Journal of Neuroscience, 23, 1886–1893. Côté, M.-P., & Gossard, J.-P. (2004). Step-training dependent plasticity in spinal cutaneous pathways. Journal of Neuroscience, 24, 11317–11327. Côté, M.-P., Ménard, A., & Gossard, J.-P. (2003). Spinal cats on the treadmill: Changes in load pathways. Journal of Neuroscience, 23, 2789–2796. Cowley, K. C., & Schmidt, B. J. (1997). Regional distribution of the locomotor pattern-generating network in the neonatal rat spinal cord. Journal of Neurophysiology, 77, 247–259. Degtyarenko, A. M., Simon, E. S., Norden-Krichmar, T., & Burke, R. E. (1998). Modulation of oligosynaptic cutaneous and muscle afferent reflex pathways during fictive locomotion and scratching in the cat. Journal of Neurophysiology, 79, 447–463. Del Negro, C. A., Kam, K., Hayes, J. A., & Feldman, J. L. (2009). Asymetric control of inspiratory and expiratory phases by excitability in the respiratory network of neonatal mice in vitro. Journal of Physiology, 587, 1217–1231. Deliagina, T. G., Feldman, A. G., Gelfand, I. M., & Orlovsky, G. N. (1975). On the role of central program and afferent inflow in the control of scratching movements in the cat. Brain Research, 100, 297–313. Deliagina, T. G., Orlovsky, G. N., & Perret, C. (1981). Efferent activity during fictitious scratch reflex in the cat. Journal of Neurophysiology, 45, 595–604. Dietz, V., & Duysens, J. (2000). Significance of load receptor input during locomotion: A review. Gait and Posture, 1, 102–110. Donga, R., & Lund, J. P. (1991). Discharge patterns of trigeminal commissural last-order interneurons during fictive mastication in the rabbit. Journal of Neurophysiology, 66, 1564–1578. Drew, T. (1991). Functional organization within the medullary reticular formation of the intact unanesthetized cat. III. Microstimulation during locomotion. Journal of Neurophysiology, 66, 919–938. Drew, T., & Rossignol, S. (1984). Phase-dependent responses evoked in limb muscles by stimulation of medullary reticular formation during locomotion in thalamic cats. Journal of Neurophysiology, 52, 653–675. Duysens, J. (1977). Reflex control of locomotion as revealed by stimulation of cutaneous afferents in spontaneously walking premammillarv cats. Journal of Neurophysiology, 40, 737–751. Duysens, J., Clarac, F., & Cruse, H. (2000). Load-regulating mechanisms in gait and posture: Comparative aspects. Physiological Review, 80, 83–133. Duysens, J., & Pearson, K. G. (1980). Inhibition of flexor burst generation by loading ankle extensor muscles in walking cats. Brain Research, 187, 321–332. Feldman, J. L., & Del Negro, C. A. (2006). Looking for inspiration: New perspective on respiratory rhythm. Nature Reviews Neuroscience, 7, 232–242.
27 Fleshman, J. W., Lev-Tov, A., & Burke, R. E. (1984). Peripheral and central control of flexor digitorium longus and flexor hallucis longus motoneurons: The synaptic basis of functional diversity. Experimental Brain Research, 54, 133–149. Frigon, A. (2009). Reconfiguration of the spinal interneuronal network during locomotion in vertebrates. Journal of Neurophysiology, 101, 2201–2203. Frigon, A., & Gossard, J.-P. (2009). Asymmetric control of cycle period by the spinal locomotor rhythm generator in the adult cat. Journal of Physiology, 587, 4617–4628. Frigon, A., & Gossard, J.-P. (2010). Evidence for specialized rhythm-generating mechanisms in the adult mammalian spinal cord. Journal of Neuroscience, 30, 7061–7071. Frigon, A., Sirois, J., & Gossard, J.-P. (2010). Effects of ankle and hip muscle afferent inputs on rhythm generation during fictive locomotion. Journal of Neurophysiology, 103, 1591–1605. Gelfand, I. M., Orlovsky, G. N., & Shik, M. L. (1988). Locomotion and scratching in tetrapods. In A. H. Cohen, S. Rossignol & S. Grillner (Eds.), Neural control of rhythmic movements in vertebrates (pp. 167–199). New York: Wiley. Goslow, G. E., Reinking, R. M., & Stuart, D. G. (1973). The cat step cycle: Hind limb joint angles and muscle lengths during unrestrained locomotion. Journal of Morphology, 141, 1–42. Gossard, J.-P., Brownstone, R. M., Barajon, I., & Hultborn, H. (1994). Transmission in a locomotor-related group Ib pathway from hindlimb extensor muscles in the cat. Experimental Brain Research, 98, 213–228. Grillner, S. (1981). Control of locomotion in bipeds, tetrapods, and fish. In J. M. Brookhart & V. B. Mountcastle (Eds.), Handbook of physiology. The nervous system II (pp. 1179–1236). Bethesda, MD: American Physiological Society. Grillner, S., & Dubuc, R. (1988). Control of locomotion in vertebrates: Spinal and supraspinal mechanisms. In S. G. Waxman (Ed.), Functional recovery in neurological disease, Vol. 453 (pp. 425–453). New York: Raven Press. Grillner, S., Halbertsma, J., Nillsson, J., & Thorstensson, A. (1979). The adaptation to speed in human locomotion. Brain Research, 165, 177–182. Grillner, S., & Rossignol, S. (1978). On the initiation of the swing phase of locomotion in chronic spinal cats. Brain Research, 146, 269–277. Grillner, S., & Zangger, P. (1979). On the central generation of locomotion in the low spinal cat. Experimental Brain Research, 34, 241–261. Grillner, S., & Zangger, P. (1984). The effect of dorsal root transection on the efferent motor pattern in the cat's hindlimb during locomotion. Acta Physiologica Scandinavica, 120, 393–405. Guertin, P., Angel, M. J., Perreault, M.-C., & McCrea, D. A. (1995). Ankle extensor group I afferents excite extensors
throughout the hindlimb during fictive locomotion in the cat. Journal of Physiology, 487, 197–209. Halbertsma, J. M. (1983). The stride cycle of the cat: The modelling of locomotion by computerized analysis of automatic recordings. Acta Physiologica Scandinavica Supplement, 521, 1–75. Harkema, S. J., Hurley, S. L., Patel, U. K., Requejo, P. S., Dobkin, B., & Edgerton, V. R. (1997). Human lumbosacral spinal cord interprets loading during stepping. Journal of Neurophysiology, 77, 797–811. Hiebert, G. W., & Pearson, K. G. (1999). Contribution of sensory feedback to the generation of extensor activity during walking in the decerebrate cat. Journal of Neurophysiology, 81, 758–770. Hultborn, H., Conway, B., Gossard, J.-P., Brownstone, R., Fedirchuk, B., & Schomburg, E. D. (1998). How do we approach the locomotor network in the mammalian spinal cord? Annals of New York Academy of Sciences, 860, 70–82. Janczewski, W. A., & Feldman, J. L. (2006). Distinct rhythm generators for inspiration and expiration in the juvenile rat. Journal of Physiology, 570, 407–420. Jordan, L. M., Liu, J., Hedlund, P. B., Akay, T., & Pearson, K. G. (2008). Descending command systems for the initiation of locomotion in mammals. Brain Research Review, 57, 183–191. Juvin, L., Simmers, J., & Morin, D. (2007). Locomotor rhythmogenesis in the isolated rat spinal cord: A phasecoupled set of symmetrical flexion extension oscillators. Journal of Physiology, 583, 115–128. Kiehn, O. (2006). Locomotor circuits in the mammalian spinal cord. Annual Review of Neuroscience, 29, 279–306. Kiehn, O., & Kjaerulff, O. (1996). Spatiotemporal characteristics of 5-HT and dopamineinduced rhythmic hindlimb activity in the in vitro neonatal rat. Journal of Neurophysiology, 75, 1472–1482. Koshland, G. F., & Smith, J. L. (1989). Mutable and immutable features of paw-shake responses after hindlimb deafferentation in the cat. Journal of Neurophysiology, 62, 162–173. Kriellaars, D. J., Brownstone, R. M., Noga, B. R., & Jordan, L. M. (1994). Mechanical entrainment of fictive locomotion in the decerebrate cat. Journal of Neurophysiology, 71, 2074–2086. Kuhta, P. C., & Smith, J. L. (1990). Scratch responses in normal cats: Hindlimb kinematics and muscle synergies. Journal of Neurophysiology, 64, 1653–1667. Leblond, H., Ménard, A., & Gossard, J.-P. (2000). Bulbospinal control of spinal cord pathways generating locomotor extensor activities in the cat. Journal of Physiology, 525, 225–240. Leblond, H., Ménard, A., & Gossard, J.-P. (2001). Corticospinal control of locomotor pathways generating extensor activities in the cat. Experimental Brain Research, 138, 173–184.
28 Lennard, P. R. (1985). Afferent perturbations during “monopodal” swimming movements in the turtle: Phasedependent cutaneous modulation and proprioceptive resetting of thelocomotor rhythm. Journal of Neuroscience, 5, 1434–1445. Lennard, P. R., & Hermanson, J. W. (1985). Central reflex modulation during locomotion. Trends in Neurosciences, 8, 483–486. McCrea, D. A. (1998). Neuronal basis of afferent-evoked enhancement of locomotor activity. Annals of New York Academy of Sciences, 860, 216–225. McCrea, D. A., & Rybak, I. A. (2008). Organization of mammalian locomotor rhythm and pattern generation. Brain Research Review, 57, 134–146. Musselman, K. E., & Yang, J. F. (2007). Loading the limb during rhythmic leg movements lengthens the duration of both flexion and extension in human infants. Journal of Neurophysiology, 97, 1247–1257. Orlovsky, G. N. (1969). Spontaneous and induced locomotion of the thalamic cat. Biophysics, 14, 1154–1162. Orlovsky, G. N. (1972). The effect of different descending systems on flexor and extensor activity during locomotion. Brain Research, 40, 359–371. Orsal, D., Cabelguen, J.-M., & Perret, C. (1990). Interlimb coordination during fictive locomotion in the thalamic cat. Experimental Brain Research, 82, 536–546. Pearson, K. G. (2007). Role of sensory feedback in the control of stance duration in walking cats. Brain Research Review, 57, 222–227. Pearson, K. G., & Duysens, J. (1976). Function of segmental reflexes in the control of stepping in coackroaches and cats. In R. M. Hermanet al. (Ed.) Neural control of locomotion (pp. 519–537). New York: Plenum Press. Pearson, K. G., Misiaszek, J. E., & Fouad, K. (1998). Enhancement and resetting of locomotor activity by muscle afferents. Annals of New York Academy of Sciences, 860, 203–215. Pearson, K. G., Ramirez, J. M., & Jiang, W. (1992). Entrainment of the locomotor rhythm by group Ib afferents from ankle extensor muscles in spinal cats. Experimental Brain Research, 90, 557–566. Pearson, K. G., & Rossignol, S. (1991). Fictive motor patterns in chronic spinal cats. Journal of Neurophysiology, 66, 1874–1887. Perreault, M.-C., Angel, M. J., Guertin, P., & McCrea, D. A. (1995). Effects of stimulation of hindlimb flexor group II afferents during fictive locomotion in the cat. Journal of Physiology, 487, 211–220. Perreault, M.-C., Enriquez-Denton, M., & Hultborn, H. (1999). Proprioceptive control of extensor activity during fictive scratching and weight support compared to fictive locomotion. Journal of Neuroscience, 19, 10966–10976. Perreault, M.-C., Rossignol, S., & Drew, T. (1994). Microstimulation of the medullary reticular formation
during fictive locomotion. Journal of Neurophysiology, 71, 229–245. Rho, M.-J., Lavoie, S., & Drew, T. (1999). Effects of red nucleus microstimulation on the locomotor pattern and timing in the intact cat: A comparison with the motor cortex. Journal of Neurophysiology, 81, 2297–2315. Rossignol, S., Dubuc, R., & Gossard, J.-P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Review, 86, 89–154. Rossignol, S., Lund, T., & Drew, T. (1988). The role of sensory inputs in regulating patterns of rhyhtmical movements in higher vertebrates: A comparison between locomotion, respiration and mastication. In A. H. Cohen, S. Rossignol & S. Grillner (Eds.), Neural control of rhythmic movements in vertebrates (pp. 201–283). New York: Wiley. Russel, D. F., & Zajac, F. E. (1979). Effects of stimulating Deiter's nucleus and medial longitudinal fasciculus on the timing of the fictive locomotor rhythm induced in cats by DOPA. Brain Research, 177, 588–592. Schomburg, E. D., Petersen, N., Barajon, I., & Hultborn, H. (1998). Flexor reflex afferents reset the step cycle during fictive locomotion in the cat. Experimental Brain Research, 122, 339–350. Shefchyk, S. J., & Jordan, L. M. (1985). Excitatory and inhibitory post-synaptic potentials in alpha-motoneurons produced during fictive locomotion by stimulation of the mesencephalic locomotor region. Journal of Neurophysiology, 53, 1345–1355. Shik, M. L., Severin, F. V., & Orlovsky, G. N. (1966). Control of walking and running by means of electrical stimulation of the mid-brain. Biophysics, 11, 756–765. Sinkjaer, T., Andersen, J. B., Ladouceur, M., Christensen, L. O. D., & Nielsen, J. (2000). Major role for sensory feedback in soleus EMG activity in the stance phase of walking in man. Journal of Physiology, 523, 817–827. Stecina, K., Quevedo, J., & McCrea, D. A. (2005). Parallel reflex pathways from flexor muscle afferents evoking resetting and flexion enhancement during fictive locomotion and scratch in the cat. Journal of Physiology, 569, 275–290. Stein, R. B., Misiaszek, J. E., & Pearson, K. G. (2000). Functional role of muscle reflexes for force generation in the decerebrate walking cat. Journal of Physiology, 525, 781–791. Stein, P. S. G., & Smith, J. L. (1997). Neural and biochemical control strategies for different forms of vertebrate hindlimb locomotor tasks. In P. S. G. Stein, S. Grillner, A. I. Selverston & D. G. Stuart (Eds.), Neurons, networks and motor behaviour (pp. 61–73). Cambridge, MA: MIT Press. Tsuboi, A., Kolta, A., Chen, C. C., & Lund, J. P. (2003). Neurons of the trigeminal main sensory nucleus participate in the generation of rhythmic motor patterns. European Journal of Neuroscience, 17, 229–238.
29 Wetzel, M. C., & Stuart, D. G. (1976). Ensemble characteristics of cat locomotion and its neural control. Progress in Neurobiology, 7, 1–98. Whelan, P. J. (1996). Control of locomotion in the decerebrate cat. Progress in Neurobiology, 49, 481–515. Whelan, P. J., Hiebert, G. W., & Pearson, K. G. (1995). Stimulation of the group I extensor afferents prolongs the
stance phase in walking cats. Experimental Brain Research, 103, 20–30. Yakovenko, S., McCrea, D. A., Stecina, K., & Prochazka, A. (2005). Control of locomotor cycle durations. Journal of Neurophysiology, 94, 1057–1065.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 3
Networks within networks: The neuronal control of breathing Alfredo J. Garcia III1, Sebastien Zanella1, Henner Koch1, Atsushi Doi and Jan-Marino Ramirez* Center for Integrative Brain Research, Seattle Children's Research Institute, Seattle, Washington, USA Department of Neurological Surgery, University of Washington, Seattle, Washington, USA
Abstract: Breathing emerges through complex network interactions involving neurons distributed throughout the nervous system. The respiratory rhythm generating network is composed of micro networks functioning within larger networks to generate distinct rhythms and patterns that characterize breathing. The pre-Bötzinger complex, a rhythm generating network located within the ventrolateral medulla assumes a core function without which respiratory rhythm generation and breathing cease altogether. It contains subnetworks with distinct synaptic and intrinsic membrane properties that give rise to different types of respiratory rhythmic activities including eupneic, sigh, and gasping activities. While critical aspects of these rhythmic activities are preserved when isolated in in vitro preparations, the pre-Bötzinger complex functions in the behaving animal as part of a larger network that receives important inputs from areas such as the pons and parafacial nucleus. The respiratory network is also an integrator of modulatory and sensory inputs that imbue the network with the important ability to adapt to changes in the behavioral, metabolic, and developmental conditions of the organism. This review summarizes our current understanding of these interactions and relates the emerging concepts to insights gained in other rhythm generating networks. Keywords: Breathing; Respiratory rhythm generation; Pre-Botzinger complex and interactions. Not surprisingly, a large degree of plasticity characterizes all levels of neuronal integration from the molecular, cellular to the network and ultimately behavioral level. Common principles of behavioral plasticity have been described in numerous invertebrates (Harris et al., 2010; Marder and Goaillard, 2006; Nadim et al., 2008; Nusbaum, 2002; Ramirez and Pearson, 1993) and
Introduction Behaviors are continuously adapted to changes in an organism's internal and external environment.
*Corresponding author, 1equal contribution. Tel.: 206-884-8188;Fax:206-884-1210. E-mail:
[email protected] DOI: 10.1016/B978-0-444-53825-3.00008-5
31
32
mammalian model systems including humans (Lee et al., 2009; Macfarlane and Mitchell, 2009; Millhorn et al., 1980; Peng et al., 2003). In this review, we will focus on modulatory processes that are critical for the neuronal control of mammalian breathing, but we will also describe how insights gained in the respiratory system relate to other networks and behaviors. Breathing is well integrated with many other behaviors and needs to adapt continuously to changes in the metabolic and behavioral demands of an organism. Both the respiratory frequency and amplitude of breathing are adapted to behavioral conditions like posture, physical activity, sleep, or speech. In fact, breathing is so sensitive to an organism's internal state that the characteristics of breathing can reveal whether someone is calm, agitated, or scared. As in most animal behaviors, plasticity in the respiratory system seems to depend on a variety of amines, steroids, and peptides that exert their modulatory actions by acting on a large number of ion channels, receptors, and second messenger systems (Doi and Ramirez, 2008). Neuromodulatory processes play equally important roles during well-oxygenated and hypoxic conditions, and when disturbed neuromodulation has been associated with a number of pathophysiological conditions ranging from Rett syndrome (Viemari et al., 2005a) to Sudden Infant Death Syndrome (SIDS) (Paterson et al., 2009). The mammalian respiratory network and the nuclei controlling neuromodulation are widely distributed along the neural axis. Important for breathing are the neuronal networks located in the ventral respiratory column (VRC) (Alheid et al., 2002; Feldman and Del Negro, 2006; Feldman et al., 2003; McCrimmon et al., 2004). These networks (Fig. 1) include, from rostral to caudal, the retrotrapezoid nucleus / parafacial respiratory group complex (RTN/pFRG), the Bötzinger complex, the pre-Bötzinger complex (pre-BötC), the rostral ventral respiratory group (rVRG), and the caudal VRG (cVRG) (Alheid et al., 2002; Feldman and Del Negro, 2006; Feldman et al., 2003; McCrimmon et al., 2004;
Onimaru and Homma, 2003; Onimaru et al., 2006; Smith et al., 1991; Thoby-Brisson et al., 2009). The distinction between these networks is based on a histological characterization (Alheid et al., 2002; Gray et al., 1999; Guyenet et al., 2002), and their distinct functional properties (Gray et al., 1999, 2001; Janczewski and Feldman, 2006; Janczewski et al., 2002; McCrimmon et al., 2004; Onimaru and Homma, 2003; Onimaru et al., 2006; ThobyBrisson et al., 2005, 2009). An important role in respiratory rhythm generation and modulation has also been described for the Kölliker-Fuse nucleus and the parabrachial complex; both are located in the dorsal pons (Alheid et al., 2004; Dick et al., 1994; Kobayashi et al., 2005; Milsom et al., 2004). The role of these neurons in the modulation and generation of eupnea (Chamberlin and Saper, 1994; Dutschmann and Herbert, 1996; St-John and Paton, 2004; Von Euler and Trippenbach, 1976) and in the transition phase between inspiration and expiration (Cohen, 1971; Morschel and Dutschmann, 2009; Von Euler and Trippenbach, 1976) has received considerable attention. Other areas involved in breathing include the cerebellum (Harper, 2000a; Harper et al., 2000b), the neocortex (Davenport et al., 2010; Von Leupoldt et al., 2010), as well as the periaqueductal gray, which is particularly important for the integration of speech and breathing (Subramanian and Holstege, 2010). An area that is both essential and sufficient for generating the respiratory rhythm is the pre-BötC, a network located within the ventrolateral medulla (Smith et al., 1991). Disruption of rhythmic activity in the pre-BötC causes irreversible loss or major disruption of breathing in vivo (Gray et al., 2001; McKay et al., 2005; Ramirez et al., 1998c; Tan et al., 2008). Isolated in a brainstem slice preparation, the pre-BötC continues to generate respiratory activity (Fig. 2) (Ramirez et al., 1996; Smith et al., 1991). In these slices, respiratory rhythmic activity can be recorded directly from the surface of pre-BötC (Lieske et al., 2000) or from the hypoglossal motor nucleus, which receives respiratory rhythmic input via an interneuronal population located outside the
33
Network Interactions Cortex Cerebellum PAG Kolliker-Fuse cVRG rVRG RTN/pFRG
Raphe magnus Raphe obscurus Raphe pallidus
51A HT , 2C 2A, ,4, 7
NeuroModulation
Reconfiguration
[Ca++]
Homeostatic Plasticity
N
NE 2 a1 a b
EupneicSighGly GABA ITASK GluR ICa-P NMDA INaP mGluR8 INaP ICAN GluR Gaspingrhythm GluR INap I KATP
R HI OS F1 a
RT
pre-BotC
SPNk1
NF BD rkB T
A6 A1, A2 A5
Ca ro tid Ra N ph TS Bod e y
Longterm Plasticity
TNFa PGE2
Motor nuclei XII Phrenic Sig
h Eupnea
g
pin
Gas
Behavior
Environment Fig. 1. Networks within networks. A schematic illustrating some of the interactions that are involved in the neuronal control of breathing. As complex as this schematic may appear, it still represents only a fraction of the known interactions. The interactions depicted in this figure are clustered according to their function in the neuronal control of breathing. Network interactions that are important for the generation of the respiratory rhythm include the interactions between the cortex, cerebellum, PAG (Periaqueductal Gray), Kölliker-Fuse, cVRG (caudal Ventral Respiratory Group), rVRG (rostral Ventral Respiratory Group), RTN (retrotrapezoid nucleus)/pFRG (parafacial respiratory group), and the pre-Bötzinger complex (pre-BötC). While the preBötC is capable of generating three distinct rhythmic activities via network reconfiguration (eupneic-, sigh-, and gasping-rhythm) even in isolation (see also Fig. 2), in the intact animal other networks will also contribute to the reconfiguration and the shaping of the respiratory rhythms that are transmitted to the motor nuclei. The extent and significance of these contributions to the overall motor output and ultimately the behavior is a topic of intense investigations. Respiratory rhythm generation is also the target of three types of modulatory processes: Neuromodulation characterizes the modulatory processes occurring on a momentto-moment basis and it is mediated via numerous aminergic and peptidergic substances acting on various receptor subtypes. For simplicity, the schematic illustrates only three neuromodulators: NE (norepinephrine) acting on a1, a2, and b noradrenergic receptors; 5-HT (serotonin) acting on 5-HT1A, 5-HT2A, 5-HT2C, 5-HT4, and 5-HT7 receptors; and SP (substance P) acting on the NK1 receptor. These neuromodulators are released by nuclei that include the Raphe magnus, obscurus, and pallidus and the noradrenergic regions: A1, A2, A5, and A6. For further explanations see text. Long-term plasticity characterizes the modulatory
34
hypoglossal nucleus (Koizumi et al., 2008).This in vitro approach has greatly facilitated our understanding of the cellular mechanisms of respiratory rhythm generation and neuromodulation. Although we still lack a deep understanding of how the different areas of the respiratory network interact and how they are altered by neuromodulation, the pre-BötC itself has provided an important avenue to study not only the fundamental elements of respiratory rhythm generation, but also how a single neuronal network can generate multiple rhythmic activity patterns as previously demonstrated in invertebrate neuronal networks (Harris-Warrick and Johnson, 2010; Marder and Goaillard, 2006; Meyrand et al., 1991; Ramirez, 1998a). This review will focus on our current understanding of respiratory rhythm generation in the pre-BötC and its alteration by reconfiguration, neuromodulation, and plasticity (Fig. 1). We will also consider the possible roles of these mechanisms in physiological and pathophysiological conditions. Network reconfiguration The notion that the respiratory network can reconfigure to generate different forms of breathing such as eupnea, sighs, and gasps was introduced approximately a decade ago (Lieske et al., 2000). Since that time, much has been learned not only about the mechanisms of reconfiguration within the pre-BötC (Pena et al., 2004a), but also about the role of the pons and other areas in reconfiguring the network from the eupneic into the gasping state (Paton et al., 2006).
Under well-oxygenated conditions, the preBötC generates two distinct rhythms: a faster small amplitude rhythm (“fictive eupnea”) and a much slower large amplitude rhythm (Fig. 2; “fictive sighs”). Several studies have shown that these activities originate from a multifunctional network located within the pre-BötC that can be partly preserved in the in vitro transverse medullary slice (Lieske and Ramirez, 2006a; Lieske et al., 2000; Ruangkittisakul et al., 2008; Tryba et al., 2008). Although the majority of neurons are activated during both eupneic and sigh rhythmic activities, pharmacological manipulations suggest that both activities emerge through distinct mechanisms. Fictive sigh, but not eupneic activity, is critically dependent on synaptic mechanisms involving the P/Q type calcium channels (Cav2.1). Interestingly, only a relatively small subpopulation of respiratory neurons receive glutamatergic inputs that depend on P/Q type calcium currents (Lieske and Ramirez, 2006a), suggesting that the pool of respiratory neurons contains a subset of neurons with specialized synapses that are critical for sigh rhythm generation. These synapses depend also on the activation of mGluR8 receptors (Lieske and Ramirez, 2006b), while NMDA-dependent mechanisms play an important role in the generation of eupneic activity (Lieske and Ramirez, 2006b; McCrimmon et al., 1997). Thus, in the respiratory network an overlapping pool of neurons generates different rhythmic activities using different behavior-specific types of synaptic mechanisms. This is a common principle that has also been described in the spinal cord (Berkowitz et al., 2010), the neocortex (Kramer et al., 2008;
processes that lead to long-term changes in respiratory activity, which includes, for example, long-term facilitation of the respiratory frequency and amplitude which is evoked by intermittent hypoxia (see text for details). Long-term plasticity is mediated among other molecules by BDNF acting on the TrkB receptor and reactive oxygen species (ROS) as well as HIF1a. Critical areas involved in long-term plasticity are the carotid body and the pre-BötC as well as a variety of motor nuclei. Important chemosensory areas include the NTS (nucleus tractus solitarius), the RTN, and the Raphe nuclei besides the carotid body. These chemosensitive areas are important sensors for inputs from the environment, which is in part affected by the breathing behavior itself. Homeostatic plasticity characterizes regulatory processes that are critical for stabilizing network activity in the context of respiratory rhythm generation as well as neuromodulation and long-term plasticity. Unfortunately, little is known about the homeostatic mechanisms that are specifically relevant for the neuronal control of breathing, but much is already known in other networks in particular in the networks located in the neocortex and hippocampus.
35 (a) In vivo whole animal a Control condition
b Initial phase of hypoxia Sigh
c Late phase of hypoxia
d Sigh
Gasping
Gasping
Eupnea
Augmentation
Depression
Eupnea
1s
(b) In vitro medulla slice a Control condition Fictive eupnea
b Initial phase of hypoxia Fictive sigh
Augmentation
c Late phase of hypoxia Fictive gasping
Depression
4s
d Fictive gasping
Fictive sigh
Fictive eupnea
Fig. 2. Isolating the pre-Bötzinger complex in a single medullary brainstem slice preserves multiple rhythms that reflect breathing rhythms found inÐ vivo and illustrates reconfiguration of a multifunctional network. (A) Integrated traces from an in vivo electromyogram ( EMG) recorded from respiratory muscles show multiple rhythmic behaviors for breathing. (a) In control conditions, eupnea, “the normal breathing rhythm” conditions, is characterized by augmented bell-shaped waveform. (b) During the transition to hypoxia (initial phase of hypoxia), the frequency of eupnea and sigh rhythms become faster (i.e., augmentation). The sigh rhythm consists of large amplitude complex waveforms. While slower in frequency, sigh rhythms are generated together with the eupnea rhythm (for detail see text). (c) During hypoxia, the frequency of the breathing rhythm is slow (late phase of hypoxia) (i.e., depression) and the waveforms can be clearly distinguished from the eupneic bursts based on their fast rise time. Collectively, these waveforms, during the depression, are described as the gasping rhythm. (d) Overlay of representative eupneic (black line), sigh (red line), and gasping (blue line) waveforms found in the respiratory rhythm in vivo. (B) Integrated activity Ð ( pre-BötC) of extracellular recordings obtained from the surface of the pre-Bötzinger complex within the medullary brainstem slice illustrates that this neuronal network generates several patterns of neuronal activity that are reminiscent to breathing behaviors found in vivo. (a) In control conditions, fictive eupnea is generated. The waveforms in fictive eupnea are similar to that of eupnea, having an augmented bell-shaped waveform. (b) Large amplitude, complex bursts, described as resembling sighs in both frequency and waveform. Similar to that in vivo, the fictive sigh rhythm is commonly increased during the augmentation phase during the transition to hypoxia (initial phase of hypoxia). (c) During hypoxia, the frequency of the fictive respiratory rhythm slows (late phase of hypoxia, i.e., depression) and the waveforms of the population rhythm changes from the augmented bell-shaped waveform of fictive eupnea to a fictive gasping waveform. Similar to the gasping rhythm in vivo, fictive gasping possess waveforms that have fast rise times. (d) Overlay of representative fictive eupneic (black line), fictive sighs (red line), and fictive gasping (blue line) waveforms illustrates the ability of pre-Bötzinger complex to generate multiple rhythms that likely contribute to in vivo breathing rhythms. Moreover, these waveform patterns represent the summation of different, yet overlapping mechanisms involving network reconfiguration and neuromodulation of the pre-Bötzinger complex (see text for full discussion).
Wulff et al., 2009), and various invertebrate species (Haque et al., 2006; Jing and Gillette, 2003; Wood et al., 2000). As the respiratory network responds to hypoxia, the breathing frequency in vivo transitions into an augmentation followed by depression
(Fig. 2A; England et al., 1995; Haddad and Mellins, 1984; Neubauer et al., 1990), a sequence that is also seen in the isolated pre-BötC (Telgkamp and Ramirez, 1999). During the depression phase, the inspiratory burst changes from an augmenting, bell-shaped burst to a
36
decrementing burst which is one of the characteristic features of gasping (Fig. 2A and B; Lieske et al., 2000; Pena et al., 2004a). The transition into the gasping activity pattern can be gradual, both in vitro and in vivo. During the transition, bursts with varying rise time and burst duration are typical, a characteristic that was referred to as pregasping in vivo (Wang et al., 1993). Hypoxia causes also a characteristic decrease in synaptic inhibition, which has been observed under both in vivo and in vitro conditions (England et al., 1995; Ramirez et al., 1998b; Richter et al., 1991; Thoby-Brisson and Ramirez, 2000; Völker et al., 1995). The depression of synaptic inhibition contributes to the reconfiguration of the respiratory network by altering the discharge pattern of a variety of neurons. Late-inspiratory neurons discharge earlier during inspiration (England et al., 1995). Many expiratory and inspiratory neurons in the ventrolateral medulla become inactive before cessation of phrenic and/or hypoglossal (XII) activity (Ballanyi et al., 1994; England et al., 1995; Ramirez, 1998a; Richter et al., 1991; Telgkamp and Ramirez, 1999; Thoby-Brisson and Ramirez, 2000). While XII neurons exhibit a massive potentiation of the rhythmic bursts both in vivo and in vitro (Ramirez et al., 1997; Telgkamp and Ramirez, 1999), respiratory neurons in the ventrolateral medulla respond inconsistently and become either weakly de- or hyperpolarized (Ramirez et al., 1998b; Richter et al., 1991; Thoby-Brisson and Ramirez, 2000). Within the pre-BötC, neurons can be differentiated into nonpacemaker and pacemaker neurons dependent on their ability to intrinsically generate bursting activity. Since nonpacemaker and pacemaker neurons are differentially affected by hypoxia and neuromodulators, they will be considered in more detail in this review. In isolation from fast synaptic transmission, nonpacemaker neurons either enter a tonic firing state or become quiescent while pacemaker neurons retain spontaneous bursting properties (Pena et al., 2004a). Pacemaker neurons can be further identified into cadmium sensitive (CS) and cadmium insensitive
(CI) pacemaker neurons. Bursting in CS pacemakers seems to depend on a nonspecific cation current (ICAN) while burst properties of CI pacemaker appear to be mediated via the persistent sodium current (INaP) (Pena et al., 2004a). Inhibition of these currents in the respective pacemaker subtypes eliminates their ability to spontaneously burst in synaptic isolation (Chevalier et al., 2008; Del Negro et al., 2002a, 2005; Pena et al., 2004a; Tryba and Ramirez, 2004). However, it is important to emphasize that neither ICAN nor INaP are exclusive currents mediating pacemaker properties. These inward currents play also critical roles in amplifying synaptic inputs not only in pacemaker neurons, but also nonpacemaker neurons (Del Negro et al., 2002b, 2005; Ramirez et al., 1996; Rubin et al., 2009). Indeed, the role of bursting properties in amplifying synaptic inputs has been described in many neuronal systems, such as the locust flight system in which sensory synaptic inputs are amplified in a nonlinear manner by intrinsic bursting mechanisms (Ramirez and Pearson, 1991, 1993). As will be discussed later in this review, bursting properties can be induced and suppressed by neuromodulators, which imbues neuronal networks with the ability to amplify specific synaptic pathways in a state and behavioral dependent manner. In the locust flight system, bursting properties are induced at the onset of flight to nonlinearly amplify sensory inputs (Ramirez and Pearson, 1993). During locomotion intrinsic membrane properties boost synaptic input also in the spinal cord (Brownstone et al., 1992). These inward currents play important roles also in a variety of other mammalian systems, such as in networks involved in mastication (Kolta et al., 2007). In the neocortex, these currents depolarize neurons toward their firing threshold (Pennartz et al., 1997) and boost synaptic input (Schwindt and Crill, 1999; Stuart and Sakmann, 1995). Pharmacological blockade of INaP has been shown to modulate locomotion generated in the spinal cord (Darbon et al., 2004; Tazerart et al., 2007) and it can suppress slow oscillations generated in cortical networks (van Drongelen et al., 2006).
37
It is impossible to negate the fact that a substantial portion of neurons that are activated during respiration possess bursting properties. Moreover, the majority of neurons possess INaP or ICAN or both, and the balance between these two major inward currents and a set of outward currents determines whether a neuron is a pacemaker or nonpacemaker (Koizumi et al., 2010). This conclusion has important implications for rhythm generation in general (Hudson and Prinz, 2010), which makes it difficult to differentiate between pacemakers and nonpacemakers as the balance between inward and outward currents is gradual. Nonpacemakers without any bursting properties and pacemakers with strong bursting properties are the extremes of a gradient of neurons possessing different degrees of bursting properties. Consequently, it is probably impossible to unambiguously dissect the relative contribution of pacemaker versus nonpacemakers. The ratio between pacemakers and nonpacemakers is not fixed, but dependent on the metabolic and modulatory state of the network, since neuromodulators, such as NE, SP, or 5-HT, can induce bursting in nonpacemakers (Pena and Ramirez, 2002, 2004b; Viemari and Ramirez, 2006). In an attempt to determine the relative contribution of these inward currents to respiratory rhythm generation, various laboratories have employed substances that are known to block these currents. Blocking either INaP with Riluzole or the ICAN by Flufenamic acid alone does not block fictive eupnea (Pena et al., 2004a), but both substances together lead to the cessation of respiratory rhythmic activity indicating that these conductances are critical for respiratory rhythm generation. A recent study extended these observations by demonstrating that applying Substance P or low doses of AMPA in the presence of Riluzole and Flufenamic acid could restimulate rhythmogenesis (Del Negro et al., 2005). Hence, the relative importance of pacemaker neurons in well-oxygenated conditions is the matter of an ongoing debate. It is, however, clear that the relative contribution of these currents changes considerably during
the transition from eupneic to gasping activity (Fig. 1). While eupneic activity involves the activation of neurons possessing INaP and ICANdependent bursting mechanisms, pacemaker neurons that depend on ICAN selectively hyperpolarize during hypoxia rendering the network more dependent on INaP during gasping, a finding that has been confirmed in vitro, in situ, and in vivo (Paton et al., 2006; Pena and Aguileta, 2007; Pena et al., 2004a, 2008). Pacemaker neurons may also differentially contribute to the generation of eupneic versus sigh activity, because the generation of sighs is more sensitive to the manipulation of the INaP current (Tryba et al., 2008). Thus, lessons learned from the respiratory network indicate that the different network states are characterized by differential contribution of different types of bursting mechanisms.
Neuromodulation and rhythm generation In the respiratory network, the same neuromodulator differentially acts on a variety of receptor subtypes, thereby exerting specific and sometimes even diverging effects on different parameters of the network output. Perhaps, best understood in the neuronal control of breathing is the role of catecholaminergic neurons that are found in the brainstem and project onto neurons of the respiratory network (Doi and Ramirez, 2008; Hilaire, 2006; Hokfelt et al., 1984; VanderHorst and Ulfhake, 2006; Viemari and Hilaire, 2002; Viemari and Ramirez, 2006). These neurons are clustered in noradrenergic (Fig. 1; A1, A2, A5, A6, and A7) and adrenergic (C1 to C3) nuclei. In the pons, the activity of A6 (i.e., the locus coeruleus) neurons is modulated by hypoxia both in vivo (Guyenet et al., 1993) and in vitro (Nieber et al., 1995; Yang et al., 1997). Although some A6 neurons are excited while others inhibited during hypoxia, A6 exerts an overall stimulatory effect on breathing. Electrical stimulation of the locus coeruleus increases
38
breathing frequency (Doi and Ramirez, 2010), whereas genetic alteration of A6 decreases respiratory activity (Viemari et al., 2004; Viemari et al., 2005a). Unlike the A6 nucleus, all neurons in A5 appear to be activated by hypoxia in vivo (Guyenet et al., 1993) and inhibit breathing. Thus, lesions of A5 increase the respiratory frequency in vitro (Viemari and Hilaire, 2002) and reduce the posthypoxic frequency depression (Coles and Dick, 1996). Thus, while both A5 and A6 contain noradrenergic neurons that modulate respiratory network activity, their effects are very diverse even though they are modulated under similar conditions. In mutants where A5 or A6 neurons are altered (Viemari et al., 2004, 2005a), the hypoxic response is blunted supporting a role of both groups in oxygen sensing. A5 and A6 are not the only noradrenergic nuclei modulating respiration: In the medulla, endogenous release of NE from A1/C1 neurons stimulates the respiratory rhythm in vitro (Zanella et al., 2006). Exogenously applied NE on medullary slices stimulates the respiratory rhythm by a direct effect on inspiratory neurons (Viemari and Ramirez, 2006). Because hypoxia increases TH expression (Peyronnet et al., 2003; Roux et al., 2000, 2003) and Fos-like immunoreactivity (Erickson and Millhorn, 1994; Teppema et al., 1997), it is likely that neurons in the A1/C1 and A2/C2 region are activated by low oxygen levels. This modulatory role may be disturbed in certain neurological disorders. In Mecp2 mutant mice, a model for Rett syndrome, the number of tyrosine hydroxylase positive neurons in A1/C1 and A2/C2 and the level of norepinephrine in the medulla is decreased (Viemari et al., 2005b), yet these mice show an increased ventilatory response to hypoxia (Roux et al., 2008; Voituron et al., 2009). Mechanistically, NE stimulates inspiratory nonpacemaker and pacemaker neurons contained within the pre-BötC acting presumably via a1, a2, and b-noradrenergic mechanisms (Viemari and Ramirez, 2006). NE induces ICAN-dependent bursting properties in active nonpacemaker neurons, and it depolarizes CI pacemakers and
increases their burst frequency. In CS pacemakers, NE increases only the amplitude of the depolarizing drive potential and the number of action potentials during the burst. However, in contrast to the situation in CI pacemakers NE does not affect the burst frequency in CS pacemakers. This differential effect is preserved at the network level, since only the modulation of the burst amplitude but not the frequency depends on the activation of ICAN (Viemari and Ramirez, 2006). This leads to the important conclusion that different network parameters are differentially modulated by the same neuromodulator acting on different cellular targets. NE is not the only bioamine acting on the respiratory network. Like other catecholaminergic neurons, serotonergic neurons are also found in the brainstem and project to neurons involved in breathing (Dahlstrom and Fuxe, 1964; Fuxe, 1965; Holtman et al., 1990; VanderHorst and Ulfhake, 2006). Serotonergic neurons are contained in nuclei numbered from B1 to B9, from the caudal to the rostral axis (Dahlstrom and Fuxe, 1964). The nuclei can also be referred to as raphe pallidus (B1), raphe obscurus (B2), and raphe magnus (B3). The action of these groups on breathing is diverse and often data are contradictory probably due to species differences, state of the animals (awake or sleep), or type and level of anesthesia (Besnard et al., 2009; Doi and Ramirez, 2010; Holtman et al., 1986; Lalley, 1986; Sessle et al., 1981). For instance, in rats under volatile anesthesia (Besnard et al., 2009) electric stimulation of raphe magnus and obscurus induced apnea, whereas stimulation of the raphe pallidus induced tachypnea. On the other hand, in mice anesthetized with urethane (i.p.) electric stimulation of the raphe magnus increases respiratory frequency (Doi and Ramirez, 2010). In cats, stimulation of raphe pallidus and obscurus (Holtman et al., 1986; Lalley, 1986) increases phrenic discharge amplitude and frequency, whereas they decrease during stimulation of the raphe magnus (Lalley, 1986; Sessle et al., 1981). Similarly, exogenous application of serotonergic agents in vitro has various actions on respiratory activity (Di Pasquale et al., 1992, 1994; Hilaire
39
et al., 1997; Morin et al., 1990; Schwarzacher et al., 2002). In brainstem slices containing the pre-BötC, exogenous application of 5HT2A agonist or blockade of serotonergic reuptake increases inspiratory frequency (Pena and Ramirez, 2002). Consistent with these results, blockade of the activation of 5HT2A receptors by endogenous release of serotonin decreases respiratory frequency. This has been recently confirmed in rats using transverse brainstem slices and in situ preparations and extended to 5HT2C and 5HT4 receptors (Ptak et al., 2009). Indeed, neurons from the raphe obscurus show a tonic activity and emit projections to the pre-BötC, which innervates the raphe reciprocally. Neuromodulation cannot be discussed without also emphasizing the importance of peptidergic modulation. Perhaps, the most studied peptide in the respiratory network is substance P (Gray et al., 2001; Hayes and Del Negro, 2007; Morgado-Valle and Feldman, 2004; Pena and Ramirez, 2004b). Neurons releasing substance P are localized in the nucleus of solitary tract (NTS), the nucleus ambiguous (NA), the raphe, the dorsal motor nucleus of the vagus (X), and the hypoglossal nucleus (Ribeiro-da-Silva and Hokfelt, 2000). Substance P is often coreleased with other neurotransmitter. For example, neurons in the raphe contain 5-HT and substance P (Kachidian et al., 1991; Ptak et al., 2009) and they project directly on the pre-BötC. NK-1 receptors are strongly expressed on pre-BötC neurons as well as serotonergic neurons of the raphe (Alheid and McCrimmon, 2008; Gray et al., 2001; Stornetta et al., 2003; Wang et al., 2001). Substance P activates the inspiratory frequency at the network and behavioral level (Del Negro et al., 2005; Doi and Ramirez, 2010; Gray et al., 2001; Pena and Ramirez, 2004b; Thoby-Brisson et al., 2005). At the cellular level, substance P slowly depolarizes nonpacemaker neurons, leading to an increase of the firing rate of action potentials. Substance P also dramatically activates CS pacemakers and to a lesser extent CI pacemakers, causing an increase in burst amplitude, frequency, and duration (Pena and Ramirez, 2004b).
This discussion leads to the important conclusion that every parameter in rhythm generation is controlled by multiple modulators. It must be emphasized that for simplicity we focused in this review only on three neuromodulators, even though there are many additional modulators that play equally important roles in the neuronal control of breathing (Doi and Ramirez, 2008). Another important conclusion is that the same modulator can exert many different effects on rhythmic activity. Thus, the influence of any given neuromodulator occurs in concert with many other aminergic and peptidergic substances (Fig. 1). The complexity of neuromodulation as described here for the respiratory network is reminiscent to the complexity that is well documented in invertebrate neuronal networks (Nusbaum, 2002; Thirumalai and Marder, 2002) and has important implications for all neuronal networks. Different sets of modulators with diverse, convergent, and divergent actions will define different states of a rhythm generating network that may change dependent on the metabolic, developmental or behavioral conditions of an animal. These network states involve a complex orchestration of large sets of different ion-channels, multiple receptors, and numerous neuropeptides and biogenic amines in addition to those described here (Doi and Ramirez, 2008; Grashow et al., 2009; Thoby-Brisson and Simmers, 1998). A modulatory state will, therefore, not only determine how a rhythm is generated, but will also determine the responsiveness of neuronal networks to inputs. Many recent experimental and computational studies suggest that neuronal networks respond differently to inputs if their state is altered by modulators (Destexhe and Contreras, 2006; MacLean et al., 2005; Nadim et al., 2008; Prescott and De Koninck, 2003). In the respiratory network, numerous excitatory and inhibitory inputs that arise from multiple anatomical regions of the brain to modulate breathing in amplitude, frequency, and regularity have been described (Doi and Ramirez, 2008).
40
Unfortunately, only little is known how different competing inputs interact in the context of different neuromodulators. In the respiratory network, excitatory inputs mediated by substance P are only critical for respiratory rhythm generation when levels of serotonin or NE are low, that is, under conditions that resemble more closely the sleep state (Jones, 2005). By contrast, when 5-HT2a receptors and a-1 receptors are fully activated, that is in a state when serotonin and norepinephrine levels are high, inputs mediated by substance P are not critical for respiratory rhythm generation (Doi and Ramirez, 2010). Thus, the convergence of different neuromodulatory inputs (Fig. 1) provides a safety net in case a given modulator or receptor is disturbed. It may also explain the state-dependency of a variety of disorders. In SIDS, there is mounting evidence for a role for 5-HT (Broadbelt et al., 2009; Duncan et al., 2010; Kinney et al., 2009; Paterson et al., 2006; Rand et al., 2007). SIDS (and also sleep apnea) occurs during sleep, when aminergic and presumably also SP levels are reduced. Thus, it seems that the respiratory network is critically dependent on an individual neuromodulator only in a reduced “modulatory state.” By contrast, in the awake animal, it is unlikely that a disruption of a single neuromodulator will have much of a critical effect, since the level of other converging modulators is relatively high. Thus, understanding the convergence of different neuromodulators is not only an interesting basic scientific issue, but will also be important for gaining insights into the neuropathology of breathing disorders. The data summarized above indicate that it is not possible to associate specific network states with the action of individual neuromodulators. Instead, a network state is defined by a complex set of different modulators with highly diverse, convergent, and divergent actions. A particular neuromodulator can differentially act on a variety of receptor subtypes, thereby exerting specific, yet sometimes diverging effects on different parameters of the respiratory network (e.g.,
frequency vs. amplitude). Conversely, different neuromodulators, such as serotonin, SP, or NE, can exert similar effects via different cellular mechanisms. The specificity and types of these modulatory effects are not necessarily preserved across different animal species even if compared between related rodent species and strains. These conclusions are reminiscent of findings in invertebrate systems. Like in the respiratory system, modulatory responses can vary between different species and even across individuals of the same species at different developmental stages (Newcomb and Katz, 2007, 2009; Rehm et al., 2008). Also similar to the situation in the respiratory network is the observation that different neuromodulatory inputs can exert distinct or comparable activity patterns from the same neuronal ensemble (Saideman et al., 2007). Like in these invertebrate networks, it is remarkable that despite the large number of concurrent and rather diverse modulatory processes the various parameters of rhythm generating networks are relatively tightly maintained even when exposed to extreme conditions. This can be illustrated for the control of the respiratory frequency in humans. The breathing frequency of children under normal conditions is 45 13 breaths/min. Acute exposure to 3109 m increases the frequency range on average by only 6 breaths/min to 51.9 15 breaths/min (Yaron et al., 2003). Under extreme altitudes, such as climbing to the Everest (8000 m), only one individual of the American Medical Research Expedition raised the breathing rate to up to 86 breaths/min (West, 2010). Thus, despite the presence of numerous modulatory systems affecting breathing frequency, the range over which this particular parameter is modulated is relatively narrow under normal as well as extreme conditions. For the invertebrate model system, it has been suggested that homeostatic mechanisms play critical roles in maintaining different parameters in a tightly controlled range (Rehm et al., 2008). Although little is known about how these mechanisms regulate the various parameters of
41
the mammalian respiratory rhythm, they must play critical roles in the homeostasis of breathing (Fig. 1) as will be discussed below.
Homeostatic plasticity and other forms of long-term plasticity Homeostatic plasticity is a fundamental mechanism that has been demonstrated in a variety of neural networks. It is crucial to maintain network stability and it affects nearly every aspect of circuit development and function (Fig. 1). One principle mechanism is the activity dependent scaling of neuronal receptors to maintain neurons in a certain firing range (Turrigiano et al., 1998). But, multiple forms of homeostatic plasticity have been described in a variety of brain areas (Aizenman et al., 2003; Desai et al., 1999; Ibata et al., 2008; Koch et al., 2010; Marder and Goaillard, 2006; Stellwagen and Malenka, 2006). Homeostatic plasticity has been shown to regulate pre- and postsynaptic scaling of excitation and inhibition. Such changes involve alterations in ion-channel composition which tune and anchor intrinsic neuronal excitability to a defined range of activity. The mechanisms that define and regulate these setpoints of cellular activity are not well understood. Alterations in internal calcium levels [Ca2þ]i have been discussed as a possible mechanism balancing synaptic homeostasis in neuronal networks (Turrigiano, 2008). In fact, since calcium fluctuations are directly related to the activity of the cells, it has been hypothesized that calcium plays important roles in regulating many forms of plasticity (Grubb and Burrone, 2010; Lisman et al., 2002; Malenka and Bear, 2004; Zhang and Linden, 2003). Consistent with this hypothesis, blockade of calcium channels can trigger upscaling of excitatory synaptic terminals (Ibata et al., 2008). Moreover, pharmacological blockade of calcium dependent kinases (i.e., the CaMK- family) prevents the effects of activity deprivation on excitatory synaptic transmission (Thiagarajan et al., 2002). Besides intracellular
calcium as an activity sensor, multiple other molecules have been implicated to be involved to mediate homeostatic synaptic plasticity (Aoto et al., 2008; Koch et al., 2010; Rutherford et al., 1998; Stellwagen and Malenka, 2006). Interestingly, some of these molecules are part of the inflammatory pathway and include prostaglandin-E2 (PGE2) and tumor necrosis factor-a (TNF-a) (Koch et al., 2010; Stellwagen and Malenka, 2006). Activation of glia cells and TNF-a signaling are necessary for synaptic upscaling in cortical neurons, since blocking the activation of TNF-a receptors prevents the upscaling (Stellwagen and Malenka, 2006). Recently PGE2, the major reaction product of the Cyclooxygenase-2 enzymes (COX-2) was reported to acutely inhibit network activity in the neocortex, but if chronically applied led to a predominantly presynaptic increase in excitatory synaptic transmission (Koch et al., 2010). The COX-2 pathway is directly activated by hypoxia through the hypoxia induced factor-1a (HIF-1a) and PGE2 (Fig. 1) has been reported to be an important regulator of the respiratory rhythm generator (Hofstetter et al., 2007). Thus, the link of activity-dependent scaling and inflammation could potentially be important for understanding pathophysiological changes that occur in the cardiorespiratory control following chronic intermittent hypoxia (IH). In the context of cardio-respiratory homeostasis, several studies described long-term regulatory processes for the NTS, an area that is critical for sensory integration not only in the context of breathing but also other autonomic functions (Greenberg et al., 1999a, 1999b; Kline et al., 2007; Zhou et al., 1997). Activity-dependent long-term depression (LTD) can be elicited in a subset of NTS neurons (Zhou et al., 1997), and chronic exposure to IH causes an activity dependent synaptic downscaling of excitatory synaptic transmission (Kline et al., 2007). Perhaps, the best studied form of long-term plasticity (Fig. 1) in the respiratory system is the response to intermittent hypoxia. Investigating
42
the neuronal consequences of repetitive episodes of hypoxia (Acute Intermittent Hypoxia, AIH) is clinically very relevant, as this condition is associated with a variety of breathing disorders including Rett syndrome and obstructive sleep apnea (Lee et al., 2009; Weese-Mayer et al., 2006). AIH causes persistent increases in respiratory frequency and amplitude of integrated motor neuronal bursts in vivo (Baker and Mitchell, 2000; Hayashi et al., 1993; Millhorn et al., 1980; Turner and Mitchell, 1997). These changes persisting for 90 min are collectively referred to as long-term facilitation (LTF) (Fuller et al., 2000; Powell et al., 1998). The degree of influence varies with preparation (Bach and Mitchell, 1996; Turner and Mitchell, 1997), animal strain (Fuller et al., 2000), gender (Zabka et al., 2006), and experimental conditions (Baker and Mitchell, 2000). AIH causes changes at multiple levels of the respiratory system, including the carotid body (Dogas et al., 1995; Powell et al., 1998), and motor nuclei (Kinkead et al., 1998). The likely site for the long-term frequency modulation is the pre-BötC, since intermittent hypoxia causes a long-lasting frequency increase within the preBötC that persists for 90 min after repetitive hypoxic episodes (Blitz and Ramirez, 2002). AIH causes immediate changes in the modulatory milieu and long-term changes involving gene expression. For this to occur, the intermittent pattern and not duration of hypoxia is critical, as even prolonged hypoxic exposure does not evoke LTF (Baker and Mitchell, 2000). AIH causes the intermittent production of reactive oxygen species (MacFarlane and Mitchell, 2009; Pawar et al., 2009) and release of aminergic neuromodulators. Blockade of serotonin receptors abolishes both motor amplitude and frequency LTF in vivo (Bach and Mitchell, 1996; Kinkead et al., 1998). It has been hypothesized (Bach and Mitchell, 1996; Fuller et al., 2000), that LTF is induced by chemoreceptor activation of serotonergic raphe neurons (Erickson and Millhorn, 1994; McCrimmon et al., 1997; Pawar et al., 2009). However, not only serotonin, but also repeated
exposure to norepinephrine elicits LTF in vitro (Bocchiaro and Feldman, 2004; Neverova et al., 2007). Both modulatory systems seem to interact (Kinkead et al., 2001). There is accumulating evidence indicating that long-lasting changes in intracellular signaling molecules involving a reactive oxygen species activated PKC pathway cause new synthesis of BDNF acting on TrkB receptors (Fig. 1), which in turn will increase postsynaptic glutamate receptor density and thereby increases glutamatergic synaptic transmission (BakerHerman et al., 2004). Thus, taken together these studies suggest that neuronal networks involved in respiratory control are regulated by multiple forms of long-term synaptic plasticity.
Conclusion The convergence of reconfiguration, neuromodulation, state dependency, and homeostatic plasticity (Fig. 1) provides multiple mechanisms by which the pre-BötC is capable of generating multiple, dynamically responsive yet stable sets of respiratory rhythmic activity. As discussed, the pre-BötC retains its ability to generate stable rhythmicity throughout changes in the oxygen environment using both network reconfiguration and neuromodulation. The ability of the pre-BötC to retain stable rhythmicity and transmit this activity to a variety of motor outputs is not a trivial property, as many neuronal networks within the mammalian nervous system shut down during hypoxia in an attempt to conserve the energy during oxygen limiting conditions. This is readily seen in the hippocampus (Krnjevic and Leblond, 1987; Garcia et al., 2010), neocortex (Jiang and Haddad, 1992), and striatum (Calabresi et al., 1995). Hence, sustaining stable rhythmicity in the pre-BötC even during severe levels of hypoxia underscores the importance of this network, as loss of the respiratory rhythmic activity would ultimately lead to death. This robustness may be one reason why the pre-BötC is capable of generating behaviorally relevant rhythmicity even when isolated from the
43
rest of the nervous system (Fig. 2), thus facilitating a rigorous cellular and network analysis. In the intact animal, the pre-BötC constitutes a core network which operates within a larger network of interconnected nuclei that contribute not only to the generation of the respiratory rhythm, but also to the plasticity and state-dependency that characterizes breathing (Fig. 1). The principles gained in the respiratory network apply to all neuronal network functions not only to the mammalian nervous system, but also to the networks of invertebrates. The emerging conclusions gained by studying the respiratory network are surprisingly complex, and we are clearly only touching the surface of our understanding. Yet, this degree of complexity is obviously necessary to guarantee that the respiratory network is very adaptive and at the same time sufficiently stable to maintain regular network activity even under adverse environmental and behavioral conditions. Acknowledgment We acknowledge the financial support from various grants awarded by the National Institute of Health to JMR. References Aizenman, C. D., Akerman, C. J., Jensen, K. R., & Cline, H. T. (2003). Visually driven regulation of intrinsic neuronal excitability improves stimulus detection in vivo. Neuron, 39, 831–842. Alheid, G. F., Gray, P. A., Jiang, M. C., Feldman, J. L., & McCrimmon, D. R. (2002). Parvalbumin in respiratory neurons of the ventrolateral medulla of the adult rat. Journal of Neurocytology, 31, 693–717. Alheid, G. F., & McCrimmon, D. R. (2008). The chemical neuroanatomy of breathing. Respiratory Physiology & Neurobiology, 164, 3–11. Alheid, G. F., Milsom, W. K., & McCrimmon, D. R. (2004). Pontine influences on breathing: An overview. Respiratory Physiology & Neurobiology, 143, 105–114. Aoto, J., Nam, C. I., Poon, M. M., Ting, P., & Chen, L. (2008). Synaptic signaling by all-trans retinoic acid in homeostatic synaptic plasticity. Neuron, 60, 308–320.
Bach, K. B., & Mitchell, G. S. (1996). Hypoxia-induced long-term facilitation of respiratory activity is serotonin dependent. Respiration Physiology, 104, 251–260. Baker, T. L., & Mitchell, G. S. (2000). Episodic but not continuous hypoxia elicits long-term facilitation of phrenic motor output in rats. The Journal of Physiology, 529(Pt 1), 215–219. Baker-Herman, T. L., Fuller, D. D., Bavis, R. W., Zabka, A. G., Golder, F. J., Doperalski, N. J., et al. (2004). BDNF is necessary and sufficient for spinal respiratory plasticity following intermittent hypoxia. Nature Neuroscience, 7, 48–55. Ballanyi, K., Volker, A., & Richter, D. (1994). Anoxia induced functionalinactivation of neonatal respiratory neurons in vitro. Neuroreport, 6, 165–168. Berkowitz, A., Roberts, A., & Soffe, S. R. (2010). Roles for multifunctional and specialized spinal interneurons during motor pattern generation in tadpoles, zebrafish larvae, and turtles. Frontiers in Behavioral Neuroscience, 4, 36. Besnard, S., Denise, P., Cappelin, B., Dutschmann, M., & Gestreau, C. (2009). Stimulation of the rat medullary raphe nuclei induces differential responses in respiratory muscle activity. Respiratory Physiology & Neurobiology, 165, 208–214. Blitz, D. M., & Ramirez, J. M. (2002). Long-term modulation of respiratory network activity following anoxia in vitro. Journal of Neurophysiology, 87, 2964–2971. Bocchiaro, C. M., & Feldman, J. L. (2004). Synaptic activity-independent persistent plasticity in endogenously active mammalian motoneurons. Proceedings of the National Academy of Sciences of the United States of America, 101, 4292–4295. Broadbelt, K. G., Barger, M. A., Paterson, D. S., Holm, I. A., Haas, E. A., Krous, H. F., et al. (2009). Serotonin-related FEV gene variant in the sudden infant death syndrome is a common polymorphism in the African-American population. Pediatric Research, 66, 631–635. Brownstone, R. M., Jordan, L. M., Kriellaars, D. J., Noga, B. R., & Shefchyk, S. J. (1992). On the regulation of repetitive firing in lumbar motoneurones during fictive locomotion in the cat. Experimental Brain Research, 90, 441–455. Calabresi, P., Pisani, A., Mercuri, N. B., & Bernardi, G. (1995). Hypoxia-induced electrical changes in striatal neurons. Journal of Cerebral Blood Flow and Metabolism, 15, 1141–1145. Chamberlin, N. L., & Saper, C. B. (1994). Topographic organization of respiratory responses to glutamate microstimulation of the parabrachial nucleus in the rat. The Journal of Neuroscience, 14, 6500–6510. Chevalier, M., Ben-Mabrouk, F., & Tryba, A. K. (2008). Background sodium current underlying respiratory rhythm regularity. The European Journal of Neuroscience, 28, 2423–2433.
44 Cohen, M. I. (1971). Switching of the respiratory phases and evoked phrenic responses produced by rostral pontine electrical stimulation. The Journal of Physiology, 217, 133–158. Coles, S. K., & Dick, T. E. (1996). Neurones in the ventrolateral pons are required for post-hypoxic frequency decline in rats. The Journal of Physiology, 497(Pt 1), 79–94. Dahlstrom, A., & Fuxe, K. (1964). Localization of monoamines in the lower brain stem. Experientia, 20, 398–399. Darbon, P., Yvon, C., Legrand, J. C., & Streit, J. (2004). INaP underlies intrinsic spiking and rhythm generation in networks of cultured rat spinal cord neurons. The European Journal of Neuroscience, 20, 976–988. Davenport, P. W., Reep, R. L., & Thompson, F. J. (2010). Phrenic nerve afferent activation of neurons in the cat SI cerebral cortex. The Journal of Physiology, 588, 873–886. Del Negro, C. A., Koshiya, N., Butera, R. J., Jr., & Smith, J. C. (2002a). Persistent sodium current, membrane properties and bursting behavior of pre-Botzinger complex inspiratory neurons in vitro. Journal of Neurophysiology, 88, 2242–2250. Del Negro, C. A., Morgado-Valle, C., & Feldman, J. L. (2002b). Respiratory rhythm: An emergent network property? Neuron, 34, 821–830. Del Negro, C., Morgado-Valle, C., Hayes, J., MacKay, D., Pace, R., Crowder, E., et al. (2005). Sodium and calcium current-mediated pacemaker neurons and respiratory rhythm generation. The Journal of Neuroscience, 25, 446–453. Desai, N. S., Rutherford, L. C., & Turrigiano, G. G. (1999). Plasticity in the intrinsic excitability of cortical pyramidal neurons. Nature Neuroscience, 2, 515–520. Destexhe, A., & Contreras, D. (2006). Neuronal computations with stochastic network states. Science, 314, 85–90. Di Pasquale, E., Monteau, R., & Hilaire, G. (1994). Endogenous serotonin modulates the fetal respiratory rhythm: An in vitro study in the rat. Brain Research. Developmental Brain Research, 80, 222–232. Di Pasquale, E., Morin, D., Monteau, R., & Hilaire, G. (1992). Serotonergic modulation of the respiratory rhythm generator at birth: An in vitro study in the rat. Neuroscience Letters, 143, 91–95. Dick, T. E., Bellingham, M. C., & Richter, D. W. (1994). Pontine respiratory neurons in anesthetized cats. Brain Res, 636, 259–269. Dogas, Z., Stuth, E. A., Hopp, F. A., McCrimmon, D. R., & Zuperku, E. J. (1995). NMDA receptor-mediated transmission of carotid body chemoreceptor input to expiratory bulbospinal neurones in dogs. The Journal of Physiology, 487(Pt 3), 639–651. Doi, A., & Ramirez, J. M. (2008). Neuromodulation and the orchestration of the respiratory rhythm. Respiratory Physiology & Neurobiology, 164, 96–104.
Doi, A., & Ramirez, J. (2010). State-Dependent Interactions between Excitatory Neuromodulators in the Control of Breathing. The Journal of Neuroscience, 30, 8251–8262. Duncan, J. R., Paterson, D. S., Hoffman, J. M., Mokler, D. J., Borenstein, N. S., Belliveau, R. A., et al. (2010). Brainstem serotonergic deficiency in sudden infant death syndrome. JAMA, 303, 430–437. Dutschmann, M., & Herbert, H. (1996). The Kolliker-Fuse nucleus mediates the trigeminally induced apnoea in the rat. Neuroreport, 7, 1432–1436. England, S., Melton, J., Douse, M., & Duffin, J. (1995). Activity of respiratory neurons during hypoxia in the chemodenervated cat. Journal of Applied Physiology, 78, 856–861. Erickson, J. T., & Millhorn, D. E. (1994). Hypoxia and electrical stimulation of the carotid sinus nerve induce Fos-like immunoreactivity within catecholaminergic and serotoninergic neurons of the rat brainstem. The Journal of Comparative Neurology, 348, 161–182. Feldman, J. L., & Del Negro, C. A. (2006). Looking for inspiration: New perspectives on respiratory rhythm. Nature Reviews. Neuroscience, 7, 232–242. Feldman, J. L., Mitchell, G. S., & Nattie, E. E. (2003). Breathing: Rhythmicity, plasticity, chemosensitivity. Annual Review of Neuroscience, 26, 239–266. Fuller, D. D., Bach, K. B., Baker, T. L., Kinkead, R., & Mitchell, G. S. (2000). Long term facilitation of phrenic motor output. Respiration Physiology, 121, 135–146. Fuxe, K. (1965). Evidence for the existence of monoamine neurons in the central nervous system. IV. Distribution of monoamine nerve terminals in the central nervous system. Acta physiologica Scandinavica. Supplementum, 247, 37. Garcia, A. J., 3rd, Putnam, R. W., & Dean, J. B. (2010). Hyperbaric hyperoxia and normobaric reoxygenation increase excitability and activate oxygen-induced potentiation (OxIP) in CA1 hippocampal neurons. Journal of Applied Physiology, 109, 804–819. Grashow, R., Brookings, T., & Marder, E. (2009). Reliable neuromodulation from circuits with variable underlying structure. Proceedings of the National Academy of Sciences of the United States of America, 106, 11742–11746. Gray, P. A., Janczewski, W. A., Mellen, N., McCrimmon, D. R., & Feldman, J. L. (2001). Normal breathing requires preBotzinger complex neurokinin-1 receptor-expressing neurons. Nature Neuroscience, 4, 927–930. Gray, P. A., Rekling, J. C., Bocchiaro, C. M., & Feldman, J. L. (1999). Modulation of respiratory frequency by peptidergic input to rhythmogenic neurons in the pre-Botzinger complex. Science, 286, 1566–1568. Greenberg, H. E., Sica, A., Batson, D., & Scharf, S. M. (1999a). Chronic intermittent hypoxia increases sympathetic responsiveness to hypoxia and hypercapnia. Journal of Applied Physiology, 86, 298–305.
45 Greenberg, H., Sica, A., Scharf, S., & Ruggiero, D. (1999b). Expression of c-fos in the rat brainstem after chronic inttermittent hypoxia. Brain Research, 816, 638–645. Grubb, M. S., & Burrone, J. (2010). Activity-dependent relocation of the axon initial segment fine-tunes neuronal excitability. Nature, 465, 1070–1074. Guyenet, P. G., Koshiya, N., Huangfu, D., Verberne, A. J., & Riley, T. A. (1993). Central respiratory control of A5 and A6 pontine noradrenergic neurons. The American Journal of Physiology, 264, R1035–R1044. Guyenet, P. G., Sevigny, C. P., Weston, M. C., & Stornetta, R. L. (2002). Neurokinin-1 receptor-expressing cells of the ventral respiratory group are functionally heterogeneous and predominantly glutamatergic. The Journal of Neuroscience, 22, 3806–3816. Haddad, G., & Mellins, R. (1984). Hypoxia and respiratory control in early life. Annual Review of Physiology, 46, 629–643. Haque, Z., Lee, T. K., Inoue, T., Luk, C., Hasan, S. U., Lukowiak, K., et al. (2006). An identified central patterngenerating neuron co-ordinates sensory-motor components of respiratory behavior in Lymnaea. The European Journal of Neuroscience, 23, 94–104. Harper, R. M. (2000a). Sudden infant death syndrome: A failure of compensatory cerebellar mechanisms? Pediatric Research, 48, 140–142. Harper, R. M., Woo, M. A., & Alger, J. R. (2000b). Visualization of sleep influences on cerebellar and brainstem cardiac and respiratory control mechanisms. Brain Research Bulletin, 53, 125–131. Harris, G., Mills, H., Wragg, R., Hapiak, V., Castelletto, M., Korchnak, A., et al. (2010). The monoaminergic modulation of sensory-mediated aversive responses in Caenorhabditis elegans requires glutamatergic/peptidergic cotransmission. The Journal of Neuroscience, 30, 7889–7899. Harris-Warrick, R., & Johnson, B. (2010). Checks and balances in neuromodulation. Frontiers in Behavioral Neuroscience, 4, pii 47. Hayashi, F., Coles, S. K., Bach, K. B., Mitchell, G. S., & McCrimmon, D. R. (1993). Time-dependent phrenic nerve responses to carotid afferent activation: Intact vs. decerebellate rats. The American Journal of Physiology, 265, R811–R819. Hayes, J. A., & Del Negro, C. A. (2007). Neurokinin receptorexpressing pre-Botzinger complex neurons in neonatal mice studied in vitro. Journal of Neurophysiology, 97, 4215–4224. Hilaire, G. (2006). Endogenous noradrenaline affects the maturation and function of the respiratory network: Possible implication for SIDS. Autonomic Neuroscience, 126–127, 320–331. Hilaire, G., Bou, C., & Monteau, R. (1997). Serotonergic modulation of central respiratory activity in the neonatal mouse: An in vitro study. European Journal of Pharmacology, 329, 115–120.
Hofstetter, A. O., Saha, S., Siljehav, V., Jakobsson, P. J., & Herlenius, E. (2007). The induced prostaglandin E2 pathway is a key regulator of the respiratory response to infection and hypoxia in neonates. Proceedings of the National Academy of Sciences of the United States of America, 104, 9894–9899. Hokfelt, T., Johansson, O., & Goldstein, M. (1984). Chemical anatomy of the brain. Science, 225, 1326–1334. Holtman, J. R., Jr., Anastasi, N. C., Norman, W. P., & Dretchen, K. L. (1986). Effect of electrical and chemical stimulation of the raphe obscurus on phrenic nerve activity in the cat. Brain Research, 362, 214–220. Holtman, J. R., Jr., Marion, L. J., & Speck, D. F. (1990). Origin of serotonin-containing projections to the ventral respiratory group in the rat. Neuroscience, 37, 541–552. Hudson, A. E., & Prinz, A. A. (2010). Conductance ratios and cellular identity. PLoS Computational Biology, 6, e1000838. Ibata, K., Sun, Q., & Turrigiano, G. G. (2008). Rapid synaptic scaling induced by changes in postsynaptic firing. Neuron, 57, 819–826. Janczewski, W. A., & Feldman, J. L. (2006). Distinct rhythm generators for inspiration and expiration in the juvenile rat. The Journal of Physiology, 570, 407–420. Janczewski, W. A., Onimaru, H., Homma, I., & Feldman, J. L. (2002). Opioid-resistant respiratory pathway from the preinspiratory neurones to abdominal muscles: In vivo and in vitro study in the newborn rat. The Journal of Physiology, 545, 1017–1026. Jiang, C., & Haddad, G. G. (1992). Differential responses of neocortical neurons to glucose and/or O2 deprivation in the human and rat. Journal of Neurophysiology, 68, 2165–2173. Jing, J., & Gillette, R. (2003). Directional avoidance turns encoded by single interneurons and sustained by multifunctional serotonergic cells. The Journal of Neuroscience, 23, 3039–3051. Jones, B. E. (2005). From waking to sleeping: Neuronal and chemical substrates. Trends in Pharmacological Sciences, 26, 578–586. Kachidian, P., Poulat, P., Marlier, L., & Privat, A. (1991). Immunohistochemical evidence for the coexistence of substance P, thyrotropin-releasing hormone, GABA, methionine-enkephalin, and leucin-enkephalin in the serotonergic neurons of the caudal raphe nuclei: A dual labeling in the rat. Journal of Neuroscience Research, 30, 521–530. Kinkead, R., Bach, K. B., Johnson, S. M., Hodgeman, B. A., & Mitchell, G. S. (2001). Plasticity in respiratory motor control: Intermittent hypoxia and hypercapnia activate opposing serotonergic and noradrenergic modulatory systems. Comparative Biochemistry and Physiology. Part A, Molecular & Integrative Physiology, 130, 207–218. Kinkead, R., Zhan, W. Z., Prakash, Y. S., Bach, K. B., Sieck, G. C., & Mitchell, G. S. (1998). Cervical dorsal
46 rhizotomy enhances serotonergic innervation of phrenic motoneurons and serotonin-dependent long-term facilitation of respiratory motor output in rats. The Journal of Neuroscience, 18, 8436–8443. Kinney, H. C., Richerson, G. B., Dymecki, S. M., Darnall, R. A., & Nattie, E. E. (2009). The brainstem and serotonin in the sudden infant death syndrome. Annual Review of Pathology, 4, 517–550. Kline, D. D., Ramirez-Navarro, A., & Kunze, D. L. (2007). Adaptive depression in synaptic transmission in the nucleus of the solitary tract after in vivo chronic intermittent hypoxia: Evidence for homeostatic plasticity. The Journal of Neuroscience, 27, 4663–4673. Kobayashi, S., Onimaru, H., Inoue, M., Inoue, T., & Sasa, R. (2005). Localization and properties of respiratory neurons in the rostral pons of the newborn rat. Neuroscience, 134, 317–325. Koch, H., Huh, S.-E., Elsen, F. P., Carroll, M., Hodge, R., Bedogni, F., et al. (2010). Prostaglandin E2 induced synaptic plasticity in neocortical networks of organotypic slice cultures. The Journal of Neuroscience, 30, 11678–11687. Koizumi, H., Smerin, S. E., Yamanishi, T., Moorjani, B. R., Zhang, R., & Smith, J. C. (2010). TASK channels contribute to the Kþ-dominated leak current regulating respiratory rhythm generation in vitro. The Journal of Neuroscience, 30, 4273–4284. Koizumi, H., Wilson, C., Wong, S., Yamanishi, T., Koshiya, N., & Smith, J. (2008). Functional imaging, spatial reconstruction, and biophysical analysis of a respiratory motor circuit isolated in vitro. The Journal of Neuroscience, 28, 2353–2365. Kolta, A., Brocard, F., Verdier, D., & Lund, J. P. (2007). A review of burst generation by trigeminal main sensory neurons. Archives of Oral Biology, 52, 325–328. Kramer, M. A., Roopun, A. K., Carracedo, L. M., Traub, R. D., Whittington, M. A., & Kopell, N. J. (2008). Rhythm generation through period concatenation in rat somatosensory cortex. PLoS Computational Biology, 4, e1000169. Krnjevic, K., & Leblond, J. (1987). Anoxia reversibly suppresses neuronal calcium currents in rat hippocampal slices. Canadian Journal of Physiology and Pharmacology, 65, 2157–2161. Lalley, P. M. (1986). Responses of phrenic motoneurones of the cat to stimulation of medullary raphe nuclei. The Journal of Physiology, 380, 349–371. Lee, D. S., Badr, M. S., & Mateika, J. H. (2009). Progressive augmentation and ventilatory long-term facilitation are enhanced in sleep apnoea patients and are mitigated by antioxidant administration. The Journal of Physiology, 587, 5451–5467. Lieske, S. P., & Ramirez, J. M. (2006a). Pattern-specific synaptic mechanisms in a multifunctional network. I. Effects
of alterations in synapse strength. Journal of Neurophysiology, 95, 1323–1333. Lieske, S. P., & Ramirez, J. M. (2006b). Pattern-specific synaptic mechanisms in a multifunctional network. II. Intrinsic modulation by metabotropic glutamate receptors. Journal of Neurophysiology, 95, 1334–1344. Lieske, S. P., Thoby-Brisson, M., Telgkamp, P., & Ramirez, J. M. (2000). Reconfiguration of the neural network controlling multiple breathing patterns: Eupnea, sighs and gasps [see comment]. Nature Neuroscience, 3, 600–607. Lisman, J., Schulman, H., & Cline, H. (2002). The molecular basis of CaMKII function in synaptic and behavioural memory. Nature Reviews. Neuroscience, 3, 175–190. Macfarlane, P. M., & Mitchell, G. S. (2009). Episodic spinal serotonin receptor activation elicits long-lasting phrenic motor facilitation by an NADPH oxidase-dependent mechanism. The Journal of Physiology, 587, 5469–5481. Maclean, J. N., Watson, B. O., Aaron, G. B., & Yuste, R. (2005). Internal dynamics determine the cortical response to thalamic stimulation. Neuron, 48, 811–823. Malenka, R. C., & Bear, M. F. (2004). LTP and LTD: An embarrassment of riches. Neuron, 44, 5–21. Marder, E., & Goaillard, J. M. (2006). Variability, compensation and homeostasis in neuron and network function. Nature Reviews. Neuroscience, 7, 563–574. McCrimmon, D. R., Alheid, G. F., Jiang, M., Calandriello, T., & Topgi, A. (2004). Converging functional and anatomical evidence for novel brainstem respiratory compartments in the rat. Advances in Experimental Medicine and Biology, 551, 101–105. McCrimmon, D. R., Zuperku, E. J., Hayashi, F., Dogas, Z., Hinrichsen, C. F., Stuth, E. A., et al. (1997). Modulation of the synaptic drive to respiratory premotor and motor neurons. Respiration Physiology, 110, 161–176. McKay, L. C., Janczewski, W. A., & Feldman, J. L. (2005). Sleepdisordered breathing after targeted ablation of pre-Botzinger complex neurons. Nature Neuroscience, 8, 1142–1144. Meyrand, P., Simmers, J., & Moulins, M. (1991). Construction of a pattern-generating circuit with neurons of different networks. Nature, 351, 60–63. Millhorn, D., Eldridge, F., & Waldrop, T. (1980). Prolonged stimulation of respiration by a new central neural mechanism. Respiration Physiology, 41, 87–103. Milsom, W. K., Chatburn, J., & Zimmer, M. B. (2004). Pontine influences on respiratory control in ectothermic and heterothermic vertebrates. Respiratory Physiology & Neurobiology, 143, 263–280. Morgado-Valle, C., & Feldman, J. L. (2004). Depletion of substance P and glutamate by capsaicin blocks respiratory rhythm in neonatal rat in vitro. The Journal of Physiology, 555, 783–792. Morin, D., Hennequin, S., Monteau, R., & Hilaire, G. (1990). Serotonergic influences on central respiratory activity:
47 An in vitro study in the newborn rat. Brain Research, 535, 281–287. Morschel, M., & Dutschmann, M. (2009). Pontine respiratory activity involved in inspiratory/expiratory phase transition. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 364, 2517–2526. Nadim, F., Brezina, V., Destexhe, A., & Linster, C. (2008). State dependence of network output: Modeling and experiments. The Journal of Neuroscience, 28, 11806–11813. Neubauer, J., Melton, J., & Edelman, N. (1990). Modulation of respiration during brain hypoxia. Journal of Applied Physiology, 68, 441–451. Neverova, N. V., Saywell, S. A., Nashold, L. J., Mitchell, G. S., & Feldman, J. L. (2007). Episodic stimulation of alpha1adrenoreceptors induces protein kinase C-dependent persistent changes in motoneuronal excitability. The Journal of Neuroscience, 27, 4435–4442. Newcomb, J. M., & Katz, P. S. (2007). Homologues of serotonergic central pattern generator neurons in related nudibranch molluscs with divergent behaviors. Journal of Comparative Physiology. A, Neuroethology, Sensory, Neural, and Behavioral Physiology, 193, 425–443. Newcomb, J. M., & Katz, P. S. (2009). Different functions for homologous serotonergic interneurons and serotonin in species-specific rhythmic behaviours. Proceedings. Biological Sciences, 276, 99–108. Nieber, K., Sevcik, J., & Illes, P. (1995). Hypoxic changes in rat locus coeruleus neurons in vitro. The Journal of Physiology, 486(Pt 1), 33–46. Nusbaum, M. P. (2002). Regulating peptidergic modulation of rhythmically active neural circuits. Brain, Behavior and Evolution, 60, 378–387. Onimaru, H., & Homma, I. (2003). A novel functional neuron group for respiratory rhythm generation in the ventral medulla. The Journal of Neuroscience, 23, 1478–1486. Onimaru, H., Kumagawa, Y., & Homma, I. (2006). Respiration-related rhythmic activity in the rostral medulla of newborn rats. Journal of Neurophysiology, 96, 55–61. Paterson, D. S., Hilaire, G., & Weese-Mayer, D. E. (2009). Medullary serotonin defects and respiratory dysfunction in sudden infant death syndrome. Respiratory Physiology & Neurobiology, 168, 133–143. Paterson, D. S., Trachtenberg, F. L., Thompson, E. G., Belliveau, R. A., Beggs, A. H., Darnall, R., et al. (2006). Multiple serotonergic brainstem abnormalities in sudden infant death syndrome. JAMA, 296, 2124–2132. Paton, J. F., Abdala, A. P., Koizumi, H., Smith, J. C., & St-John, W. M. (2006). Respiratory rhythm generation during gasping depends on persistent sodium current. Nature Neuroscience, 9, 311–313. Pawar, A., Nanduri, J., Yuan, G., Khan, S. A., Wang, N., Kumar, G. K., et al. (2009). Reactive oxygen species-dependent endothelin signaling is required for augmented hypoxic
sensory response of the neonatal carotid body by intermittent hypoxia. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology, 296, R735–R742. Pena, F., & Aguileta, M. A. (2007). Effects of riluzole and flufenamic acid on eupnea and gasping of neonatal mice in vivo. Neuroscience Letters, 415, 288–293. Pena, F., Meza-Andrade, R., Paez-Zayas, V., & GonzalezMarin, M. C. (2008). Gasping generation in developing Swiss-Webster mice in vitro and in vivo. Neurochemical Research, 33, 1492–1500. Pena, F., Parkis, M., Tryba, A., & Ramirez, J. (2004a). Differential contribution of pacemaker properties to the generation of respiratory rhythms during normoxia and hypoxia. Neuron, 43, 105–117. Pena, F., & Ramirez, J. M. (2002). Endogenous activation of serotonin-2A receptors is required for respiratory rhythm generation in vitro. The Journal of Neuroscience, 22, 11055–11064. Pena, F., & Ramirez, J. M. (2004b). Substance P-mediated modulation of pacemaker properties in the mammalian respiratory network. The Journal of Neuroscience, 24, 7549–7556. Peng, Y., Overholt, J., Kline, D., Kumar, G., & Prabhakar, N. (2003). Induction of sensory long-term facilitation in the carotid body by intermittent hypoxia: Implications for recurrent apneas. Proceedings of the National Academy of Sciences of the United States of America, 100, 10073–10078. Pennartz, C. M., Bierlaagh, M. A., & Geurtsen, A. M. (1997). Cellular mechanisms underlying spontaneous firing in rat suprachiasmatic nucleus: Involvement of a slowly inactivating component of sodium current. Journal of Neurophysiology, 78, 1811–1825. Peyronnet, J., Roux, J. C., Perrin, D., Pequignot, J. M., Lagercrantz, H., & Dalmaz, Y. (2003). Prenatal hypoxia and early postnatal maturation of the chemoafferent pathway. Advances in Experimental Medicine and Biology, 536, 525–533. Powell, F. L., Milsom, W. K., & Mitchell, G. S. (1998). Time domains of the hypoxic ventilatory response. Respiration Physiology, 112, 123–134. Prescott, S. A., & De Koninck, Y. (2003). Gain control of firing rate by shunting inhibition: Roles of synaptic noise and dendritic saturation. Proceedings of the National Academy of Sciences of the United States of America, 100, 2076–2081. Ptak, K., Yamanishi, T., Aungst, J., Milescu, L. S., Zhang, R., Richerson, G. B., et al. (2009). Raphe neurons stimulate respiratory circuit activity by multiple mechanisms via endogenously released serotonin and substance P. The Journal of Neuroscience, 29, 3720–3737. Ramirez, J. (1998a). Reconfiguration of the respiratory network at the onset of locust flight. Journal of Neurophysiology, 80, 3137–3147. Ramirez, J. M., & Pearson, K. G. (1991). Octopamine induces bursting and plateau potentials in insect neurones. Brain Research, 549, 332–337.
48 Ramirez, J. M., & Pearson, K. G. (1993). Alteration of bursting properties in interneurons during locust flight. Journal of Neurophysiology, 70, 2148–2160. Ramirez, J. M., Quellmalz, U. J., & Richter, D. W. (1996). Postnatal changes in the mammalian respiratory network as revealed by the transverse brainstem slice of mice. The Journal of Physiology, 491(Pt 3), 799–812. Ramirez, J., Quellmalz, U., Wilken, B., & Richter, D. (1998b). The hypoxic response of neurones within the in vitro mammalian respiratory network. The Journal of Physiology, 507, 571–582. Ramirez, J. M., Schwarzacher, S. W., Pierrefiche, O., Olivera, B. M., & Richter, D. W. (1998c). Selective lesioning of the cat pre-Botzinger complex in vivo eliminates breathing but not gasping. The Journal of Physiology, 507(Pt 3), 895–907. Ramirez, J. M., Telgkamp, P., Elsen, F. P., Quellmalz, U. J., & Richter, D. W. (1997). Respiratory rhythm generation in mammals: Synaptic and membrane properties. Respiration Physiology, 110, 71–85. Rand, C. M., Berry-Kravis, E. M., Zhou, L., Fan, W., & Weese-Mayer, D. E. (2007). Sudden infant death syndrome: Rare mutation in the serotonin system FEV gene. Pediatric Research, 62, 180–182. Rehm, K. J., Deeg, K. E., & Marder, E. (2008). Developmental regulation of neuromodulator function in the stomatogastric ganglion of the lobster, Homarus americanus. The Journal of Neuroscience, 28, 9828–9839. Ribeiro-Da-Silva, A., & Hokfelt, T. (2000). Neuroanatomical localisation of Substance P in the CNS and sensory neurons. Neuropeptides, 34, 256–271. Richter, D., Bischoff, A., Anders, K., Bellingham, M., & Windhorst, U. (1991). Response of the medullary respiratory network of the cat to hypoxia. The Journal of Physiology, 443, 231–256. Roux, J. C., Dura, E., & Villard, L. (2008). Tyrosine hydroxylase deficit in the chemoafferent and the sympathoadrenergic pathways of the Mecp2 deficient mouse. Neuroscience Letters, 447, 82–86. Roux, J. C., Mamet, J., Perrin, D., Peyronnet, J., Royer, C., Cottet-Emard, J. M., et al. (2003). Neurochemical development of the brainstem catecholaminergic cell groups in rat. Journal of Neural Transmission, 110, 51–65. Roux, J. C., Pequignot, J. M., Dumas, S., Pascual, O., Ghilini, G., Pequignot, J., et al. (2000). O2-sensing after carotid chemodenervation: Hypoxic ventilatory responsiveness and upregulation of tyrosine hydroxylase mRNA in brainstem catecholaminergic cells. The European Journal of Neuroscience, 12, 3181–3190. Ruangkittisakul, A., Schwarzacher, S. W., Secchia, L., Ma, Y., Bobocea, N., Poon, B. Y., et al. (2008). Generation of eupnea and sighs by a spatiochemically organized inspiratory network. The Journal of Neuroscience, 28, 2447–2458.
Rubin, J. E., Hayes, J. A., Mendenhall, J. L., & Del Negro, C. A. (2009). Calcium-activated nonspecific cation current and synaptic depression promote networkdependent burst oscillations. Proceedings of the National Academy of Sciences of the United States of America, 106, 2939–2944. Rutherford, L. C., Nelson, S. B., & Turrigiano, G. G. (1998). BDNF has opposite effects on the quantal amplitude of pyramidal neuron and interneuron excitatory synapses. Neuron, 21, 521–530. Saideman, S. R., Blitz, D. M., & Nusbaum, M. P. (2007). Convergent motor patterns from divergent circuits. The Journal of Neuroscience, 27, 6664–6674. Schwarzacher, S. W., Pestean, A., Gunther, S., & Ballanyi, K. (2002). Serotonergic modulation of respiratory motoneurons and interneurons in brainstem slices of perinatal rats. Neuroscience, 115, 1247–1259. Schwindt, P., & Crill, W. (1999). Mechanisms underlying burst and regular spiking evoked by dendritic depolarization in layer 5 cortical pyramidal neurons. Journal of Neurophysiology, 81, 1341–1354. Sessle, B. J., Ball, G. J., & Lucier, G. E. (1981). Suppressive influences from periaqueductal gray and nucleus raphe magnus on respiration and related reflex activities and on solitary tract neurons, and effect of naloxone. Brain Research, 216, 145–161. Smith, J., Ellenberger, H., Ballanyi, K., Richter, D., & Feldman, J. (1991). Pre-Bötzinger complex: A brainstem region that may generate respiratory rhythm in mammals. Science, 254, 726–729. Stellwagen, D., & Malenka, R. C. (2006). Synaptic scaling mediated by glial TNF-alpha. Nature, 440, 1054–1059. St-John, W. M., & Paton, J. F. (2004). Role of pontile mechanisms in the neurogenesis of eupnea. Respiratory Physiology & Neurobiology, 143, 321–332. Stornetta, R. L., Rosin, D. L., Wang, H., Sevigny, C. P., Weston, M. C., & Guyenet, P. G. (2003). A group of glutamatergic interneurons expressing high levels of both neurokinin-1 receptors and somatostatin identifies the region of the pre-Botzinger complex. The Journal of Comparative Neurology, 455, 499–512. Stuart, G., & Sakmann, B. (1995). Amplification of EPSPs by axosomatic sodium channels in neocortical pyramidal neurons. Neuron, 15, 1065–1076. Subramanian, H. H., & Holstege, G. (2010). Periaqueductal gray control of breathing. Advances in Experimental Medicine and Biology, 669, 353–358. Tan, W., Janczewski, W., Yang, P., Shao, X., Callaway, E., & Feldman, J. (2008). Silencing pre-Bötzinger complex somatostatin-expressing neurons induces persistent apnea in awake rat. Nature Neuroscience, 11, 538–540. Tazerart, S., Viemari, J. C., Darbon, P., Vinay, L., & Brocard, F. (2007). Contribution of persistent sodium
49 current to locomotor pattern generation in neonatal rats. Journal of Neurophysiology, 98, 613–628. Telgkamp, P., & Ramirez, J. M. (1999). Differential responses of respiratory nuclei to anoxia in rhythmic brain stem slices of mice. Journal of Neurophysiology, 82, 2163–2170. Teppema, L. J., Veening, J. G., Kranenburg, A., Dahan, A., Berkenbosch, A., & Olievier, C. (1997). Expression of c-fos in the rat brainstem after exposure to hypoxia and to normoxic and hyperoxic hypercapnia. The Journal of Comparative Neurology, 388, 169–190. Thiagarajan, T. C., Piedras-Renteria, E. S., & Tsien, R. W. (2002). alpha- and betaCaMKII. Inverse regulation by neuronal activity and opposing effects on synaptic strength. Neuron, 36, 1103–1114. Thirumalai, V., & Marder, E. (2002). Colocalized neuropeptides activate a central pattern generator by acting on different circuit targets. The Journal of Neuroscience, 22, 1874–1882. Thoby-Brisson, M., Karlen, M., Wu, N., Charnay, P., Champagnat, J., & Fortin, G. (2009). Genetic identification of an embryonic parafacial oscillator coupling to the pre-Botzinger complex. Nature Neuroscience, 12, 1028–1035. Thoby-Brisson, M., & Ramirez, J. (2000). Role of inspiratory pacemaker neurons in mediating the hypoxic response of the respiratory network in vitro. The Journal of Neuroscience, 20, 5858–5866. Thoby-Brisson, M., & Simmers, J. (1998). Neuromodulatory inputs maintain expression of a lobster motor patterngenerating network in a modulation-dependent state: Evidence from long-term decentralization in vitro. The Journal of Neuroscience, 18, 2212–2225. Thoby-Brisson, M., Trinh, J. B., Champagnat, J., & Fortin, G. (2005). Emergence of the pre-Botzinger respiratory rhythm generator in the mouse embryo. The Journal of Neuroscience, 25, 4307–4318. Tryba, A. K., Pena, F., Lieske, S. P., Viemari, J. C., ThobyBrisson, M., & Ramirez, J. M. (2008). Differential modulation of neural network and pacemaker activity underlying eupnea and sigh-breathing activities. Journal of Neurophysiology, 99, 2114–2125. Tryba, A. K., & Ramirez, J. M. (2004). Background sodium current stabilizes bursting in respiratory pacemaker neurons. Journal of Neurobiology, 60, 481–489. Turner, D. L., & Mitchell, G. S. (1997). Long-term facilitation of ventilation following repeated hypoxic episodes in awake goats. The Journal of Physiology, 499(Pt 2), 543–550. Turrigiano, G. G. (2008). The self-tuning neuron: Synaptic scaling of excitatory synapses. Cell, 135, 422–435. Turrigiano, G. G., Leslie, K. R., Desai, N. S., Rutherford, L. C., & Nelson, S. B. (1998). Activity-dependent scaling of quantal amplitude in neocortical neurons. Nature, 391, 892–896. Van Drongelen, W., Koch, H., Elsen, F. P., Lee, H. C., Mrejeru, A., Doren, E., et al. (2006). Role of persistent
sodium current in bursting activity of mouse neocortical networks in vitro. Journal of Neurophysiology, 96, 2564–2577. Vanderhorst, V. G., & Ulfhake, B. (2006). The organization of the brainstem and spinal cord of the mouse: Relationships between monoaminergic, cholinergic, and spinal projection systems. Journal of Chemical Neuroanatomy, 31, 2–36. Viemari, J. C., Bevengut, M., Burnet, H., Coulon, P., Pequignot, J. M., Tiveron, M. C., et al. (2004). Phox2a gene, A6 neurons, and noradrenaline are essential for development of normal respiratory rhythm in mice. The Journal of Neuroscience, 24, 928–937. Viemari, J., & Hilaire, G. (2002). Noradrenergic receptors and in vitro respiratory rhythm: Possible interspecies differences between mouse and rat neonates. Neuroscience Letters, 324, 149–153. Viemari, J. C., Maussion, G., Bevengut, M., Burnet, H., Pequignot, J. M., Nepote, V., et al. (2005a). Ret deficiency in mice impairs the development of A5 and A6 neurons and the functional maturation of the respiratory rhythm. The European Journal of Neuroscience, 22, 2403–2412. Viemari, J. C., & Ramirez, J. M. (2006). Norepinephrine differentially modulates different types of respiratory pacemaker and nonpacemaker neurons. Journal of Neurophysiology, 95, 2070–2082. Viemari, J. C., Roux, J. C., Tryba, A. K., Saywell, V., Burnet, H., Pena, F., et al. (2005b). Mecp2 deficiency disrupts norepinephrine and respiratory systems in mice. The Journal of Neuroscience, 25, 11521–11530. Voituron, N., Zanella, S., Menuet, C., Dutschmann, M., & Hilaire, G. (2009). Early breathing defects after moderate hypoxia or hypercapnia in a mouse model of Rett syndrome. Respiratory Physiology & Neurobiology, 168, 109–118. Völker, A., Ballanyi, K., & Richter, D. (1995). Anoxic disturbance of the isolated respiratory network of neonatal rats. Experimental Brain Research, 103, 9–19. Von Euler, C., & Trippenbach, T. (1976). Excitability changes of the inspiratory “off-switch” mechanism tested by electrical stimulation in nucleus parabrachialis in the cat. Acta Physiologica Scandinavica, 97, 175–188. Von Leupoldt, A., Keil, A., Chan, P. Y., Bradley, M. M., Lang, P. J., & Davenport, P. W. (2010). Cortical sources of the respiratory-related evoked potential. Respiratory Physiology & Neurobiology, 170, 198–201. Wang, W., Fung, M. L., & St John, W. M. (1993). Pontile regulation of ventilatory activity in the adult rat. Journal of Applied Physiology, 74, 2801–2811. Wang, H., Stornetta, R. L., Rosin, D. L., & Guyenet, P. G. (2001). Neurokinin-1 receptor-immunoreactive neurons of the ventral respiratory group in the rat. The Journal of Comparative Neurology, 434, 128–146. Weese-Mayer, D. E., Lieske, S. P., Boothby, C. M., Kenny, A. S., Bennett, H. L., Silvestri, J. M., et al. (2006).
50 Autonomic nervous system dysregulation: Breathing and heart rate perturbation during wakefulness in young girls with Rett syndrome. Pediatric Research, 60, 443–449. West, J. B. (2010). American medical research expedition to everest. High Altitude Medicine & Biology, 11, 103–110. Wood, D. E., Stein, W., & Nusbaum, M. P. (2000). Projection neurons with shared cotransmitters elicit different motor patterns from the same neural circuit. The Journal of Neuroscience, 20, 8943–8953. Wulff, P., Ponomarenko, A. A., Bartos, M., Korotkova, T. M., Fuchs, E. C., Bahner, F., et al. (2009). Hippocampal theta rhythm and its coupling with gamma oscillations require fast inhibition onto parvalbumin-positive interneurons. Proceedings of the National Academy of Sciences of the United States of America, 106, 3561–3566. Yang, J. J., Chou, Y. C., Lin, M. T., & Chiu, T. H. (1997). Hypoxia-induced differential electrophysiological changes in rat locus coeruleus neurons. Life Sciences, 61, 1763–1773.
Yaron, M., Niermeyer, S., Lindgren, K. N., Honigman, B., Strain, J. D., & Cairns, C. B. (2003). Physiologic response to moderate altitude exposure among infants and young children. High Altitude Medicine & Biology, 4, 53–59. Zabka, A. G., Mitchell, G. S., & Behan, M. (2006). Conversion from testosterone to oestradiol is required to modulate respiratory long-term facilitation in male rats. The Journal of Physiology, 576, 903–912. Zanella, S., Roux, J. C., Viemari, J. C., & Hilaire, G. (2006). Possible modulation of the mouse respiratory rhythm generator by A1/C1 neurones. Respiratory Physiology & Neurobiology, 153, 126–138. Zhang, W., & Linden, D. J. (2003). The other side of the engram: Experience-driven changes in neuronal intrinsic excitability. Nature Reviews. Neuroscience, 4, 885–900. Zhou, Z., Champagnat, J., & Poon, C. S. (1997). Phasic and long-term depression in brainstem nucleus tractus solitarius neurons: Differing roles of AMPA receptor desensitization. The Journal of Neuroscience, 17, 5349–5356.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 4
Supraspinal control of locomotion: The mesencephalic locomotor region Didier Le Ray}, Laurent Juvin{,}, Dimitri Ryczko{ and Réjean Dubuc{,{,* {
{ Département de kinanthropologie, Université du Québec à Montréal, Montréal, Québec, Canada Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montréal, Québec, Canada } Laboratoire Mouvement Adaptation Cognition, Université de Bordeaux, CNRS, Bordeaux, France
Abstract: Locomotion is a basic motor function generated and controlled by genetically defined neuronal networks. The pattern of muscle synergies is generated in the spinal cord, whereas neural centers located above the spinal cord in the brainstem and the forebrain are essential for initiating and controlling locomotor movements. One such locomotor control center in the brainstem is the mesencephalic locomotor region (MLR), first discovered in cats and later found in all vertebrate species tested to date. Over the last years, we have investigated the cellular mechanisms by which this locomotor region operates in lampreys. The lamprey MLR is a well-circumscribed region located at the junction between the midbrain and hindbrain. Stimulation of the MLR induces locomotion with an intensity that increases with the stimulation strength. Glutamatergic and cholinergic monosynaptic inputs from the MLR are responsible for excitation of reticulospinal (RS) cells that in turn activate the spinal locomotor networks. The inputs are larger in the rostral than in the caudal hindbrain RS cells. MLR stimulation on one side elicits symmetrical excitatory inputs in RS cells on both sides, and this is linked to bilateral projections of the MLR to RS cells. In addition to its inputs to RS cells, the MLR activates a well-defined group of muscarinoceptive cells in the brainstem that feeds back strong excitation to RS cells in order to amplify the locomotor output. Finally, the MLR gates sensory inputs to the brainstem through a muscarinic mechanism. It appears therefore that the MLR not only controls locomotor activity but also filters sensory influx during locomotion. Keywords: locomotion; supraspinal control; mesencephalic locomotor region; sensory modulation; cholinergic transmission; lamprey.
*Corresponding author. Tel.: þ1-514-343-5729; Fax: þ1-514-343-6611 DOI: 10.1016/B978-0-444-53825-3.00009-7
51
52
Introduction The neural organization underlying locomotion— one of the most basic motor acts—is remarkably similar in different species of vertebrates. The muscle synergies responsible for propulsion are generated by neural networks in the spinal cord interacting with sensory signals (Grillner, 1981, 1985; Rossignol, 1996; Rossignol et al., 2006). These spinal networks, known as Central Pattern Generators (CPG) for locomotion, are activated and controlled by specific supraspinal structures, which also receive sensory inputs (Armstrong, 1986; Orlovsky et al., 1999; Rossignol, 1996; Rossignol et al., 2006; Shik and Orlovsky, 1976). The supraspinal control of locomotion includes forebrain structures, specific locomotor centers in the forebrain and brainstem, and command cells in the lower brainstem that activate the spinal CPGs (Fig. 1). The detailed contribution of forebrain structures to locomotion has not been resolved yet. However, the role of motor cortex is better known; it contributes to precision walking requiring an exact foot placement, such as on an uneven terrain (Beloozerova and Sirota, 1993a,b; Bretzner and Drew, 2005; reviewed in Drew et al., 2008). The basal ganglia are believed to play a role in the selection of locomotor behaviors (Grillner et al., 1997, 2008). A striking feature relative to the supraspinal control of locomotion is the presence of forebrain and brainstem locomotor centers specifically dedicated to initiating and controlling locomotion (for reviews, see: Armstrong, 1986; Dubuc, 2009; Dubuc et al., 2008; Grillner and Dubuc, 1988; Grillner et al., 1997; Jordan, 1986, 1998; Orlovsky et al., 1999; Whelan, 1996). One such region is located in the diencephalon and another in the mesencephalon. These two locomotor control centers are respectively referred to as the diencephalic locomotor region (DLR; El Manira et al., 1997; Grillner et al., 2008) and the mesencephalic locomotor region (MLR; Shik et al., 1966). This review chapter will focus primarily on the MLR and on recent findings obtained from one species of vertebrates, the lamprey.
The mesencephalic locomotor region in vertebrates The MLR has been identified in the 1960s by a research group in Moscow (Shik et al., 1966). This region receives inputs from the basal ganglia, the lateral hypothalamus, and the medial hypothalamus through the periaqueductal gray matter (Jordan, 1998). The mammalian MLR consists of two nuclei, the nucleus cuneiformis (CN) and the pedunculopontine nucleus (PPN). Electrical or chemical stimulation of the MLR induces bouts of locomotion (Garcia-Rill et al., 1985; Shik et al., 1966) via the activation of the reticulospinal (RS) pathways (Garcia-Rill and Skinner, 1987a,b; Orlovsky, 1970; Steeves and Jordan, 1984). The mechanisms that underlie the activation of the MLR when the basal ganglia are activated have not been fully understood. It appears that the selection of a relevant motor program by the basal ganglia would rely on disinhibition (Hikosaka, 1991). It was proposed that the ventral pallidum and the substantia nigra reticulata (SNr) could act similarly and inhibit the MLR (Grillner et al., 1998; Takakusaki, 2008). As such, locomotion would result from a disinhibition of the MLR by the ventral pallidum or the SNr. Experimental findings support this hypothesis. For instance, electrical stimulation of the SNr prevents MLR-inducted locomotion (Takakusaki, 2003), suggesting that the SNr inhibits locomotor activity. In addition, injections of the GABAA antagonist bicuculline in the MLR induce bouts of locomotion (Garcia-Rill et al., 1990). Classically, electrical or chemical stimulation of the MLR in decerebrate cats elicits motor output that has been subdivided into two phases. First, there is an increase in muscle tone allowing the animal to fully support its weight; this is followed by the locomotor phase (Shik et al., 1966). In the initial experiments describing the MLR of cats, it was found that the frequency of locomotion was graded in relation to the stimulation intensity. At low MLR stimulation intensities, the animals walked; as the intensity increased, they trotted
53
Motor cortex
Basal ganglia
DLR Medial and lateral hypothalamus
MLR
Pontine RF
Medullary RF
Fig. 1. The supraspinal control of locomotion. The general organization of the supraspinal control of locomotion has been described in mammals and some of the relevant structures and their connections are schematically illustrated on a sagittal view of the forebrain and brainstem. DLR, diencephalic locomotor region; MLR, mesencephalic locomotor region; RF, reticular formation.
and then galloped. This initial stunning observation provided the basis for qualifying this particular brainstem region as “dedicated to control a locomotor output.” The exact anatomical substrate of the MLR has been on the other hand subjected to debate. The most effective site to induce locomotion was a region comprising the CN and possibly a part of the PPN (Mori et al., 1989). Activation of the CN in a walking cat increased the speed of locomotion (Sterman and Fairshild, 1966). Similarly in rats, chemical
activation of the PPN elicited only brief episodes of locomotion in comparison to those elicited by activation of the CN (Garcia-Rill et al., 1985, 1990). The CN was thus proposed as the most effective site eliciting locomotion (for review, see Grillner et al., 1997). Whether a part or the entire PPN participated in the initiation of locomotion was not as clear; because of the proximity of these two structures, experimental results have been difficult to interpret. Several studies have tried to dissect the specific role of the CN and
54
the PPN in the control of locomotion. Because locomotion is also strongly modulated by its behavioral context, it was suggested that the MLR could be subdivided in different modules that would be activated in a context-dependent manner (Sinnamon, 1993). For instance, it was proposed that the MLR elicited locomotion in three different contexts and could therefore be subdivided into three main functional areas: “an exploratory system,” “an appetitive system,” and “a defensive system” (Sinnamon, 1993). This concept of organization was supported by experimental findings. For example, injection of glutamate in the CN induced a sequence of freezing, darting, and fast running (Mitchell et al., 1988a, 1988b). An “escape” locomotor behavior was observed both in cats and rats when the CN was stimulated (Depoortere et al., 1990a,b; Mori et al., 1989; Sirota and Shik, 1973). On the other hand, injections of GABA antagonist in the PPN induced locomotion that was apparently more related to startle (Ebert and Ostwald, 1991; Garcia-Rill et al., 1990). It was proposed that the PPN would itself be divided into two specific regions, a ventral and a dorsal component (Milner and Mogenson, 1988). The dorsal part of the PPN would be part of the MLR as a locomotor controlling region, whereas the ventral part would consist of a muscle tone inhibitory system (Takakusaki, 2003). Chemical activation of the dorsal part of the PPN was shown to increase locomotion in intact rats, while an opposite effect was observed when the ventral part of the PPN was activated (Milner and Mogenson, 1988). Functional magnetic resonance imaging recently revealed an increased BOLD signal in the MLR during mental imagery of walking and running in healthy volunteers (Jahn et al., 2008). In addition, recent clinical trials have shown improvement in posture and gait after stimulating the PPN in Parkinson’s patients (Lozano and Snyder, 2008; Mazzone et al., 2005; Stefani et al., 2007). Moreover, there is new evidence that cholinergic mesencephalic neurons would be involved in gait and postural disorders in Parkinson’s disease (Karachi et al., 2010).
The MLR does not project directly to the spinal cord, but it activates hindbrain RS cells that in turn activate spinal cord locomotor networks (Garcia-Rill, 1991; Garcia-Rill and Skinner, 1987a,b; Grillner, 1981; for review, see Rossignol, 1996). After its first discovery in cats (Shik et al., 1966), the MLR was also identified in many other species of vertebrates, including rats (Skinner and Garcia-Rill, 1984), stingrays (Bernau et al., 1991), guinea pigs (Marlinsky and Voitenko, 1991), lampreys (Sirota et al., 2000), salamanders (Cabelguen et al., 2003), and rabbits (Musienko et al., 2008). The most remarkable feature of the MLR was the locomotor activity that increased in speed as the stimulation of the MLR was increased. This was seen in all the animal species investigated. Moreover, it was shown in salamanders that MLR stimulation elicited the two modes of locomotor activity displayed by these animals, walking and swimming. At low MLR stimulation, the motor output elicited was characterized by limb movements associated with stepping. As the MLR stimulation was increased, the stepping movements increased in frequency. With further increases in stimulation strength, the limbs moved backward along the body and swimming movements were elicited (Cabelguen et al., 2003). Altogether, these observations indicate that the MLR is responsible for the initiation of swimming, walking, trotting, or galloping in different species of vertebrates. Moreover, observations made in salamanders indicate that this brainstem region can control two different modes of locomotion in the same animal.
The supraspinal control of locomotion in lampreys The lamprey model has provided first-hand information on the cellular mechanisms of vertebrate locomotion. The general organization of the lamprey nervous system is strikingly similar to that of mammals, but the presence of considerably fewer
55
neurons results in a reduced complexity that has been very useful to examine the cellular mechanisms underlying motor behavior. The lamprey model has paved the way for several important discoveries. One of these is the detailed characterization of a vertebrate CPG for locomotion (Buchanan and Grillner, 1987; for review, see Grillner et al., 1998). The brainstem mechanisms responsible for initiating and controlling locomotion have also been successfully uncovered in lampreys with an array of in vitro techniques, with the added benefit of including all relevant brain structures needed for locomotor control, and the ability to monitor the active locomotor behavior. One common feature of all vertebrate species is the crucial role played by RS cells in relaying MLR inputs to the spinal CPGs for locomotion. The RS cells receive peripheral and central inputs, integrate these signals, and generate a coherent descending motor command. In lampreys, RS cells have been described anatomically and physiologically. They constitute about 90% of the neurons projecting to the spinal cord (Bussières, 1994; Davis and McClellan, 1994a,b; Stefani et al., 2007). The lamprey RS cells are located in one mesencephalic reticular nucleus (MRN) and in three rhombencephalic reticular nuclei, the anterior (ARRN), the middle (MRRN), and the posterior (PRRN; Brodin et al., 1988; Davis and McClellan, 1994a,b; Nieuwenhuys, 1972, 1977; Swain et al., 1993). There are around 2500 RS cells and 85% of them reside in the MRRN and PRRN (Bussières, 1994). There are clear homologies with reticular nuclei in other vertebrate species, including mammals (Cruce and Newman, 1984). The ARRN and MRRN are located in the “lamprey pons” and contain large RS cells (Müller cells; Rovainen, 1967) that send their axons medially in the spinal cord. These two nuclei are similar to the superior and middle reticular nuclei in fish, which are respectively homologous to nuclei pontis oralis and caudalis of mammals (Cruce and Newman, 1984). The PRRN, located in the “lamprey medulla oblongata,” contains RS cells that send axons laterally,
similarly to the nucleus gigantocellularis in mammals (also discussed in Brocard and Dubuc, 2003). The axons of large RS neurons make synaptic contacts with several classes of spinal neurons and some of these are part of the locomotor CPG (Buchanan, 1982; Ohta and Grillner, 1989; Rovainen, 1974). The prevalent neurotransmitter is glutamate, although 5-HT and neuropeptides are also present in some RS cells (reviewed in Brodin et al., 1988). As in other species, the MLR was first characterized functionally in the lamprey by its ability to initiate and control locomotion when electrically stimulated (Sirota et al., 2000). The neuroanatomical substrate of the MLR has remained more elusive, but recent studies have provided new insights relative to this. Stimulation experiments demonstrated that the MLR of lampreys is located at the mesopontine border, close to the wall of the mesencephalic ventricle (Brocard and Dubuc, 2003; Brocard et al., 2005, 2010; Le Ray et al., 2003; Sirota et al., 2000; Smetana et al., 2010). The most efficient area for eliciting locomotion is a region containing a group of cholinergic cells close to a large RS cell, I1. This area corresponds to the caudal pole of the laterodorsal tegmental nucleus (LDT). Increasing the MLR stimulation intensity elicited faster and faster swimming movements of greater amplitude (Sirota et al., 2000; Fig. 2). The forebrain projections to the lamprey MLR have not been as extensively studied as in mammals. There are GABAergic inputs from the caudal portion of the medial pallium (Ménard et al., 2007), a region that could correspond to the amygdala of mammals. As shown in mammals, the lamprey MLR is under a tonic inhibition. Locomotion induced by glutamate ejection in the MLR is suppressed by co-ejection of GABA receptor agonist. Moreover, the GABA receptor antagonist, gabazine, elicits bouts of swimming when injected in the MLR of a semi-intact lamprey (Ménard et al., 2007). The forebrain connections to the MLR as well as their neurochemical identity have still not been fully identified.
56 (a1)
(b1)
Rostral Mid
high frequency. Second, the synaptic responses in RS cells, in a Ringer’s solution enriched in divalent cations, still followed a 25-Hz MLR stimulation (Brocard and Dubuc, 2003; Le Ray et al., 2003).
Caudal Stim
Glutamatergic and cholinergic MLR outputs
1.5 mA
(b2)
(a2) Rostral Mid Caudal Stim 2.0 mA
(a3)
(b3)
Rostral Mid Caudal Stim 2.5 mA 2s
Fig. 2. The velocity of swimming movements is correlated to the intensity of MLR stimulation. EMG recordings of rostral, middle, and caudal body segments during MLR-induced swimming at different intensities of stimulation (a) with sketches of the body swimming movements (b). The MLR was electrically stimulated with trains of stimuli of 1 ms duration (10 Hz pulses) at 1.5 mA (a1, b1), 2 mA (a2, b2), 2.5 mA (a3, b3) (adapted from Sirota et al., 2000).
Downstream effects of the mesencephalic locomotor region Inputs to the MLR in lampreys are just starting to be explored, but the output projections from the MLR have been defined more extensively. The MLR projections are relayed within the hindbrain reticular formation. Several lines of evidence indicate that the MLR projections to RS neurons are monosynaptic. First, electrical stimulation of the MLR evokes short-latency EPSPs that are maintained during repetitive MLR stimulation at
The MLR of mammals contains populations of cholinergic neurons (see Jordan, 1998). It is also the case in lampreys where cholinergic neurons have been found within the isthmic region (Pombal et al., 2001). The possibility that the cholinergic neurons would be located within the MLR was addressed directly using immunohistochemical staining for choline acetyltransferase (ChAT) at the site of electrical stimulation that was found to initiate and control locomotion in lampreys (Le Ray et al., 2003). Two distinct groups of ChAT-immunoreactive neurons were observed in an area that included the isthmus and the caudal mesencephalon. One group consisted of densely clustered cells located medially close to the ventricular border. Another group of cholinergic cells was more loosely distributed further rostrally and laterally within the tegmentum. Comparison with other species (cats: Mitani et al., 1988; rats: Jones, 1990; and amphibians: Marín et al., 1997) suggested that the first group would correspond to the laterodorsal tegmental nucleus (LDT), whereas the second one to the PPN. In mammals (Garcia-Rill and Skinner, 1987a; Lai et al., 1999; Mesulam et al., 1983; Skinner et al., 1990), cholinergic neurons in these nuclei project to the reticular formation. This suggests a role for acetylcholine (ACh) in the MLR control of locomotion. However, numerous neurons projecting to RS cells are located in regions of the lamprey MLR that do not contain ChAT-positive neurons (Brocard et al., 2010). This implies that other neurotransmitter systems may also be involved (see below). Electrophysiological experiments are also in accord with a role of cholinergic transmission in mediating MLR effects onto RS cells. In mammals and birds, local
57
injections of cholinergic agonists in the reticular formation elicit locomotion (Garcia-Rill and Skinner, 1987a; Sholomenko et al., 1991). In these animal species, however, a direct link between a cholinergic command originating from the MLR and locomotion was not established. The role of cholinergic inputs was examined in semi-intact preparations of lampreys. The MLR was stimulated at 5 Hz, and the evoked EPSPs in RS cells summed up until the threshold for spiking was reached and swimming activity was elicited (see also Sirota et al., 2000). In the presence of D-tubocurarine the nicotinic antagonist (30–50 mM), applied selectively to the brainstem and not to the spinal cord using a partitioned recording chamber, the membrane potential of RS neurons remained mostly below the spiking threshold and swimming was prevented even under MLR stimulation intensity that would have normally induced swimming. In fact, swimming could only be induced by significantly increasing the MLR stimulation intensity. Conversely, the application of the selective cholinesterase inhibitor physostigmine (100 mM) to the “brainstem” chamber partition largely enhanced the MLR-evoked compound EPSPs and facilitated the RS cell depolarization and the occurrence of swimming (Le Ray et al., 2003). The amplitude and slope of the MLR-evoked monosynaptic EPSP were largely, but not completely blocked by the nicotinic antagonists D-tubocurarine (30–100 mM) or a-bungarotoxin (0.1 mM). Adding a mixture of NMDA and non-NMDA glutamate receptor antagonists (200 mM AP5 and 25 mM CNQX, respectively) further reduced the EPSPs. Taken together, these results suggested that MLR inputs to RS cells use cholinergic and glutamatergic transmission. Additional support for this was the observation that a direct application of nicotinic receptor agonists in either the MRRN or the PRRN evoked swimming in a semi-intact preparation. Comparable results were obtained in an in vitro isolated brainstem-spinal cord preparation in which fictive locomotion was induced by nicotinic receptor agonists. In addition, when
applied on preparation already displaying a slow fictive locomotor activity under NMDA perfusion, cholinergic agonists speeded up the locomotor rhythm early after their application. This effect did not occur when the fictive locomotor rhythm had already stabilized to its faster rhythm under NMDA perfusion or after a previous ejection of the cholinergic agonist had already accelerated the fictive locomotor rhythm. TTX-resistant depolarizing responses were generated in intracellularly recorded RS neurons by local application of nicotinic agonists onto the recorded cell. When repeated before the membrane potential of the RS neuron returned to resting value, the nicotinic responses showed summation properties allowing the neuron to reach the spiking threshold and to generate a sustained firing of action potentials. Such temporal summation of responses may be important for the slow buildup of RS cell depolarization and the delayed swimming onset that occur at low intensities of MLR stimulation. Indeed, in lampreys (Sirota et al., 2000) as in mammals (Garcia-Rill and Skinner, 1987b; Iwakiri et al., 1995), several seconds of repeated MLR stimulation are required to induce locomotion. Furthermore, the onset delay of swimming decreases as the intensity or frequency of stimulation increases, probably due to the enhancement of RS depolarization by the nicotinic response summation. According to this, the buildup of the response to a 5 Hz stimulation of the MLR was dramatically reduced in the presence of a nicotinic antagonist and largely increased in the presence of the cholinesterase inhibitor physostigmine, which resulted in the blockade or the facilitation of the initiation of swimming activity, respectively (Le Ray et al., 2003).
Differential MLR inputs to RS cells in the MRRN and PRRN Although it is well established that the MLR elicits locomotor behavior by activating RS cells in different species of vertebrates, the detailed
58
connectivity between the MLR and the RS cells has remained unresolved. Such connectivity was examined in lampreys by comparing the recruitment of large MRRN and PRRN RS cells using paired intracellular recordings and increasing stimulation strength of the MLR (Brocard and Dubuc, 2003); such an approach was not used in any other vertebrate species previously. MRRN cells displayed spiking activity at low MLR stimulation strength, whereas PRRN cells began to discharge at higher intensities, when MRRN cells have already reached their maximal spiking frequency. The respective contribution of the MRRN and PRRN to locomotor control was also investigated by selectively injecting a mixture of ionotropic glutamate receptor antagonists in each of the two reticular nuclei (Brocard and Dubuc, 2003). Injections over the entire MRRN prevented locomotion, even during MLR stimulation at high intensities. Injections over the PRRN only decreased locomotion intensity. According to these observations, RS cells receive differential inputs from the MLR, such that RS cells in the rostral hindbrain discharge more importantly at low swimming intensities and RS cells located more caudally begin spiking at higher swimming speeds.
Bilateral MLR inputs to RS cells Another interesting aspect of the MLR is that it elicits bilaterally symmetrical locomotion even when it is stimulated only on one side (Shik and Orlovsky, 1976; Shik et al., 1966; Sirota et al., 2000). This has been a feature observed in the different animal species investigated to date. A unilateral electrical (E) or chemical (C) activation of the MLR produces symmetrical locomotion in the guinea pig (C: Marlinsky and Voitenko, 1991), lamprey (E and C: Sirota et al., 2000), rabbit (E: Musienko et al., 2008), rat (E: Skinner and Garcia-Rill, 1984), salamander (E: Cabelguen et al., 2003), and stingray (E: Bernau et al., 1991; for review, see Orlovsky et al., 1999). One question relates to how a unilateral stimulation of the MLR
is converted into a bilateral symmetrical locomotor output At which neural level (or levels) does the symmetry appear? We recently addressed these questions in a study combining anatomical, electrophysiological, Ca2þ imaging, and kinematic analysis in lampreys. We found that MLR inputs to RS cells are at least partly responsible for the transformation of a unilateral MLR stimulation into bilaterally symmetrical locomotor output (Brocard et al., 2010). Direct evidence for the symmetry of MLR inputs to RS cells was provided by simultaneously recording the intracellular responses of bilateral pairs of identifiable homologous RS cells from the MRRN and the PRRN to stimulation of the MLR on one side (Brocard et al., 2010). The synaptic responses on both sides were very similar in shape, amplitude, and threshold intensity. Increasing the intensity of MLR stimulation produced a strikingly similar increase in the magnitude of the responses on both sides (Fig. 3). Because the technique of intracellular recordings limits our conclusions to a small number of large-size-paired RS cells, Ca2þ imaging experiments were performed on brainstem-isolated preparations. In accord to what was found in the large cells, a bilaterally symmetrical activation of smaller-sized RS cells of the MRRN and PRRN was seen when unilaterally stimulating the MLR. Monosynaptic inputs from the MLR to RS neurons were known to be present in lampreys (Brocard and Dubuc, 2003). In Brocard et al. (2010), it was shown that MLR projects monosynaptically to RS cells not only on the ipsilateral but also on the contralateral side. In a high-divalent cation solution, the synaptic responses of simultaneously recorded homologous RS cells persisted and exhibited a constant latency during high-frequency stimulation. Moreover, during gradual replacement of normal Ringer’s solution with a Ca2þ-free solution, the magnitude of responses showed a gradual reduction with a similar time course in the homologous RS cells. These results provided strong evidence for monosynaptic inputs from the MLR to RS cells on both sides.
59 MRRN (a2)
(a1) RS ipsi
Stim MLR
RS contra
RS ipsi
RS contra
2 mV
2 mV
2 mV
2 mV
20 ms
20 ms
20 ms
20 ms
(b3)
(a3) 1200
700 600
1000 Response area (mV ms)
Response area (mV ms)
PRRN (b2)
(b1)
800 600 400 RS ipsi RS contra
200 0
500 400 300 200
RS ipsi RS contra
100 0
4
6
8
10
12
Stimulation intensity (mA)
1
5 9 12 Stimulation intensity (mA)
16
Fig. 3. Postsynaptic potentials recorded in homologous RS cells on both sides in response to unilateral stimulation of the MLR. (a) Data from ipsilateral (ipsi) and contralateral (contra) reticulospinal (RS) cells from the middle rhombencephalic reticular nucleus (MRRN). (b) Data from ipsi and contra RS cells from the posterior rhombencephalic reticular nucleus (PRRN). (a1, a2, b1, b2) Graded responses recorded simultaneously from bilaterally homologous RS cells to increasing intensity of stimulation of the MLR on one side. All traces are averages of three sweeps. (a3, b3) Relationship between the intensity of MLR stimulation and the area of postsynaptic potentials elicited in ipsilateral (black squares) and contralateral (blue circles) RS cells. Data from RS cells in the MRRN and in the PRRN are from different preparations (adapted from Brocard et al., 2010).
Simultaneous recordings of homologous RS cells of the MRRN on both sides also revealed a symmetrical output in frequency when bouts of symmetrical swimming are generated by unilateral stimulation of the MLR in a semi-intact preparation (Fig. 4; Brocard et al., 2010). When increasing the MLR stimulation, the increase in discharge frequency was identical for the left and right RS cells, and the swimming frequency proportionally increased (Brocard et al., 2010). Anatomical experiments were carried out to examine the MLR projections to RS cells. A
unilateral injection of a retrograde tracer into the MRRN revealed labeled cells bilaterally in the MLR (Brocard et al., 2010; see also Sirota et al., 2000). Bilateral injections of two different retrograde tracers in the MRRN revealed that the same MLR cells very rarely projected to the MRRN on both sides. The anatomical projections were bilaterally asymmetrical: retrograde markers injected in the MRRN on one side always revealed fewer labeled cells in the contralateral MLR, indicating that the descending projections from the MLR to RS neurons were slightly biased
60 (a)
(b)
4.5 mA
RS ipsi
RS contra
40 mV 40 mV
1s
5s
Stim MLR
120
RS discharge frequency (% of max)
(d) RS discharge frequency (% of max)
(c) RS ipsi RS contra
80 40 0 1.5
2.5 3.5 Stim MLR (mA)
4.5
120
RS ipsi RS contra
80 40 0 20
40
60 80 100 Stim MLR (% of max)
Fig. 4. Homologous reticulospinal (RS) cells display a symmetrical output in frequency when swimming is generated by a unilateral stimulation of the MLR. (a) Simultaneous paired intracellular recordings from homologous ipsilateral (ipsi; black) and contralateral (contra; blue) RS cells of the middle rhombencephalic reticular nucleus (MRRN) during unilateral MLR electrical stimulation at 4.5 mA in a semi-intact preparation of larval lamprey. The duration of the electrical stimulation applied to the MLR is indicated by the stimulation bar (Stim MLR) below the recordings. Note the enlargement illustrating the antiphasic relationship between ipsilateral and contralateral RS neurons. (b) Kinematic analysis (15 frames/s) of one representative swimming cycle elicited by unilateral MLR stimulation (4.5 mA). Tracking positions of markers equidistantly distributed along the body of the animal using software analysis revealed that left and right maximal bending angles of the body are not statistically different, indicating the bilateral symmetry of swimming movements in response to unilateral stimulation of the MLR. (c) Discharge frequencies of homologous ipsilateral (black squares) and contralateral (blue circles) RS neurons in the same animal. Each dot illustrates the mean S.E.M. discharge frequency for a 20 s bout of MLR stimulation. Each intensity was tested three times. RS discharge frequencies are expressed in percentage of the maximal RS discharge frequency. (d) Relationships between the intensity of unilateral stimulation of the MLR and the discharge frequencies of ipsilateral (black squares) and contralateral (blue circles) RS neurons in five animals. Both ipsilateral and contralateral data followed a highly similar cubic polynomial function (black and blue solid lines, respectively). The dotted lines illustrate the prediction intervals for each fit at 95%. Data in a, b, c are from the same animal (adapted from Brocard et al., 2010).
ipsilaterally (Brocard et al., 2010). Because the MLR inputs to RS cells are perfectly symmetrical and the swimming behavior elicited upon a unilateral stimulation of the MLR, the anatomical asymmetry must then be physiologically compensated.
Bilateral descending projections from the MLR to RS neurons have also been described anatomically in the cat (Garcia-Rill et al., 1983; Steeves and Jordan, 1984) and rat (Garcia-Rill et al., 1986). Electrophysiological experiments in cats revealed bilateral inputs (Garcia-Rill and
61
Skinner, 1987b; Orlovsky, 1970), whereas the bilateral anatomical projections were found largely asymmetrical (Steeves and Jordan, 1984). Injections of [3H]proline and [3H]leucine into the MLR revealed descending neurons from the MLR that were located mainly on the ipsilateral side. In line with this anatomical observation, Garcia-Rill and Skinner (1987b) provided electrophysiological evidence in the cat that the MLR projected mainly to ipsilateral RS cells, that in turn projected to the ipsilateral spinal cord. Noga et al. (2003) showed that the pattern of postsynaptic responses measured intracellularly in a-motoneurons (innervating flexor, extensor, or bifunctional muscles) in the L6–L7 spinal cord segments was similar whether the ipsilateral or the contralateral MLR was stimulated. The segmental latency of the first locomotor EPSP detected in motoneuronal response was also similar on both sides. They proposed that the slight asymmetry of the descending signal generated by unilateral MLR stimulation could be compensated by descending RS neurons projecting contralaterally directly or indirectly through spinal commissural neurons, to end up with a symmetrical motor output (Noga et al., 2003). As indicated above, a similar anatomical bilateral asymmetry in MLR-RS projections was observed in lampreys, although physiological experiments revealed a striking symmetry in the MLR inputs to RS cells. Whether this is the case in mammals remains to be established. The physiological connections between the MLR in RS cells on both sides have not been examined directly. In an attempt to do so, Orlovsky used intracellular recordings of RS cells and stimulated the MLR on each side with two different stimulating electrodes (Orlovsky, 1970). However, small differences in the positioning of the stimulating electrodes and/or differences in the impedance of the two stimulating electrodes could account for the differences that he reported in the size of synaptic responses of RS cells on both sides. It appears therefore that unilateral activation of the MLR produces a bilateral symmetrical
locomotor output. The bilateral projections from the MLR to RS cells are likely to play a crucial role in the symmetrical locomotor activity. Further experiments are needed to establish the contribution of bilateral connections in the spinal cord to this symmetry as well as the possible contribution of sensory inputs. Moreover, it is not known whether crossing connections innervating the RS cells themselves in the hindbrain could also strengthen the bilateral symmetry observed during locomotion induced by a unilateral MLR stimulation. These issues should be examined further.
A locomotor-boosting mechanism within the brainstem The MLR has traditionally been described as a neural control area in the brainstem, which activates specific populations of reticulospinal neurons to initiate and control locomotion. As such, the MLR has been considered as part of a serial control system for locomotion (forebrain structures ! MLR ! RS cells ! spinal CPGs) Recently, we described in lampreys a parallel projection from the MLR to a group of hindbrain neurons that, in turn, provide additional excitation to reticulospinal cells to amplify the locomotor output. These interesting findings were made as we were examining the role of muscarinic receptor activation of RS cells (Smetana et al., 2007). We found that lamprey RS cells were activated by bath application of a cholinergic muscarinic agonist. Muscarine elicited sustained membrane depolarizations ( 5 s duration) in RS cells that recurred at a periodicity of 60 s. Such depolarizations occurred simultaneously in pairs of homologous RS cells recorded intracellularly, and calcium imaging showed that entire populations of RS cells were activated synchronously. In addition, the sustained depolarizations were associated with ventral discharges, suggesting that they could somehow participate in locomotion.
62
The effects of muscarine disappeared when TTX was added to the perfusion Ringer’s, suggesting that muscarine was not acting directly onto RS cells but most likely through neurons that were presynaptic to the RS cells. Indeed, the depolarizations resulted from the activation a group of hindbrain muscarinoceptive interneurons that projected to RS neurons. Lesion studies as well as local muscarine injections revealed that the population of muscarinoceptive cells was located between the caudal border of the MRRN and rostral border of the PRRN. Bilateral injections of atropine in this region prevented the depolarization of RS neurons when muscarine was bath-applied. There were cells in this region showing immunoreactivity for muscarinic receptors. In addition, calcium imaging experiments revealed that cells in this region displayed sustained rises in intracellular calcium to bath application of muscarine. These calcium responses persisted in the presence of TTX (Smetana et al., 2010). It was also shown that the muscarinoceptive cells provided strong bilateral glutamatergic inputs to the RS cells. Paired recordings of RS neurons and muscarinoceptive interneurons showed that depolarizations of the muscarinoceptive cell induced short-latency EPSCs in the RS cell, suggesting that the connections were monosynaptic (Smetana et al., 2010). The physiological significance of these connections was then determined. First, it was found that these muscarinoceptive cells received atropine-sensitive inputs from the MLR, and direct projections were anatomically suggested by the observation of cells labeled in the MLR on both sides after biocytin injection into the muscarinoceptive cell region. These results confirmed that the MLR not only projected to RS neurons but also to a group of muscarinoceptive cells that in turn sent excitation to RS neurons. This parallel pathway amplifies significantly the synaptic input received by the RS cells during the activation of the MLR. The specific role was confirmed using a semi-intact swimming preparation. Bilateral injections of atropine in the muscarinoceptive cell area considerably depressed
the depolarization responses of RS cells to MLR stimulation and modified the locomotor output (Fig. 5). As a result, the slope of the linear correlation between swimming frequency and MLR stimulation intensity was dramatically reduced by the bilateral injection of atropine in the muscarinoceptive neurons area (Smetana et al., 2010). There is a strikingly linear relationship between the MLR stimulation intensity and locomotion frequency in many animal species. We have now found in lampreys that when the MLR is intensively activated, a population of muscarinoceptive cells located in the hindbrain is recruited to literally “boost” the locomotor output. Whether this mechanism is present in other vertebrate species remains to be determined. In rats, the ejection of cholinergic agonists (carbachol, methacholine, and arecoline) in the medioventral medulla induces locomotion that is prevented by atropine, thus suggesting a role for muscarinic receptors (Kinjo et al., 1990). The neural substrate is however still unknown in mammals.
Gating of sensory inputs on RS neurons by the MLR During locomotion, sensory inputs shape the activity of the central neural networks according to external and internal constraints. In turn, the central networks gate sensory influx (Graham Brown, 1911; Grillner, 1973; Grillner and Rossignol, 1978; Rossignol and Gauthier, 1980). Gating has been described extensively in the mammalian spinal cord (Hultborn, 2001; Krawitz et al., 2001 Rossignol et al., 1981), but the underlying cellular mechanisms have not been identified yet. In lampreys, we found that muscarinic receptor activation depressed sensory inputs to RS cells (Le Ray et al., 2004). The efficacy of the sensorymotor connection between trigeminal afferents and RS neurons was then tested in the context of a MLR-induced locomotor behavior (Le Ray et al., 2010). For this purpose, intracellular
63 EMG
(a1)
(a3)
Atropine stim. MLR
5s
(a2)
Control
RS cell EMG
Atropine
Atropine Washout
1.2
0.8
0.4
0 2
stim.
Boost mechanism
20 mV
Swimming frequency (Hz)
Control 1.6
4
6
8
10
Stimulation frequency (Hz)
Fig. 5. Inactivation of muscarinoceptive neurons slows down swimming movement velocity. (a1) Schematic representation of the experimental set-up. Swimming is induced in a semi-intact preparation by the stimulation of the MLR. The activity of RS neurons is recorded intracellularly and EMG is used to assess the swimming activity. After ejection of atropine over the muscarinoceptive neurons, the swimming frequency and the oscillation of RS neuron membrane potentials are slowed down in comparison to control (compare a2 left and right). (a3) Graphic representation of the swimming frequency as a function of MLR stimulation intensity. The ejection of atropine over the muscarinoceptive neurons prevents the production of fast swimming even when the MLR is stimulated at high intensity (adapted from Smetana et al., 2010).
recordings were made from RS cells in isolated brainstem preparations while the trigeminal sensory root on one side was electrically stimulated at a low frequency (0.1 Hz). The trigeminal EPSP was monitored before and after a 3 Hz train of stimuli was applied to the MLR for 15 s. Consecutively to MLR tonic activation, a powerful depression of the trigeminal EPSP was observed in RS cells (Fig. 6). On average, both the peak and amplitude of the synaptic responses were significantly decreased and showed a progressive recovery after about 30 min (Le Ray et al., 2010). It was found that the depression of the trigeminalevoked EPSPs also depended on the level of MLR activation (Le Ray et al., 2010). Different frequencies of stimulation were randomly applied to the MLR in order to reproduce different levels of locomotor activity in the isolated brainstem. We found that the higher the MLR stimulation frequency, the stronger was the trigeminal EPSP depression until a maximal depression was obtained at 7–8 Hz. Increasing the stimulation frequency further did not produce more depression.
Interestingly, the sensory depression was maximal for MLR stimulation frequencies that were submaximal for swimming speeds (see Sirota et al., 2000). The time course of the depressive effects produced by the MLR activation suggested the involvement of metabotropic mechanisms, and because the MLR contains cholinergic neurons, the implication of muscarinic receptors was tested. In the presence of the muscarinic receptor antagonist, atropine, MLR stimulation produced a far less depression of trigeminal EPSPs in RS cells. There was still a residual depression, suggesting that other neurotransmitter systems can be involved. Altogether, these results indicate that the MLR-induced depression of the RS neuron response to trigeminal nerve stimulation is largely mediated by muscarinic receptors. This is supported by the immunohistochemical demonstration of the presence of muscarinic receptors on RS cells (Le Ray et al., 2010). Immunohistochemical labeling was also seen on cells located in the trigeminal descending tract (Northcutt, 1979), which was found to contain the second-order
64 (a)
Normal Ringer’s
(b)
(1) Trig. St
MLR St. 2 mV 10 ms
200 EPSP area (mV ms)
MLR St.
(2)
0
400
Atropine
(2)
(1)
MLR St. (1)
RS cell
(2)
(1)
400
200
0
10
2 mV 10 ms
(2)
30
50 Time (min)
70
90
110
Fig. 6. Synaptic transmission of trigeminal sensory nerve inputs to RS neurons is depressed by MLR stimulation. (a) Schematic representation of the experimental design. Stimulation (St.) electrodes are positioned within the MLR and on the trigeminal nerve, and the activity of a RS neuron is recorded using an intracellular electrode. (b, top) The area of EPSP recorded from RS neurons in response to stimulation of the trigeminal nerve is depressed for several minutes by the stimulation of the MLR (arrow). To compare, see the difference between superimposed EPSP in control (1) and post-MLR-stimulation (2) conditions. (b, bottom) Perfusion of atropine (10 mM) prevented the EPSP depression (adapted from Le Ray et al., 2010).
trigeminal sensory neurons that relay directly trigeminal inputs to RS neurons (Viana Di Prisco et al., 2005). Local pressure application of ACh or its muscarinic agonist pilocarpine onto the recorded RS neuron or in the trigeminal relay area reproduced the depressive effects of the MLR activation onto the trigeminal EPSP, without affecting the membrane potential of the recorded RS cell (Le Ray et al., 2004). Conversely, EPSP enhancement was observed when atropine (or scopolamine, another muscarinic antagonist) was substituted for the muscarinic agonists in the ejection pipette or when it was bath-applied. This suggests the existence of a tonic inhibitory control exerted via muscarinic receptors onto the trigemino-reticular connection. In addition, preincubation with atropine totally prevented the depressive effects of the muscarinic agonists. Several lines of evidence also indicate
that the muscarinic modulation is predominantly exerted on the NMDA receptor-mediated component of the glutamatergic EPSP elicited by trigeminal stimulation, without affecting the glycinergic one (see Viana Di Prisco et al., 1995): the depolarizing responses to direct application of NMDA onto the recorded RS cells were enhanced by atropine, whereas the responses to AMPA application were not. Moreover, blocking NMDA receptors with AP5 abolished the effects of muscarinic agonists and antagonists on the trigeminalevoked EPSP; muscarinic drug applications usually had little effect on the early part of the synaptic responses. In contrast to most of the cases reported in the literature (Bellingham and Berger, 1996; Jiang and Dun, 1986; Scanziani et al., 1995; Smolders et al., 1997), the muscarinic modulation of the RS responses to glutamate relies mainly on postsynaptic mechanisms in lampreys (i.e., at the
65
level of the RS neuron itself). Whether this is also the case within the trigeminal relay area where muscarinic modulation also clearly occurs will need to be examined. Strong trigeminal stimulation induces sustained depolarizations in RS neurons, which trigger swimming activity in the spinal cord (Viana Di Prisco et al., 1997, 2000). There is evidence that a tonic muscarinic inhibition of this sustained activity occurs (Le Ray et al., 2004): the duration of the depolarizing plateaus displayed a fivefold increase when atropine was perfused on the brainstem; the threshold for inducing a sustained depolarization was reduced by half; the firing rate during maximal responses was increased more than twice. Because the RS neuron input resistance, resting potential, and afterhyperpolarization were not affected by atropine perfusion, the enhancement of the sustained depolarizations likely resulted from a specific action on the synaptic response to trigeminal excitatory inputs. Interestingly, when depolarizations were evoked by local ejection of NMDA onto the recorded RS neuron, atropine unmasked membrane potential oscillations that occurred on top of the depolarizing plateau, and spiking occurred on top of each oscillation during the whole duration of the NMDA-evoked oscillatory behavior (Le Ray et al., 2004). The persistence of these oscillations under TTX suggested an intrinsic nature for the NMDA-induced activity. In the lamprey RS neurons, sustained depolarizations require the activation of NMDA receptor by trigeminal inputs (Viana Di Prisco et al., 2000), and a 30-Hz electrical stimulation of a trigeminal nerve could also trigger membrane potential oscillations in RS cells in the presence of atropine (Le Ray et al., 2004). Experiments performed on the isolated hindbrain demonstrated that RS oscillations could also occur in the absence of atropine and were blocked by the local ejection of a muscarinic agonist onto the recorded RS neuron (Le Ray et al., 2004). Because the spinal cord was removed, the NMDA-induced oscillations could not result from
any ascending spinal inputs (Dubuc and Grillner, 1989; Vinay and Grillner, 1993; Vinay et al., 1998a,b). Conclusions The control of locomotion relies in a large part on neural regions within the brainstem and forebrain specifically dedicated to locomotion. The MLR is one such region at the mesopontine border identified in all vertebrate species examined to date. It is believed to channel brainstem and forebrain inputs to then activate populations of RS cells in order to initiate and control locomotion. Examining the detailed downstream effects of the MLR in lampreys has yielded new information relative to the role of this region (Fig. 7). Inputs from the MLR to RS cells are strikingly symmetrical on both sides and this will play a significant role in generating symmetrical locomotion. The MLR does not only project to RS cells but also sends powerful inputs to a population of muscarinoceptive cells that provide additional excitation of RS cells, considerably amplifying locomotor output. In addition to controlling locomotor output, the MLR has now been shown to gate sensory inputs in the brainstem. The detailed mechanisms by which this sensory modulation operates have not been fully identified, but muscarinic receptors are involved. Future research will be needed to elucidate the detailed mechanisms involved as well as the functional importance of such sensory modulation. Acknowledgments We are very grateful to Danielle Veilleux for her help with the chapter, to Christian Valiquette for his skilful programming of data analysis software, and to Frédéric Bernard for the design of the figures. We thank François Auclair for his useful comments on the chapter. R. D. receives funding from the Canadian Institutes of Health Research (CIHR; individual and group grants), from the
66
MLR GLUergic AChergic
RS MRRN
MLR
SENSORY RELAY RS
d.V
PRRN MUSCARINOCEPTIVE
CPG
Fig. 7. Schematic representation of the proposed brainstem locomotor control circuitry (based on the results reviewed in this chapter). A first series of experiments established the contribution of bilateral glutamatergic and cholinergic inputs from the MLR to RS neurons. Further experiments brought to light a group of muscarinoceptive cells located at the pontomedullary border that receives direct input from the MLR and increases RS cell activity and locomotor output. Finally, the MLR modulates sensory transmission in the brainstem by likely acting at the level of both RS neurons and trigeminal sensory relay cells.
Natural Sciences and Engineering Research Council of Canada (NSERC), and from the Fonds de la Recherche en Santé du Québec (FRSQ; group grant). References Armstrong, D. (1986). Supraspinal contributions to the initiation and control of locomotion. Progress in Neurobiology, 26, 273–361.
Bellingham, M. C., & Berger, A. J. (1996). Presynaptic depression of excitatory synaptic inputs to rat hypoglossal motoneurons by muscarinic M2 receptors. Journal of Neurophysiology, 76, 3758–3770. Beloozerova, I. N., & Sirota, M. G. (1993a). The role of the motor cortex in the control of accuracy of locomotor movements in the cat. Journal de Physiologie, 461, 1–25. Beloozerova, I. N., & Sirota, M. G. (1993b). The role of the motor cortex in the control of vigour of locomotor movements in the cat. Journal de Physiologie, 461, 27–46. Bernau, N. A., Puzdrowski, R. L., & Leonard, R. B. (1991). Identification of the midbrain locomotor region and its
67 relation to descending locomotor pathways in the Atlantic stingray, Dasyatis sabina. Brain Research, 557, 83–94. Bretzner, F., & Drew, T. (2005). Contribution of the motor cortex to the structure and the timing of hindlimb locomotion in the cat: A microstimulation study. Journal of Neurophysiology, 94, 657–672. Brocard, F., Bardy, C., & Dubuc, R. (2005). Modulatory effect of substance P to the brainstem locomotor command in lampreys. Journal of Neurophysiology, 93, 2127–2141. Brocard, F., & Dubuc, R. (2003). Differential contribution of reticulospinal cells to the control of locomotion induced by the mesencephalic locomotor region. Journal of Neurophysiology, 90, 1714–1727. Brocard, F., Ryczko, D., Fenelon, K., Hatem, R., Gonzales, D., Auclair, F., et al. (2010). The transformation of a unilateral locomotor command into a symmetrical bilateral activation in the brainstem. The Journal of Neuroscience, 30, 523–533. Brodin, L., Grillner, S., Dubuc, R., Ohta, Y., Kasicki, S., & Hökfelt, T. (1988). Reticulospinal neurons in lamprey: Transmitters, synaptic interactions and their role during locomotion. Archives Italiennes de Biologie, 126, 317–345. Buchanan, J. T. (1982). Identification of interneurons with contralateral, caudal axons in the lamprey spinal cord: Synaptic interactions and morphology. Journal of Neurophysiology, 47, 961–975. Buchanan, J. T., & Grillner, S. (1987). Newly identified ‘glutamate interneurons’ and their role in locomotion in the lamprey spinal cord. Science, 236, 312–314. Bussières, N. (1994). Les systèmes descendants chez la lamproie. Étude anatomique et fonctionnelle. (PhD thesis) Université de Montréal, Montréal, Canada. Cabelguen, J. M., Bourcier-Lucas, C., & Dubuc, R. (2003). Bimodal locomotion elicited by electrical stimulation of the midbrain in the salamander Notophthalmus viridescens. The Journal of Neuroscience, 23, 2434–2439. Cruce, W. L. R., & Newman, D. B. (1984). Evolution of motor systems: The reticulospinal pathways. American Zoologist, 24, 733–753. Davis, G. R., Jr., & McClellan, A. D. (1994a). Long distance axonal regeneration of identified lamprey reticulospinal neurons. Experimental Neurology, 127, 94–105. Davis, G. R., Jr., & McClellan, A. D. (1994b). Extent and time course of restoration of descending brainstem projections in spinal cord-transected lamprey. The Journal of Comparative Neurology, 344, 65–82. Depoortere, R., Di Scala, G., Angst, M. J., & Sandner, G. (1990a). Differential pharmacological reactivity of aversion induced by stimulation of periaqueductal gray or mesencephalic locomotor region. Pharmacology, Biochemistry and Behavior, 37, 311–316. Depoortere, R., Di Scala, G., & Sandner, G. (1990b). Treadmill locomotion and aversive effects induced by electrical
stimulation of the mesencephalic locomotor region in the rat. Brain Research Bulletin, 25, 723–727. Drew, T., Andujar, J. E., Lajoie, K., & Yakovenko, S. (2008). Cortical mechanisms involved in visuomotor coordination during precision walking. Brain Research Reviews, 57, 199–211. Dubuc, R. (2009). Locomotor regions in the midbrain (MLR) and diencephalon (DLR). In M. D. Binder, N. Hirokawa & U. Windhorst (Eds.), Encyclopedia of neuroscience. Berlin Heidelberg: Springer-Verlag GmbH. Dubuc, R., Brocard, F., Antri, M., Fénelon, K., Gariépy, J. F., Smetana, R., et al. (2008). Initiation of locomotion in lampreys. Brain Research Reviews, 57, 172–182. Dubuc, R., & Grillner, S. (1989). The role of spinal cord inputs in modulating the activity of reticulospinal neurons during fictive locomotion in the lamprey. Brain Research, 483, 196–200. Ebert, U., & Ostwald, J. (1991). The mesencephalic locomotor region is activated during the auditory startle response of the unrestrained rat. Brain Research, 565, 209–217. El Manira, A., Pombal, M. A., & Grillner, S. (1997). Diencephalic projection to reticulospinal neurons involved in the initiation of locomotion in adult lampreys Lampetra fluviatilis. The Journal of Comparative Neurology, 389, 603–616. Garcia-Rill, E. (1991). The pedunculopontine nucleus. Progress in Neurobiology, 36, 363–389. Garcia-Rill, E., Kinjo, N., Atsuta, Y., Ishikawa, Y., Webber, M., & Skinner, R. D. (1990). Posterior midbraininduced locomotion. Brain Research Bulletin, 24, 499–508. Garcia-Rill, E., & Skinner, R. D. (1987a). The mesencephalic locomotor region. I. Activation of a medullary projection site. Brain Research, 411, 1–12. Garcia-Rill, E., & Skinner, R. D. (1987b). The mesencephalic locomotor region. II. Projections to reticulospinal neurons. Brain Research, 411, 13–20. Garcia-Rill, E., Skinner, R. D., Conrad, C., Mosley, D., & Campbell, C. (1986). Projections of the mesencephalic locomotor region in the rat. Brain Research Bulletin, 17, 33–40. Garcia-Rill, E., Skinner, R. D., & Fitzgerald, J. A. (1985). Chemical activation of the mesencephalic locomotor region. Brain Research, 330, 43–54. Garcia-Rill, E., Skinner, R. D., Gilmore, S. A., & Owings, R. (1983). Connections of the mesencephalic locomotor region (MLR). II. Afferents and efferents. Brain Research Bulletin, 10, 63–71. Graham Brown, T. (1911). Studies in the physiology of the nervous system. VIII: Neural balance and reflex reversal, with a note on progression in the decerebrate guinea pig. Quarterly Journal Experimental Physiology, 4, 273–288. Grillner, S. (1973). Locomotion in the spinal cat. In R. B. Stein, K. G. Pearson, R. S. Smith & J. B. Redford (Eds.), Control of posture and locomotion (pp. 515–533). New York: Plenum Press.
68 Grillner, S. (1981). Control of locomotion in bipeds, tetrapods, and fish. Sect. 1. In: V. B. Brooks (Ed.), Handbook of physiology. Section 1. The nervous system. Motor control (Vol. 2, pp. 1179–1236). Bethesda, MD: American Physiological Society. Grillner, S. (1985). Neurobiological bases of rhythmic motor acts in vertebrates. Science, 228, 143–149. Grillner, S., & Dubuc, R. (1988). Control of locomotion in vertebrates: Spinal and supraspinal mechanisms. Advances in Neurology, 47, 425–453. Grillner, S., Ekeberg, O., El Manira, A., Lansner, A., Parker, D., Tegner, J., et al. (1998). Intrinsic function of a neuronal network—A vertebrate central pattern generator. Brain Research Reviews, 26, 184–197. Grillner, S., Georpopoulos, A. P., & Jordan, L. M. (1997). Selection and initiation of motor behavior. In P. S. G. Stein, S. Grillner, A. I. Selverston & D. G. Stuart (Eds.), Neurons, networks, and motor behavior (pp. 3–19). Cambridge: The MIT Press. Grillner, S., & Rossignol, S. (1978). Contralateral reflex reversal controlled by limb position in the acute spinal cat injected with clonidine i.v. Brain Research, 144, 411–414. Grillner, S., Wallén, P., Saitoh, K., Kozlov, A., & Robertson, B. (2008). Neural bases of goal-directed locomotion in vertebrates—An overview. Brain Research Reviews, 57, 2–12. Hikosaka, O. (1991). Role of the forebrain in oculomotor function. Progress in Brain Research, 87, 101–107. Hultborn, H. (2001). State-dependent modulation of sensory feedback. Journal de Physiologie, 533, 5–13. Iwakiri, H., Oka, T., Takakusaki, K., & Mori, S. (1995). Stimulus effects of the medial pontine reticular formation and the mesencephalic locomotor region upon medullary reticulospinal neurons in acute decerebrate cats. Neuroscience Research, 23, 47–53. Jahn, K., Deutschländer, A., Stephan, T., Kalla, R., Wiesmann, M., Strupp, M., et al. (2008). Imaging human supraspinal locomotor centers in brainstem and cerebellum. Neuroimage, 39, 786–792. Jiang, Z. G., & Dun, N. J. (1986). Presynaptic suppression of excitatory postsynaptic potentials in rat ventral horn neurons by muscarinic agonists. Brain Research, 381, 182–186. Jones, B. E. (1990). Immunohistochemical study of choline acetyltransferase-immunoreactive processes and cells innervating the pontomedullary reticular formation in the rat. The Journal of Comparative Neurology, 295, 485–514. Jordan, L. M. (1986). Initiation of locomotion from the mammalian brainstem. In S. Grillner, P. S. G. Stein, D. G. Stuart, H. Forssberg & R. M. Herman (Eds.), Neurobiology of vertebrate locomotion (pp. 21–37). London: Macmillan. Jordan, L. M. (1998). Initiation of locomotion in mammals. Annals of the New York Academy of Sciences, 860, 83–93.
Karachi, C., Grabli, D., Bernard, F. A., Tandé, D., Wattiez, N., Belaid, H., et al. (2010). Cholinergic mesencephalic neurons are involved in gait and postural disorders in Parkinson disease. Journal of Clinical Investigation, 120, 2745–2754. Kinjo, N., Atsuta, Y., Webber, M., Kyle, R., Skinner, R. D., & Garcia-Rill, E. (1990). Medioventral medulla-induced locomotion. Brain Research Bulletin, 24, 509–516. Krawitz, S., Fedirchuk, B., Dai, Y., Jordan, L. M., & McCrea, D. A. (2001). State-dependent hyperpolarization of voltage threshold enhances motoneurone excitability during fictive locomotion in the cat. Journal de Physiologie, 532, 271–281. Lai, Y. Y., Clements, J. R., Wu, X. Y., Shalita, T., Wu, J. P., Kuo, J. S., et al. (1999). Brainstem projections to the ventromedial medulla in cat: Retrograde transport horseradish peroxidase and immunohistochemical studies. The Journal of Comparative Neurology, 408, 419–436. Le Ray, D., Brocard, F., Bourcier-Lucas, C., Auclair, F., Lafaille, P., & Dubuc, R. (2003). Nicotinic activation of reticulospinal cells involved in the control of swimming in lampreys. The European Journal of Neuroscience, 17, 137–148. Le Ray, D., Brocard, F., & Dubuc, R. (2004). Muscarinic modulation of the trigemino-reticular pathway in lampreys. Journal of Neurophysiology, 92, 926–938. Le Ray, D., Juvin, L., Boutin, T., Auclair, F., & Dubuc, R. (2010). A neuronal substrate for a state-dependent modulation of sensory inputs in the brainstem. The European Journal of Neuroscience, 32, 53–59. Lozano, A. M., & Snyder, B. J. (2008). Deep brain stimulation for parkinsonian gait disorders. Journal of Neurology, 255(Suppl. 4), 30–31. Marín, O., Smeets, W. J., & González, A. (1997). Distribution of choline acetyltransferase immunoreactivity in the brain of anuran (Rana perezi, Xenopus laevis) and urodele (Pleurodeles waltl) amphibians. The Journal of Comparative Neurology, 382, 499–534. Marlinsky, V. V., & Voitenko, L. P. (1991). The effect of procaine injection into the medullary reticular formation on forelimb muscle activity evoked by mesencephalic locomotor region and vestibular stimulation in the decerebrated guinea-pig. Neuroscience, 45, 753–759. Mazzone, P., Lozano, A., Stanzione, P., Galati, S., Scarnati, E., Peppe, A., et al. (2005). Implantation of human pedunculopontine nucleus: A safe and clinically relevant target in Parkinson’s disease. Neuroreport, 16, 1877–1881. Ménard, A., Auclair, F., Bourcier-Lucas, C., Grillner, S., & Dubuc, R. (2007). GABAergic projections to the mesencephalic locomotor region in the lamprey Petromyzon marinus. The Journal of Comparative Neurology, 501, 260–273. Mesulam, M. M., Mufson, E. J., Levey, A. I., & Wainer, B. H. (1983). Cholinergic innervation of cortex by the basal forebrain: Cytochemistry and cortical connections of the septal
69 area, diagonal band nuclei, nucleus basalis (substantia innominata), and hypothalamus in the rhesus monkey. The Journal of Comparative Neurology, 214, 170–197. Milner, K. L., & Mogenson, G. J. (1988). Electrical and chemical activation of the mesencephalic and subthalamic locomotor regions in freely moving rats. Brain Research, 452, 273–285. Mitani, A., Ito, K., Hallanger, A. E., Wainer, B. H., Kataoka, K., & McCarley, R. W. (1988). Cholinergic projections from the laterodorsal and pedunculopontine tegmental nuclei to the pontine gigantocellular tegmental field in the cat. Brain Research, 451, 397–402. Mitchell, I. J., Dean, P., & Redgrave, P. (1988a). The projection from superior colliculus to cuneiform area in the rat. II. Defence-like responses to stimulation with glutamate in cuneiform nucleus and surrounding structures. Experimental Brain Research, 72, 626–639. Mitchell, I. J., Redgrave, P., & Dean, P. (1988b). Plasticity of behavioural response to repeated injection of glutamate in cuneiform area of rat. Brain Research, 460, 394–397. Mori, S., Sakamoto, T., Ohta, Y., Takakusaki, K., & Matsuyama, K. (1989). Site-specific postural and locomotor changes evoked in awake, freely moving intact cats by stimulating the brainstem. Brain Research, 505, 66–74. Musienko, P. E., Zelenin, P. V., Lyalka, V. F., Orlovsky, G. N., & Deliagina, T. G. (2008). Postural performance in decerebrated rabbit. Behavioural Brain Research, 190, 124–134. Nieuwenhuys, R. (1972). Topological analysis of the brain stem of the lamprey Lampetra fluviatilis. The Journal of Comparative Neurology, 145, 165–178. Nieuwenhuys, R. (1977). The brain of the lamprey in a comparative perspective. Annals of the New York Academy of Sciences, 299, 97–145. Noga, B. R., Kriellaars, D. J., Brownstone, R. M., & Jordan, L. M. (2003). Mechanism for activation of locomotor centers in the spinal cord by stimulation of the mesencephalic locomotor region. Journal of Neurophysiology, 90, 464–1478. Northcutt, R. G. (1979). Experimental determination of the primary trigeminal projections in lampreys. Brain Research, 163, 323–327. Ohta, Y., & Grillner, S. (1989). Monosynaptic excitatory amino acid transmission from the posterior rhombencephalic reticular nucleus to spinal neurons involved in the control of locomotion in lamprey. Journal of Neurophysiology, 62, 1079–1089. Orlovsky, G. N. (1970). Work of the reticulo-spinal neurones during locomotion. Biofizika, 4, 728–737. Orlovsky, G. N., Deliagina, T. G., & Grillner, S. (1999). Neural control of locomotion, from mollusc to man. New York: Oxford University Press, 322pp. Pombal, M. A., Marín, O., & González, A. (2001). Distribution of choline acetyltransferase-immunoreactive structures in the lamprey brain. The Journal of Comparative Neurology, 431, 105–126.
Rossignol, S. (1996). Neural control of stereotypic limb movements. In L. B. Rowel & J. T. Sheperd (Eds.), Handbook of physiology. Section 12. Exercise: Regulation and integration of multiple systems (pp. 173–215). New York: Oxford University Press. Rossignol, S., Dubuc, R., & Gossard, J. P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86, 89–154. Rossignol, S., & Gauthier, L. (1980). An analysis of mechanisms controlling the reversal of crossed spinal reflexes. Brain Research, 182, 31–45. Rossignol, S., Julien, C., & Gauthier, L. (1981). Stimulus–response relationships during locomotion. Canadian Journal of Physiology and Pharmacology, 59, 667–674. Rovainen, C. M. (1967). Physiological and anatomical studies on large neurons of central nervous system of the sea lamprey (Petromyzon marinus) I. Müller and Mauthner cells. Journal of Neurophysiology, 30, 1000–1023. Rovainen, C. M. (1974). Synaptic interactions of reticulospinal neurons and nerve cells in the spinal cord of the sea lamprey. The Journal of Comparative Neurology, 154, 207–224. Scanziani, M., Gahwiler, B. H., & Thompson, S. M. (1995). Presynaptic inhibition of excitatory synaptic transmission by muscarinic and metabotropic glutamate receptor activation in the hippocampus: Are Ca2þ channels involved? Neuropharmacology, 34, 1549–1557. Shik, M. L., & Orlovsky, G. N. (1976). Neurophysiology of locomotor automatism. Physiological Reviews, 56, 465–501. Shik, M. L., Severin, F. V., & Orlovskiĭ, G. N. (1966). Control of walking and running by means of electric stimulation of the midbrain. Biofizika, 11, 659–666. Sholomenko, G. N., Funk, G. D., & Steeves, J. D. (1991). Avian locomotion activated by brainstem infusion of neurotransmitter agonists and antagonists. I. Acetylcholine excitatory amino acids and substance P. Experimental Brain Research, 85, 659–673. Sinnamon, H. M. (1993). Preoptic and hypothalamic neurons and the initiation of locomotion in the anesthetized rat. Progress in Neurobiology, 41, 323–344. Sirota, M. G., & Shik, M. L. (1973). Locomotion of the cat on stimulation of the mesencephalon (Article in Russian). Fiziologicheskiı˘ zhurnal SSSR imeni I. M. Sechenova, 59, 1314–1321. Sirota, M. G., Viana Di Prisco, G., & Dubuc, R. (2000). Stimulation of the mesencephalic locomotor region elicits controlled swimming in semi-intact lampreys. The European Journal of Neuroscience, 12, 4081–4092. Skinner, R. D., & Garcia-Rill, E. (1984). The mesencephalic locomotor region (MLR) in the rat. Brain Research, 323, 385–389. Skinner, R. D., Kinjo, N., Ishikawa, Y., Biedermann, J. A., & Garcia-Rill, E. (1990). Locomotor projections from the pedunculopontine nucleus to the medioventral medulla. NeuroReport, 1, 207–210.
70 Smetana, R. W., Alford, S., & Dubuc, R. (2007). Muscarinic receptor activation elicits sustained, recurring depolarization in reticulospinal neurons. Journal of Neurophysiology, 97, 3181–3192. Smetana, R., Juvin, L., Dubuc, R., & Alford, S. (2010). A parallel cholinergic brainstem pathway for enhancing locomotor drive. Nature Neuroscience, 13, 731–738. Smolders, I., Bogaert, L., Ebinger, G., & Michotte, Y. (1997). Muscarinic modulation of striatal dopamine, glutamate, and GABA release, as measured with in vivo microdialysis. Journal of Neurochemistry, 68, 1942–1948. Steeves, J. D., & Jordan, L. M. (1984). Autoradiographic demonstration of the projections from the mesencephalic locomotor region. Brain Research, 307, 263–276. Stefani, A., Lozano, A. M., Peppe, A., Stanzione, P., Galati, S., Tropepi, D., et al. (2007). Bilateral deep brain stimulation of the pedunculopontine and subthalamic nuclei in severe Parkinson’s disease. Brain, 130, 1596–1607. Sterman, M. B., & Fairshild, M. D. (1966). Modification of locomotor performance by reticular formation and basal forebrain stimulation in the cat: Evidence for reciprocal systems. Brain Research, 2, 205–217. Swain, G. P., Snedeker, J. A., Ayers, J., & Selzer, M. E. (1993). Cytoarchitecture of spinal-projecting neurons in the brain of the larval sea lamprey. The Journal of Comparative Neurology, 336, 194–210. Takakusaki, K. (2003). Function of cortical basal nuclei: Pathophysiology of Parkinson’s disease. Nippon Seirigaku Zasshi, 65, 113–129.
Takakusaki, K. (2008). Forebrain control of locomotor behaviors. Brain Research Reviews, 57, 192–198. Viana Di Prisco, G., Boutin, T., Petropoulos, D., Brocard, F., & Dubuc, R. (2005). The trigeminal sensory relay to reticulospinal neurones in lampreys. Neuroscience, 131, 535–546. Viana Di Prisco, G., Ohta, Y., Bongianni, F., Grillner, S., & Dubuc, R. (1995). Trigeminal inputs to reticulospinal neurones in lampreys are mediated by excitatory and inhibitory amino acids. Brain Research, 695, 76–80. Viana Di Prisco, G., Pearlstein, É., Le Ray, D., Robitaille, R., & Dubuc, R. (2000). A cellular mechanism for the transformation of a sensory input into a motor command. The Journal of Neuroscience, 20, 8169–8176. Viana Di Prisco, G., Pearlstein, É., Robitaille, R., & Dubuc, R. (1997). Sensory-evoked NMDA plateau potentials and their role in the initiation of locomotion. Science, 278, 1122–1125. Vinay, L., Bongianni, F., Ohta, Y., Grillner, S., & Dubuc, R. (1998a). Spinal inputs from lateral columns to reticulospinal neurons in lampreys. Brain Research, 808, 279–293. Vinay, L., Bussières, N., Shupliakov, O., Dubuc, R., & Grillner, S. (1998b). Anatomical study of spinobulbar neurons in lampreys. The Journal of Comparative Neurology, 397, 475–492. Vinay, L., & Grillner, S. (1993). The spino-reticulo-spinal loop can slow down the NMDA-activated spinal locomotor network in lamprey. Neuroport, 4, 609–612. Whelan, P. J. (1996). Control of locomotion in the decerebrate cat. Progress in Neurobiology, 49, 481–515.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 5
Face sensorimotor cortex: Its role and neuroplasticity in the control of orofacial movements Barry J. Sessle* Faculty of Dentistry, University of Toronto, Toronto, Ontario, Canada
Abstract: The range and complexity of orofacial movements require sophisticated neural circuitries that provide for the coordination and control of these movements and their integration with other motor patterns such as those associated with breathing and walking. This chapter is dedicated to Jim Lund whose many research studies have made major contributions to our knowledge of the role of brainstem and cerebral cortex in orofacial motor control. Our own investigations using intracortical microstimulation (ICMS), cortical cold block, and single neuron recordings have documented that the face primary motor area (MI) and primary somatosensory area (SI) are involved in the control not only of elemental and learned orofacial movements but also of the so-called semiautomatic movements such as mastication and swallowing, the control of which have been largely attributed in the past to brainstem mechanisms. Recent studies have also documented that neuroplasticity of the face sensorimotor cortex is a feature of humans and animals trained in a novel oral motor behavior, and that it reflects dynamic and adaptive events that can be modeled by behaviorally significant experiences, including pain and other alterations to the oral environment. Furthermore, our findings of the disruptive effects of the face sensorimotor cortex cold block indicate that the face MI and SI are also critical in the successful performance of an orofacial motor skill once it is learned. Future studies aimed at the further demonstration of such changes and at their underlying mechanisms and their sequence of appearance in the face sensorimotor cortex and associated cortical areas represent crucial steps for understanding the intracortical processes underlying neuroplasticity related to oral motor learning and adaptation. In view of the role that cortical neuronal ensembles play in motor execution, learning, and adaptation (Nicolelis and Lebedev, 2009), these studies should include the properties and plasticity of neuronal ensembles in several related cortical areas in addition to a specific focus on single neurones or efferent microzones within the face MI or SI. As recently noted (Martin, 2009; Sessle et al., 2007, 2009), such research approaches are also important for developing improved rehabilitative strategies to exploit these mechanisms in humans suffering from chronic orofacial pain or sensorimotor disorders. *Corresponding author. Tel.: þ1-416-979-4921; Fax: þ1-417-979-4936 DOI: 10.1016/B978-0-444-53825-3.00010-3
71
72
Keywords: face M1 cortex; face S1 cortex; cortical masticatory area; chewing; tongue protrusion; learning orofacial movements; intracortical microstimulation.
The range and complexity of orofacial movements It is a pleasure to be able to contribute to this book dedicated to Jack Feldman, Serge Rossignol, and Jim Lund who collectively have made seminal contributions to our understanding of the mechanisms underlying breathing, walking, and chewing. This chapter is dedicated to the memory of Jim Lund whose many studies provided much of our current knowledge of the neural mechanisms involved in the genesis and control of chewing, and he also provided important insights into processes controlling other types of orofacial movements as well as orofacial pain. Jim and I shared not only many common research interests but also had somewhat similar backgrounds. We both obtained our dental training in Australia, but left Australia for Canada about 40 years ago. He then undertook his PhD training in Canada at the University of Western Ontario (with another Australian, Peter Dellow, as his supervisor) while I came to Canada with a PhD already “under my belt” (and having had post-doctoral training with Ron Dubner at the NIH in the USA). Jim and I took up academic positions in Montreal and Toronto, respectively, in the 1970s. We then also took somewhat similar academic paths, including Deanships at dental schools at our respective universities. In addition, because of our common interests and backgrounds, we also collaborated on brainstem and cortical research studies in the 1970s (e.g Lund and Sessle, 1974; Lund et al, 1979) and more recently co-authored or co-edited books and reviews dealing with orofacial pain and sensorimotor control (e.g. Sessle et al, 2008; Lund et al, 2009). Jim’s contributions to our understanding of orofacial pain and sensorimotor control were manifested not only in his numerous
important papers, books and reviews but also in his role as, in a sense, the “conscience” of the orofacial sensory and motor fields. Consequently, his sudden and unexpected death in December, 2009 represents a profound loss to these fields as well as the loss of a colleague and friend for many of us. While most of the literature on motor control has focused on mechanisms controlling the limb muscles, and reviews of the topic of motor control often completely neglect the topic of orofacial motor control, it is important to recognize that many orofacial movements necessitate complex neural processing in order to provide for the exquisite motor control that is required to coordinate the vast array of muscles in the orofacial region. It should also be appreciated that these muscles serve important and diverse functions, some of which are unique to the orofacial region and help sustain the very life of the animal, for example, the so-called semiautomatic movements of chewing and swallowing. Orofacial movements include voluntary movements that are driven by centers in the central nervous system (CNS) and that can be refined through sensory inputs to reflex circuits and to higher brain centers in the CNS. These voluntary movements can range from limited movements involving a limited number of muscles (e.g., jaw opening or tongue protrusion), to the diversity of tongue positions and shapes that some of us can achieve, to the even more complex and sophisticated movements that characterize speech which involves an elaborate bilateral integration of sensory and motor pathways and central programming of muscles of the jaw, face, tongue, palate, pharynx, and larynx and other parts of the respiratory tract. There is also a vast array of reflexes manifested in these muscles, some involving relatively “simple” reflex circuits (e.g., jaw-opening
73
reflex, blink reflex). Other orofacial reflexes are more complicated and involve several brainstem motor neuron pools bilaterally (e.g., cough, gag), and others are even more complex and require the bilateral activity of numerous muscle groups and highly integrated CNS circuits (e.g., swallowing). Some of these complex motor events may also include a voluntary component as well as a reflex component (e.g., the buccopharyngeal phase of a swallow), and some may be cyclic in nature. An example of the latter is chewing (mastication) which, like breathing and walking, involves cyclic events (e.g., alternating rhythm of jaw opening and closing) that are driven by a CNS program produced by central pattern generator (s); the generator(s) has collectively been termed the “chewing center” and has been extensively explored by Jim Lund and his colleagues (see Chapter 15). However, in considering chewing, it is also important to note that not all “masticatory” muscles show this cyclic rhythmic activity; for example, during chewing, the tongue and several of the facial muscles are important in manipulating the food bolus, but they may not necessarily manifest rhythmic activity. It is also noteworthy that these other associated muscle activities (i.e., of tongue and facial muscles) as well as the jaw-opening and jaw-closing muscle activities producing the cyclic jaw movements that characterize chewing are sensitive to afferent inputs acting through brainstem reflex circuits or higher brain centers. Indeed, virtually all of the orofacial functions mentioned above are not purely “motor” but rather are “sensorimotor” since they depend upon or utilize sensory inputs or feedback to initiate or guide them (for review, see Dubner et al., 1978; Sessle, 2009a). Some of these sophisticated orofacial motor functions are laid down at birth, others develop as the infant animal matures. For example, swallowing develops in utero and most mammals are born as suckling and swallowing animals, whereas chewing, like walking, has features of a learned motor function. As the infant matures, the jaw muscle activities characteristic of chewing
become increasingly under central control and more elaborate to allow for horizontal as well as vertical jaw movements. In addition, other muscle groups (e.g., tongue, facial) become engaged in a coordinated manner with the jaw muscles to provide for more refinement and management of chewing; it seems likely that the eruption of the teeth provides important sensory inputs from periodontal receptors that also assist in the development of this particular level of control. Thus, orofacial movements require sophisticated neural circuits that allow for bilateral integration of diverse and complex motor functions. This integration includes neural and chemical regulatory processes involved in respiration, since many orofacial muscles (e.g., genioglossus, anterior digastric) also function as accessory respiratory muscles; their respiratory-related activity is especially evident when airway patency is threatened. There is also a requirement for coordination of orofacial motor activities with locomotion as well as respiration so that indeed humans can breathe, walk, and chew at the same time! Furthermore, there are several features that distinguish these orofacial sensorimotor functions and their underlying mechanisms from spinal sensorimotor processes and movements: in addition to the large number of muscles that may be involved, often requiring bilateral muscle activities, these features include the arrangement of the various sensory nuclei and motor nuclei into distinct neuronal pools in the brainstem, and also unique aspects of the peripheral mechanisms (see Sessle, 2009a,b). Thus, neural mechanisms found to be involved in limb or spinal motor control cannot automatically be assumed to be applicable also to orofacial motor control. The role and features of the brainstem circuits and central pattern generators involved in orofacial movements are elaborated upon in other chapters in this volume (see Kolta; Chapter 15), and the remaining sections of this chapter will focus
74
on the properties of higher brain centers, particularly those of the sensorimotor cortex, in the control of orofacial movements.
Face sensorimotor cortex and orofacial motor control Jim Lund's early research focused not only on brainstem pattern generator(s) but also on the role of the cerebral cortex in the control of orofacial movements. For example, one of his first postdoctoral studies after coming to Montreal was to describe the properties of single cortical neurons during jaw and other orofacial movements in awake monkeys (Lund and Lamarre, 1974). Around this time, he and I also collaborated to document the orofacial input features of neurons in the sensorimotor cortex and the adjacent cortical regions of the cat (Lund and Sessle, 1974), and he also subsequently documented in rabbits the organization of the cortical control of rhythmic masticatory-like movements (e.g., Lund et al., 1984). Over the following three decades, studies employing intracortical microstimulation (ICMS) or recordings of peripherally evoked responses and movement-related activity patterns of single neurons in the primary motor area (MI) and somatosensory area (SI) of the cerebral cortex in monkeys, cats, and rodents, as well as stimulation, recording, or imaging studies in humans, have provided important insights into the function of the sensorimotor cortex. They have shown that MI plays a critical role in the planning, initiation, and execution of limb movements, that SI contributes to the control of these movements, that MI and SI are important in the acquisition of new motor skills involving the limbs, and that sensory inputs to these sensorimotor cortex areas from the limbs are important in the modification, control, and learning of these movements (for review, see Asanuma, 1989; Ebner, 2005; Nicolelis and Lebedev, 2009; Sanes and Donoghue, 2000).
In the case of face MI, studies by ourselves and others have shown that, consistent with stimulation and imaging studies in humans (for review, see Martin and Sessle, 1993; Martin et al., 2004; Nordstrom, 2007), short-train ICMS of monkey face MI evokes elemental movements such as jaw opening or tongue protrusion (e.g., Huang et al., 1988; Martin et al., 1999; Murray and Sessle, 1992a; Fig. 1). The ICMS findings of multiple efferent microzones or “nests” for each muscle suggest a face MI organization reflecting the different functional contingencies in which that muscle participates, and perhaps the need for numerous neuronal ensembles in order to provide a comprehensive neural encoding of several motor parameters of that muscle (e.g., Nicolelis and Lebedev, 2009). In addition, long-train ICMS in monkeys evokes various patterns of swallowing and masticatory-like movements from multiple microzones in the more lateral “cortical masticatory area” (CMA) and overlapping swallowing area as well as in the face MI itself (see Huang et al., 1988, 1989b; Martin et al., 1999; Sessle et al., 2007; Fig. 1). Furthermore, studies employing lesions or reversible cold block of synaptic activity in selected cortical areas in the monkey have shown that blockade of not only the CMA/swallow cortex but also specifically of the face MI markedly affects the jaw muscle activity characteristic of chewing (e.g., Larson et al., 1980; Luschei and Goodwin, 1975; Narita et al., 1999; Yamamura et al., 2002); cold block, specifically of the face MI, also interferes with trained voluntary tongue movements that the monkey has learned (Murray et al., 1991). In subprimates also, masticatory movements as well as elemental orofacial movements can be evoked from the face MI as well as from the CMA/swallow cortex and can be disrupted by lesions of these areas (e.g., Hiraba and Sato, 2005; Lund et al., 1984; Neafsey, 1986; Sanes and Donoghue, 2000; Sessle et al., 2007). These findings are consistent with clinical reports as well as other experimental findings of orofacial motor deficits following damage to the face MI (see
75 arc.
(a)
cen.
(b) Med. 1 2 3 4 5 6 7 8 9 10 11 12 mm Lat.
H1 RT
Rostral
sca
Rostral H1 RT 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 mm
1 mm
area 6
area 3a
area 3b
area 1
Lateral
(c)
9
Rostral 8 7
6
5
4
3
2
1
H1 RT 0 1
2
3
4
5
6
7
8
9 mm
Med. 1 Hand
2
VVV
3 4 5
CCCCCCCCCVVVVCC
VVV
6
8
II
VV lIII
VVVV l
VV C
II V V V V
VV l VV
II I II I
III
V
l l l l ll l l l l l l l l l l l
Rostral arc.
3 2 10
ICMS C RJMc Hand Face I RJMi Jaw V RJMv Tongue
mm
cen. 34 12
54
12 mm Lat.
CC
VVVV
CCC
VV
C I I VV V I I I VVVVVVVVVV
Vl
10
VVV
VV I I I CCCC
VVVVV
9
11
C CCCC
VVV
7
5 mm
Fig. 1. Distribution of sites in the right (RT) pericentral cortex of a monkey (H1) in which intracortical microstimulation (ICMS) could evoke masticatory-like movements (RJMs) or elemental twitch movements of the face, jaw, or tongue. (a) The cortical region explored by ICMS (both long train and short train). Each point of entry of an ICMS microelectrode penetration is marked with a symbol: open circle, no evoked RJMs; filled circle, RJMs evoked by long-train ICMS in this penetration. The contour lines enclose the penetrations in which an elemental twitch movement could be evoked by short-train ICMS, thus denoting the face MI. (b) The cytoarchitectonic boundaries in the “unfolded” pericentral cortex. The stippled area demarcates the region (area 4) within which large layer V pyramidal cells were found. Other cytoarchitectonic areas are delineated by the three unbroken lines. (c) The distribution pattern of RJMs in the “unfolded” pericentral cortex. 0 mm: top of the rostral bank of the central sulcus; dashed line on the right represents the bottom of the central sulcus (see insert). RJMc, vertical jaw movement with lateral deviation to side contralateral to ICMS site; RJM I, vertical jaw movement with lateral deviation to side ipsilateral to ICMS site; RJM v, vertical jaw movement with limited lateral deviation. From Huang et al. (1989b).
76 Table 1. Proportions of neurons showing activity in the face SI, MI, CMA/swallow cortex of monkey in relation to trained tongue-protrusion task, chewing, and/or swallowing
Face SI Face MI CMA/ swallow cortex
Task related (%)
Chewing related (%)
Swallow related (%)
61 71 15
83 81 88
0 26 55
Sample size ¼ 700 neurons, based on Martin et al. (1997) and Yao et al. (2002).
Barritt and Smithard, 2009; Hiraba and Sato, 2005; Martin and Sessle, 1993; Nordstrom, 2007). In addition, many MI neurons recorded at these same ICMS-defined sites in the face MI show activity related to chewing, swallowing, and/or to a trained task involving tongue or jaw muscles (e.g., Table 1), and most receive mechanosensory inputs especially from tongue, lips, and teeth (e.g., Hiraba and Sato, 2005; Hoffman and Luschei, 1980; Martin et al., 1999; Murray and Sessle, 1992b; Sessle et al., 2007; Yao et al., 2002). A close spatial matching of afferent input features and motor output effects at intracortical face MI sites was also a characteristic finding, suggesting the existence of intracortical mechanisms for orofacial sensorimotor integration (Huang et al., 1989a; Murray and Sessle, 1992a). These various findings collectively reveal that the face MI is involved in the control not only of elemental and learned orofacial movements but also of complex, semiautomatic movements such as mastication and swallowing that in the past have been largely attributed to brainstem control mechanisms. This is consistent with evidence that aspects of breathing and locomotion are also under the influence of the sensorimotor cortex. Analogous experiments employing ICMS, cortical cold block, or single neuron recordings in monkey face SI and CMA/swallow cortex during these task or semiautomatic orofacial movements (Huang et al., 1989b; Lin and Sessle, 1994; Lin
et al., 1998; Martin et al., 1997, 1999; Sessle et al., 2007) have documented that the face SI receives extensive orofacial sensory inputs and is involved in the fine control of these orofacial movements, consistent with the findings of analogous studies of the subprimate face sensorimotor cortex (e.g., Hiraba and Sato, 2005; Lund and Sessle, 1974; Lund et al., 1984). In contrast to the face SI and MI, very few CMA/swallow cortex neurons in monkeys show tongue task-related activity, although most do exhibit activity related to masticatory and/or swallow movements (e.g., Table 1), suggesting that the primate CMA/swallow cortex plays little if any role in voluntary trained movements but is more critical for the initiation and regulation of these semiautomatic movements. Such documentation of the properties and role of the face sensorimotor cortex in motor control and performance of orofacial movements provided fundamental details of these features that have allowed us to carry out recent experiments aimed at exploring whether and how these features may undergo neuroplastic changes that permit the acquisition of new motor skills or adaptation to an altered oral environment. The next section provides an overview of the findings of these experiments and their implications.
Neuroplasticity of face sensorimotor cortex Neuroplasticity reflects the brain's ability to undergo structural and functional changes throughout life; such changes represent crucial processes for CNS development, memory, motor skill acquisition and learning, and adaptation following peripheral sensory or motor nerve lesions or alterations to sensory inputs to the CNS. Neuroplastic changes in these contexts have been documented in the limb sensorimotor cortex and shown to involve several intracortical neurochemicals (e.g., acetylcholine, GABA, norepinephrine, glutamate) and associated unmasking of existing connections or altered synaptic
77
efficacy that can result in short-term changes in the excitability and activity of MI and SI neurons; longer-term changes may also occur as reflected in axonal sprouting, increased synaptogenesis, and changes in cortical volume (e.g., Buonomano and Merzenich, 1998; Ebner, 2005; Fox, 2009; Holtmaat and Svoboda, 2009; May, 2008; Nicolelis and Lebedev, 2009; Sanes and Donoghue, 2000). There is also recent evidence of the neuroplasticity of the face sensorimotor cortex in relation to the acquisition of novel orofacial motor skills and in adaptive processes following alterations to intraoral tissues and sensory inputs, as outlined below.
Acquisition of orofacial motor skills The extensive literature on the limb sensorimotor cortex has underscored the importance of MI and SI neuroplasticity and of sensory inputs to the sensorimotor cortex in the acquisition of limb motor skills (see Buonomano and Merzenich, 1998; May, 2008; Nicolelis and Lebedev, 2009; Sanes and Donoghue, 2000). The subject of cortical neuroplasticity as it relates to orofacial motor skill acquisition has received little attention until recently, except for reports of clinical and animal behavioral experiments that neuroplasticity or progressive return of function of the face sensorimotor cortex can occur following cortical damage or manipulations of orofacial sensory inputs such as those from the rodent vibrissae to the face SI (see Barritt and Smithard, 2009; Ebner, 2005; Fox, 2009; Martin, 2009; Martin and Sessle, 1993; Sanes and Donoghue, 2000). In addition, a recent ICMS study that involved the training of rats in a semiautomatic oral (tonguelicking) behavior reported that while tongue force training alone does not produce lasting changes in the size of the orofacial cortical motor representation, tongue-force training decreases the current thresholds necessary for eliciting an ICMSevoked motor response (Guggenmos et al., 2009). We have also recently tested for the
Table 2. Proportion of neurons with tongue protrusion-related activity in the face SI, MI, and CMA/swallow cortex of monkey before and after training in tongue-protrusion task
Face SI Face MI CMA/swallow cortex
Before training (%)
After training (%)
25 23 40
54* 79* 35
Sample size ¼ 290 neurons, based on Sessle et al. (2005, 2007). *P < 0.001.
possible neuroplasticity of the face MI (and the SI and CMA/swallow cortex) related to the learning of a novel orofacial motor skill (Sessle et al., 2007). ICMS and neuronal recordings were made in awake monkeys over a 1–2-month period before, and again after, a 2–6-week period of their training in a novel tongue-protrusion task; this task was comparable to that used in our earlier face MI studies (see above). Lingual sensory input appears critical to the monkey's ability to carry out the tongue-protrusion task, since selective bilateral local anesthetic block of the lingual nerve produced a marked reduction (from 60% to < 20%) in successful performance of the task, consistent with earlier findings of substantial lingual mechanosensory inputs to the face MI as well as the SI (see above). Compared with pretraining data in trained animals and data in control untrained animals, the trained animals had a significant increase in the proportion of MI efferent microzones for tongue protrusion (as revealed by ICMS) and a significant decrease in microzones for lateral tongue movement. Furthermore, there were significant increases in the proportions of neurons in MI, and also in SI, showing lingual mechanosensory inputs and in the proportions manifesting tongue protrusionrelated activity (e.g., Table 2). These findings are consistent with limb MI studies showing that the tuning properties of MI neurons change and indeed may gradually evolve during the learning of a motor task (see Ganguly and Carmena,
78
2009; Nicolelis and Lebedev, 2009; Sanes and Donoghue, 2000). Interestingly, there were no analogous changes apparent in the CMA/swallow cortex, suggesting that a differential expression of task-related neuroplasticity may occur in these three cortical areas. We have also carried out translational studies using transcranial magnetic stimulation (TMS) in humans and have shown that training humans for less than 1 h in a tongue task analogous to that used in our monkey studies is associated with a significantly increased TMS-defined tongue representation in MI and significantly decreased threshold and increased amplitude of TMS-evoked motor potentials in the tongue musculature (Boudreau et al., 2007; Svensson et al., 2006). These findings indicate that face MI neuroplasticity occurs when humans and animals are trained in a novel orofacial motor task. They suggest that, like limb MI, the face MI is integrally involved in the acquisition as well as in the performance of new motor skills. Intraoral alterations As noted above, neuroplasticity also occurs in limb MI and SI following peripheral sensory or motor nerve lesions or manipulations of sensory inputs. Alterations to or deafferentation of orofacial sites, including removal of the incisor tooth or vibrissae, induces neuroplastic changes in the face SI (e.g., Ebner, 2005; Fox, 2009; Henry et al., 2005; Holtmaat and Svoboda, 2009). In the case of the face MI, it has been reported that altering the vibrissal afferent input to the rodent cortex or lesioning the facial nerve that provides the motor supply to the vibrissal muscles results in ICMS threshold changes or a reduction in the vibrissal representation and an expansion of the adjacent forelimb representation in MI (e.g., Diamond et al., 2008; Ebner, 2005; Franchi, 2001; Sanes and Donoghue, 2000). No studies appear to have provided any detailed information of the effects of modifying orofacial sensory
inputs on the tongue or jaw representations in the face MI, except for recent TMS studies reporting changes in face MI output effects following local anesthesia of selected orofacial sites in humans (Halkjaer et al., 2006; Yildiz et al., 2004; cf. Ernberg et al., 2009). Yet, adaptive, and in some cases, maladaptive behaviors occur when the dental occlusion is modified or intraoral (e.g., lingual) nerves are damaged (see Feine and Carlsson, 2003; Robinson et al., 2004; Sessle et al., 2009). Therefore, we recently tested the effects of such manipulations on the ICMS-defined jaw (anterior digastric) and tongue (genioglossus) motor representations in the face MI in anesthetized rats. These manipulations included modifying the dental occlusion (by trimming of the mandibular incisors to take them out of occlusion with the maxillary incisors, or by extracting one of the incisors), and transecting the lingual nerve. Both types of intraoral alterations produced neuroplastic changes in the face MI. In contrast to our findings of depressive effects on MI excitability of orofacial noxious stimulation (see below), extraction of the mandibular incisor was associated 1 week later with a significant increase of the ICMS-defined jaw motor representation in the face MI compared with control animals (Avivi-Arber et al., 2009; Sessle et al., 2007; see Chapter 9). In contrast, trimming both rat mandibular incisors induced neuroplastic changes reflected in a significant reduction in the jaw and tongue representation that could be restored if the incisors were allowed to erupt back into occlusion, consistent with reports of face MI neuroplasticity with vibrissae trimming and regrowth in rodents (Ebner, 2005; Keller et al., 1996). Face MI neuroplasticity was also evident for 1–4 weeks after lingual nerve transection (Adachi et al., 2007). We also documented a small jaw and tongue motor representation in histologically confirmed sites in the face SI that could also express neuroplastic changes as a result of incisor extraction. A recent imaging study in humans has also reported neuroplasticity of the
79
face sensorimotor cortex following the placement of dental implants (Yan et al., 2008). Another factor affecting the face sensorimotor cortex could conceivably be pain. Pain can decrease so-called static (e.g., force, endurance, recovery) and dynamic (agonist/antagonist complementary function) motor activities as well as alter reflexes, motor strategy/planning and stability, and control of limb, trunk, neck, or jaw movements, often markedly impacting on the patient's quality of life in the case of chronic pain, with “maladaptive” changes contributing to the recurrence or maintenance of the chronic pain (see Boudreau et al., 2007; Hodges, 2008; Svensson and Graven-Nielsen, 2001). In the case of animal models of orofacial pain, many studies, including several by Lund et al. and by ourselves in awake or anesthetized animals, have also shown that acute orofacial pain induces or modifies motor behavior (see Boudreau et al., 2007; Sessle, 2006; Svensson and Graven-Nielsen, 2001; Chapter 15). While several pain-related changes in motor behavior involve principally local segmental or brainstem reflex circuits, there is a growing body of evidence pointing to the involvement of higher CNS centers such as the sensorimotor cortex; for example, in healthy humans, decreased MI excitability has been shown in association with acute pain induced by capsaicin applied to the skin or hypertonic saline injected into a limb (see Boudreau et al., 2007; Hodges, 2008; May, 2008). And whereas human face MI excitability has been reported to be unaffected by hypertonic saline-induced masseter muscle pain or capsaicin-induced facial or lingual pain (Halkjaer et al., 2006; Romaniello et al., 2000), we recently documented (Boudreau et al., 2007) that capsaicin-induced intraoral pain interferes with successful performance by humans of a novel tongue-protrusion task (also see below) and concomitantly reduces tongue MI excitability. These findings in humans are consistent with our data in a recent rat study where we found significantly increased ICMS thresholds for the tongue (genioglossus) motor representation for several
hours after acute noxious stimulation of the rat tongue (Adachi et al., 2008); the data suggest that this decreased MI excitability induced by noxious lingual sensory inputs reflects intracortical changes specifically on motor outputs to the region (tongue) in the vicinity of the noxious stimulus site, consistent with our earlier findings of close input–output coupling for the tongue MI representation (Huang et al., 1989a; Murray and Sessle, 1992a,b). It should be noted that these pain-related face MI studies showing a decreased MI excitability suggest that the aforementioned increase in face MI motor representation following incisor extraction is unlikely due to pain experienced by the animal from the extraction. Nonetheless, the pain-related face MI studies tested the effects only of acute orofacial pain, and it remains to be determined whether and how a more persistent orofacial pain also affects MI, as well as determine whether acute or chronic orofacial pain influences SI properties. Such effects are likely since there are reports of MI or SI excitability changes or reorganization in several chronic pain states in humans (see Apkarian et al., 2005; Flor, 2000; Hodges, 2008), and experiments using chronic orofacial pain animal models have documented neuroplastic changes in the trigeminal brainstem sensory nucleus complex that provides the orofacial sensory inputs to the face sensorimotor cortex (e.g., Iwata et al., 2004; Kwan et al., 1996; Okada-Ogawa et al., 2009; Sessle, 2000). Studies clarifying the cortical effects of chronic orofacial pain and their underlying mechanisms also have clinical relevance from a management standpoint: since motor training also produces MI neuroplasticity (see above), it is conceivable that applying training to reorganize the sensorimotor cortex may prove to be an effective approach to improve motor control in chronic orofacial pain patients. Another consideration arising from our recent findings of neuroplastic changes in the face sensorimotor cortex following intraoral manipulations is whether the cortical changes associated with
80
alterations to the teeth or lingual nerve function are the result of alterations in sensory inputs to the sensorimotor cortex that allow the animal to adapt to the modified intraoral state, or whether the cortical changes result from a sensorimotor behavior that has been modified by the animal to adjust to the altered intraoral condition. We are planning to test this question directly, but the literature (see above) on sensorimotor cortex neuroplasticity and motor control favors the former, suggesting that the neuroplasticity reflects dynamic, adaptive constructs responsive to alterations in the sensory environment and indeed underlies the animal's ability to acquire new motor skills or modify its motor behavior as it adjusts to the altered peripheral state. The literature also indicates that the MI of both humans and animals is specifically engaged during the acquisition phase of novel motor skills, and that the associated neuroplasticity is not the result of increased muscle or nerve excitability. Our human data and findings in earlier studies that changes in MI properties can occur within minutes as a new sensorimotor task is being learned (see Boudreau et al., 2007; Svensson et al., 2006) would further suggest that the neuroplastic changes in the face sensorimotor cortex are crucial for the adaptation and acquisition of new motor skills and behavior appropriate to the altered sensory environment.
Acknowledgments The author's studies cited above are supported by CIHR grant MOP-4918, the Canadian Foundation for Innovation, and the Ontario Ministry of Research and Innovation. The secretarial assistance of Fong Yuen and Dorothy Tsang is also gratefully acknowledged. Abbreviations CMA CNS
cortical masticatory area central nervous system
ICMS MI RJMc
RJMI
RJMv SI TMS
intracortical microstimulation primary motor area vertical jaw movement with lateral deviation to side contralateral to ICMS site vertical jaw movement with lateral deviation to side ipsilateral to ICMS site vertical jaw movement with limited lateral deviation primary somatosensory area transcranial magnetic stimulation
References Adachi, K., Lee, J. C., Hu, J. W., Yao, D., & Sessle, B. J. (2007). Motor cortex (MI) neuroplasticity associated with lingual nerve injury in rats. Somatosensory & Motor Research, 24, 97–109. Adachi, K., Murray, G. M., Lee, J.-C., & Sessle, B. J. (2008). Noxious lingual stimulation influences the excitability of the face primary motor cerebral cortex (face MI) in the rat. Journal of Neurophysiology, 100, 1234–1244. Apkarian, A. V., Bushnell, M. C., Treede, R. D., & Zubieta, J. K. (2005). Human brain mechanisms of pain perception and regulation in health and disease. European Journal of Pain, 9(4), 463–484. Asanuma, H. (1989). The motor cortex. New York: Raven Press, 189pp. Avivi-Arber, L., Lee, J. C., & Sessle, B. J. (2009). Effects of incisor extraction on jaw and tongue motor representations within face sensorimotor cortex of adult rats. The Journal of Comparative Neurology, 518, 1030–1045. Barritt, A. W., & Smithard, D. G. (2009). Role of cerebral cortex plasticity in the recovery of swallowing function following dysphagic stroke. Dysphagia, 24, 83–90. Boudreau, S., Romaniello, A., Wang, K., Svensson, P., Sessle, B. J., & Arendt-Nielsen, L. (2007). The effects of intra-oral pain on motor cortex neuroplasticity associated with short-term novel tongue-protrusion training in humans. Pain, 132, 169–178. Buonomano, D. V., & Merzenich, M. M. (1998). Cortical plasticity: From synapses to maps. Annual Review of Neuroscience, 21, 149–186. Diamond, M. E., von Heimendahl, M., Knutsen, P. M., Kleinfeld, D., & Ahissar, E. (2008). ‘Where’ and ‘what’ in the whisker sensorimotor system. Nature Reviews. Neuroscience, 9, 601–612.
81 Dubner, R., Sessle, B. J., & Storey, A. T. (1978). The neural basis of oral and facial function. New York: Plenum Press, 483pp. Ebner, F. F. (2005). Neural plasticity in adult somatic sensorymotor systems. Boca Raton, FL: CRC Press, 273pp. Ernberg, M., Serra, E., Baad-Hansen, L., & Svensson, P. (2009). Influence of topical anaesthesia on the corticomotor response to tongue training. Archives of Oral Biology, 54, 696–704. Feine, J. S., & Carlsson, G. (2003). Implant overdentures as minimum standard of care. Illinois: Quintessence Publishing Co, 162pp. Flor, H. (2000). The functional organization of the brain in chronic pain. Progress in Brain Research, 129, 313–322. Fox, K. (2009). Experience-dependent plasticity mechanisms for neural rehabilitation in somatosensory cortex. Philosophical Transactions of the Royal Society of London, 27, 185–196. Franchi, G. (2001). Persistence of vibrissal motor representation following vibrissal pad deafferentation in adult rats. Experimental Brain Research, 137, 180–189. Ganguly, K., & Carmena, J. M. (2009). Emergence of a stable cortical map for neuroprosthetic control. PLoS Biology, 7, 1–12. Guggenmos, D. J., Barbay, S., Bethel-Brown, C., Nudo, R. J., & Stanford, J. A. (2009). Effects of tongue force training on orolingual motor cortical representation. Behavioural Brain Research, 201(1), 229–232. Halkjaer, L., Melsen, B., McMillan, A. S., & Svensson, P. (2006). Influence of sensory deprivation and perturbation of trigeminal afferent fibers on corticomotor control of human tongue musculature. Experimental Brain Research, 170, 199–205. Henry, E. C., Marasco, P. D., & Catania, K. C. (2005). Plasticity of the cortical dentition representation after tooth extraction in naked mole-rats. The Journal of Comparative Neurology, 485, 64–74. Hiraba, H., & Sato, T. (2005). Cortical control of mastication in cats. 2. Deficits of masticatory movements following a lesion in the motor cortex. Somatosensory & Motor Research, 22, 183–192. Hodges, P. W. (2008). Changes in sensorimotor control in low back pain. In T. Graven-Nielsen, L. Arendt-Nielsen & S. Mense (Eds.), Fundamentals of Musculoskeletal Pain (pp. 445–459). Seattle: IASP Press. Hoffman, D. S., & Luschei, E. S. (1980). Responses of monkey precentral cortical cells during a controlled jaw bite task. Journal of Neurophysiology, 44, 333–348. Holtmaat, A., & Svoboda, K. (2009). Experience-dependent structural synaptic plasticity in the mammalian brain. Nature Reviews. Neuroscience, 10, 647–658. Huang, C. S., Sirisko, M. A., Hiraba, H., Murray, G. M., & Sessle, B. J. (1988). Organization of the primate face motor cortex as revealed by intracortical microstimulation and electrophysiological identification of afferent inputs and
corticobulbar projections. Journal of Neurophysiology, 59, 796–818. Huang, C. S., Hiraba, H., & Sessle, B. J. (1989a). Input-output relationships of the primary face motor cortex in the monkey (Macaca fascicularis). Journal of Neurophysiology, 61, 350–362. Huang, C.-S., Hiraba, H., & Sessle, B. J. (1989b). Topographical distribution and functional properties of cortically induced rhythmical jaw movements in the monkey (Macaca fascicularis). Journal of Neurophysiology, 61, 635–650. Iwata, K., Tsuboi, Y., Shima, A., Harada, T., Ren, K., Kanda, K., et al. (2004). Central neuronal changes after nerve injury: Neuroplastic influences of injury and aging. Journal of Orofacial Pain, 18, 293–298. Keller, A., Weintraub, N. D., & Miyashita, E. (1996). Tactile experience determines the organization of movement representations in rat motor cortex. NeuroReport, 7, 2373–2378. Kwan, C. L., Hu, J. W., & Sessle, B. J. (1996). Neuroplastic effects of neonatal capsaicin treatment on neurons in adult rat trigeminal nuclei principalis and subnucleus oralis. Journal of Neurophysiology, 75, 298–310. Larson, C. R., Byrd, K. E., Garthwaite, C. R., & Luschei, E. S. (1980). Alterations in the pattern of mastication after ablations of the lateral precentral cortex in rhesus macaques. Experimental Neurology, 70, 638–651. Lin, L.-D., & Sessle, B. J. (1994). Functional properties of single neurons in the primate face primary somatosensory cortex. III. Modulation of response to peripheral stimuli during trained orofacial motor behavior. Journal of Neurophysiology, 71, 2401–2413. Lin, L.-D., Murray, G. M., & Sessle, B. J. (1998). Effects on non-human primate mastication of reversible inactivation by cooling of the face primary somatosensory cortex. Archives of Oral Biology, 43, 133–141. Lund, J. P., & Lamarre, Y. (1974). Activity of neurons in lower precentral cortex during voluntary and rhythmical jaw movements in the monkey. Experimental Brain Research, 19, 282–299. Lund, J. P., & Sessle, B. J. (1974). Oral-facial and jaw muscle afferent projections to neurons in cat frontal cortex. Experimental Neurology, 45, 314–331. Lund, J. P., Sasamoto, K., Murakami, T., & Olsson, K. A. (1984). Analysis of rhythmical jaw movements produced by electrical stimulation of motor sensory cortex of rabbit. Journal of Neurophysiology, 52, 1014–1029. Luschei, E. S., & Goodwin, G. M. (1975). Role of monkey precentral cortex in control of voluntary jaw movements. Journal of Neurophysiology, 38, 146–157. Martin, R. E. (2009). Review article: Neuroplasticity and swallowing. Dysphagia, 24(2), 218–229. Martin, R. E., & Sessle, B. J. (1993). The role of the cerebral cortex in swallowing. Dysphagia, 8, 195–202.
82 Martin, R., Murray, G. M., Kemppainen, P., Masuda, Y., & Sessle, B. J. (1997). Functional properties of neurons in the primate tongue primary motor cortex during swallowing. Journal of Neurophysiology, 78, 1516–1530. Martin, R. E., Kemppainen, P., Masuda, Y., Yao, D., Murray, G. M., & Sessle, B. J. (1999). Features of cortically evoked swallowing in the awake primate (Macaca fascicularis). Journal of Neurophysiology, 82, 1529–1541. Martin, R. E., MacIntosh, B. J., Smith, R. C., Barr, A. M., Stevens, T. K., Gati, J. S., et al. (2004). Cerebral areas processing swallowing and tongue movement are overlapping but distinct: A functional magnetic resonance imaging study. Journal of Neurophysiology, 92, 2428–2443. May, A. (2008). Chronic pain may change the structure of the brain. Pain, 137, 7–15. Murray, G. M., & Sessle, B. J. (1992a). Functional properties of single neurons in the face primary motor cortex of the primate. I. Input and output features of tongue motor cortex. Journal of Neurophysiology, 67, 747–758. Murray, G. M., & Sessle, B. J. (1992b). Functional properties of single neurons in the face primary motor cortex of the primate. II. Relations with trained orofacial motor behavior. Journal of Neurophysiology, 67, 759–774. Murray, G. M., Lin, L. D., Moustafa, E. M., & Sessle, B. J. (1991). Effects of reversible inactivation by cooling of the primate face motor cortex on the performance of a trained tongue-protrusion task and a trained biting task. Journal of Neurophysiology, 65, 511–530. Narita, N., Yamamura, K., Yao, D., Martin, R. E., & Sessle, B. J. (1999). Effect of functional disruption of the lateral pericentral cerebral cortex on primate swallowing. Brain Research, 824, 140–145. Neafsey, E. J. (1986). The organization of the rat motor cortex: A microstimulation mapping study. Brain Research, 396, 77–96. Nicolelis, M. A. L., & Lebedev, M. A. (2009). Principles of neural ensemble physiology underlying the operation of brain— Machine interfaces. Nature Reviews. Neuroscience, 10, 530–540. Nordstrom, M. A. (2007). Insights into the bilateral cortical control of human masticatory muscles revealed by transcranial magnetic stimulation. Archives of Oral Biology, 52, 338–342. Okada-Ogawa, A., Suzuki, I., Sessle, B. J., Chiang, C. Y., Salter, M., Dostrovsky, J., et al. (2009). Astroglia in medullary dorsal horn (trigeminal spinal subnucleus caudalis) are involved in trigeminal neuropathic pain mechanisms. Journal of Neuroscience, 29, 11161–11171. Robinson, P. P., Boissonade, F. M., Loescher, A. R., Smith, K. G., Yates, J. M., Elcock, C., et al. (2004). Peripheral mechanisms for the initiation of pain following trigeminal nerve injury. Journal of Orofacial Pain, 18, 287–292. Romaniello, A., Cruccu, G., McMillan, A. S., ArendtNielsen, L., & Svensson, P. (2000). Effect of experimental pain from trigeminal muscle and skin on motor cortex excitability in humans. Brain Research, 882(1–2), 120–127.
Sanes, J. N., & Donoghue, J. P. (2000). Plasticity and primary motor cortex. Annual Review of Neuroscience, 23, 393–415. Sessle, B. J. (2000). Acute and chronic craniofacial pain: Brainstem mechanisms of nociceptive transmission and neuroplasticity, and their clinical correlates. Critical Reviews in Oral Biology and Medicine, 11, 57–91. Sessle, B. J. (2006). Mechanisms of oral somatosensory and motor functions and their clinical correlates. Journal of Oral Rehabilitation, 33, 243–261. Sessle, B. J. (2009a). Orofacial motor control. In: L. Squire (Ed.), Encyclopedia of Neuroscience, (Vol. 7, pp. 303–308). Oxford: Academic Press. Sessle, B. J. (2009b). Sensation from the face. In: L. Squire (Ed.), Encyclopedia of Neuroscience, (Vol. 8, pp. 585–592). Oxford: Academic Press. Sessle, B. J., Yao, D., Nishiura, H., Yoshino, K., Lee, J.-C., Martin, R. E., et al. (2005). Properties and plasticity of the primate somatosensory and motor cortex related to orofacial sensorimotor function. Clinical and Experimental Pharmacology & Physiology, 32, 109–114. Sessle, B. J., Adachi, K., Avivi-Arber, L., Lee, J., Nishiura, H., Yao, D., et al. (2007). Neuroplasticity of face primary motor cortex control of orofacial movements. Archives of Oral Biology, 52, 334–337. Sessle, B. J., Klineberg, I., & Svensson, P. (2009). A neurophysiological perspective on rehabilitation with oral implants and their potential side effects. In A. Jokstad (Ed.), Osseointegration and Dental Implants (pp. 333–334). Copenhagen: Blackwell Munksgaard. Svensson, P., & Graven-Nielsen, T. (2001). Craniofacial muscle pain: Review of mechanisms and clinical manifestations. Journal of Orofacial Pain, 15(2), 117–145. Svensson, P., Romaniello, A., Wang, K., Arendt-Nielsen, L., & Sessle, B. J. (2006). One hour of tongue-task training is associated with plasticity in corticomotor control of the human tongue musculature. Experimental Brain Research, 173, 165–173. Yamamura, K., Narita, N., Yao, D., Martin, R. E., Masuda, Y., & Sessle, B. J. (2002). Effects of reversible bilateral inactivation of face primary motor cortex on mastication and swallowing. Brain Research, 944, 40–55. Yan, C., Ye, L., Zhen, J., Ke, L., & Gang, L. (2008). Neuroplasticity of edentulous patients with implantsupported full dentures. European Journal of Oral Sciences, 116, 387–393. Yao, D., Yamamura, K., Narita, N., Martin, R. E., Murray, G. M., & Sessle, B. J. (2002). Neuronal activity patterns in primate motor cortex related to trained or semiautomatic jaw and tongue movements. Journal of Neurophysiology, 87, 2531–2541. Yildiz, N., Yildiz, S., Ertekin, C., Aydogdu, I., & Uludag, B. (2004). Changes in the perioral muscle responses to cortical TMS induced by decrease of sensory input and electrical stimulation to lower facial region. Clinical Neurophysiology, 115, 2343–2349.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 6
Motor planning of locomotor adaptations on the basis of vision: The role of the posterior parietal cortex Daniel S. Marigold{, Jacques-Etienne Andujar{, Kim Lajoie{ and Trevor Drew{,* { {
Groupe de Recherche sur le Système Nerveux Central, Département de Physiologie, Université de Montréal, Montréal, Québec, Canada Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, Canada
Abstract: In this chapter, we consider the contribution of the posterior parietal cortex (PPC) to obstacle avoidance behavior and we define a model that identifies the major planning processes that are required for this task. A key aspect of this planning process is the need to integrate information concerning the obstacle, obtained from vision, together with an estimation of body and limb state. We suggest that the PPC makes a major contribution to this process during visually guided locomotion. We present evidence from lesion and single unit recording experiments in the cat that are compatible with this viewpoint. Keywords: locomotion; posterior parietal cortex; vision; gait modification; internal model; state estimation; time to contact; motor planning.
et al., 2007; Patla, 1997; Sherk and Fowler, 2001a), (2) aid in determining self-motion and object motion, which facilitates heading direction (Britten, 2008; Logan and Duffy, 2006; Prokop et al., 1997; Sherk and Kim, 2002; Warren et al., 2001), (3) control foot placement (Drew, 1993; Drew et al., 2008; Hollands et al., 1995; Lajoie and Drew, 2007; Marigold and Patla, 2007; Patla and Greig, 2006; Sherk and Fowler, 2001a), and (4) provide visual exproprioception, that is, visual input of the body relative to the environment
Introduction Vision provides a wealth of information that can be extracted from the environment by the nervous system and used to guide locomotion. Related to this role, vision serves to (1) detect and define the characteristics of objects and ground terrain in the environment (Marigold *Corresponding author. Tel.: þ1-514-343-7061; Fax: þ1-514-343-6113 DOI: 10.1016/B978-0-444-53825-3.00011-5
83
84
(Marigold and Patla, 2008; Rietdyk and Rhea, 2006). In this chapter, we focus on the task of avoiding an obstacle in the travel path. Such a task inherently incorporates all four of these aspects of visual information. First, we will attempt to define the requirements for successful planning and execution of stepping over an obstacle. Second, we propose a model that integrates these processes. And third, we provide experimental evidence from neuronal recordings that is compatible with this model. A key cortical area we concentrate on in this latter section is the posterior parietal cortex (PPC). The PPC is often implicated in motor planning during saccade or reach tasks in nonhuman primates (Andersen and Buneo, 2002; Duhamel et al., 1992; Kalaska, 1996; Mountcastle et al., 1975) and more recently has been suggested to play a role in the motor planning of a gait modification during locomotion in the cat (Andujar et al., 2010; Drew et al., 2008; Lajoie and Drew, 2007). Here, we suggest that the role of the PPC is intimately involved with integrating an estimation of the location of the body and position of the limbs (using visual and proprioceptive input) with visual input regarding time to contact (TTC) with an obstacle to successfully perform an avoidance behavior.
additional requirement to determine the relative speed and the rate of closure of the gap between the subject and the object. Based on the initial analysis of the environment, a series of decisions need to be made including whether to avoid the obstacle or to step over it. In either case, we need to adjust our gait so that we can perform the necessary avoidance maneuver. This will normally require that the step length be initially adjusted so that the foot is placed in a position before or to the side of the obstacle to allow the final avoidance maneuver. In the case of a moving obstacle, this is particularly critical. Related to this is the requirement to initiate the gait modification at the appropriate time. When the decision is made to step over the obstacle, the essential limb trajectory must be determined and executed according to the extracted obstacle attributes and current limb position. Lastly, in the case of a quadruped, the animal has to coordinate all four limbs to step over the obstacle without hitting it, an especially difficult task with a moving obstacle. The sequence of movements to successfully step over the obstacle is illustrated in Fig. 1a.
Consideration of the behavioral processes involved in planning a step over an obstacle
Several models have been proposed to explain the control of goal-directed movements such as reaching and grasping (Desmurget and Grafton, 2000; Gritsenko et al., 2009; Kawato, 1999; Miall and Wolpert, 1996; Shadmehr and Krakauer, 2008; Wolpert and Ghahramani, 2000). Fundamental to these models is the ability of the nervous system to exploit forward models such that the sensory consequences of one's actions can be predicted based upon an efference copy of the motor command. Through integration of this prediction with reafference from various sensory systems during movement, it is then possible to form a belief about the state of the limbs and body (i.e., state estimation). This process forms the basis for rapid online corrections during reaching/grasping
The environment in which we live and travel is full of obstructions that require us to modify our gait to step around or over such obstacles in our path. In most cases, we accomplish these gait modifications effortlessly despite the fact that the processes involved in these activities can be quite complex. We can summarize these processes as follows: First, we need to use vision to detect a hazard in our path and to determine the properties of that obstacle. These attributes might include its size and shape and its spatial location relative to the body or the limbs. In the case of a moving object (see below), there is the
Motor planning: A model for obstacle avoidance during locomotion
85 (a) Object attributes and motion
Advance paw placement
Step over obstacle
Storage of obstacle properties
Forelimb–hindlimb coordination
Hindlimb–hindlimb coordination
(b)
Comparator
Motor command Efference Copy
Error Motor plan
Correction
Controller
Predicted sensory consequences Limb state estimation
Limb/body state change
Forward model
Vision
Sensory feedback (proprioceptive, vestibular)
Self-motion estimation Efference copy Temporal/spatial integration Task selection
TTC
Gap closure
Vision Hazard detector
Object attributes
Context
Fig. 1. Motor planning processes involved in an obstacle avoidance task. (a) Behavioral planning processes involved in stepping over an obstacle. Obstacle attributes (such as size and shape) and object velocity must be determined in order to plan paw placement in front of the obstacle and limb trajectory over the obstacle. Subsequently, a representation of the spatial relationship of the obstacle with the position of the limbs must be combined with obstacle motion characteristics to successfully coordinate the hindlimbs to step over the obstacle in the absence of vision. (b) Proposed model of the planning processes for obstacle avoidance. Central to this model is the temporal/spatial integration, which combines self-motion and obstacle time-tocontact (TTC) signals (temporal aspect) with an estimate of limb state (spatial aspect). The result of this integration forms the basis of the motor plan. In particular, this signal provides details on the gap closure of the animal and obstacle to allow the tasks of paw placement in front of the obstacle, and step execution over the obstacle to be triggered. See text for details.
86
and is summarized in the top portion of Fig. 1b (enclosed within the dotted line box). The models of reaching/grasping concentrate primarily on the motion of the arm from a static position. Thus, the motor plan can be made offline and without the need to estimate motion. However, in the situation examined in this chapter, there are additional processes such as integrating the state of the moving body and limbs relative to a moving obstacle. In this chapter, we have attempted to illustrate in a model the complex integration and planning processes involved in stepping over an obstacle that we outlined in the previous paragraphs. We identify three processes that may also be applicable to reaching/grasping movements toward moving objects.
Analysis of the obstacle and its attributes The first requirement is to detect the obstacle using visual information. This hazard detector provides information about objects in the environment that require a change of the base locomotor activity (Marigold, 2008). In essence, it identifies the context of the obstruction and it primes the system for a future gait modification. As the animal moves toward the obstacle, or vice versa, one then needs to determine both its spatial (where it is and its size) and its temporal (when it will be encountered) attributes. One mechanism by which this may be accomplished is through processing of optic flow created during self- and object motion. In fact, extensive research has shown that animals are capable of extracting relevant task details from optic flow, defined as the pattern of motion (or flow of textures) of surfaces and objects in the environment generated during movement (Britten, 2008; Duffy and Wurtz, 1991, 1995, 1997; Gibson, 1958; Gu et al., 2008; Logan and Duffy, 2006; Rauschecker et al., 1987; Sherk and Fowler, 2001a,b; Sherk and Kim, 2002; Sun et al., 1992). This optic flow may provide the animal with
object motion, self-motion, and TTC. For instance, Sun et al. (1992) have demonstrated that gerbils use optic flow to control deceleration during locomotion to a target. In this study, they measured the velocity of the gerbils running toward a target, which either expanded (thus providing the illusion of the target approaching) or shrank (providing the illusion of the target retreating). Whereas the expanding target resulted in the animals reducing their velocity earlier than control trials, the shrinking target delayed deceleration. In the context of obstacle avoidance, TTC may represent the best means of timing the required gait modification. Two different mechanisms for calculating TTC may work alone or in concert: changes in tau or binocular retinal disparity and vergence cues. Tau is the ratio between retinal image size at a given instant and the rate of expansion of the image, which is extracted from optic flow (Lee, 1976). For example, Warren et al. (1986) have shown that tau can be used to control the vertical impulse applied to the ground by the leg to ensure accurate foot placement onto targets during treadmill running. Research has demonstrated that there are classes of neurons in the optic tectum and nucleus rotundus of the pigeon and area 7a of the monkey that encode various optical variables related to signaling TTC to looming objects including tau (Merchant et al., 2001, 2004; Siegel and Read, 1997; Sun and Frost, 1998; Wang and Frost, 1992; Wu et al., 2005). In addition, neurons in the lateral intraparietal (LIP) cortex have been shown to encode both vergence and retinal disparity, with vergence angle (representing fixation distance) modulating the retinal disparity activity (Genovesio and Ferraina, 2004). Similar results have also recently been shown with neurons in the parietal reach region (PRR) (Bhattacharyya et al., 2009). By relating vergence and disparity signals over time as the object approaches it may be possible to determine TTC (Regan, 2002; Wann, 1996). Several brain regions have been implicated in processing optic flow information including the
87
extrastriate visual cortex and PPC. In the cat, Sherk and Kim (2002) have demonstrated neurons in the lateral suprasylvian visual area that respond to optic flow simulating turns during locomotion (see also Rauschecker et al., 1987). In fact, research from Sherk and Fowler (2001b) has also shown how optic flow may be used during locomotion. Specifically, low-frequency strobe lighting, which eliminated image motion, resulted in significantly decreased accuracy in cats trained to avoid stepping on small objects in the travel path during locomotion. This suggests that optic flow provides critical details for obstacle avoidance. In the monkey, dorsal medial superior temporal cortex (MSTd) neurons are sensitive to optic flow signaling stimulus motion in depth (Duffy and Wurtz, 1991) and heading direction (Duffy and Wurtz, 1995; Gu et al., 2006). Both MSTd neurons and medial temporal (MT) cortical neurons are also sensitive to the speed of visual motion (Duffy and Wurtz, 1997; Maunsell and Van Essen, 1983). Interestingly, Gu et al. (2008) have found that MSTd neuronal thresholds for heading discrimination with both optic flow (visual input) and passive motion in the dark (vestibular input) cues combined were significantly correlated with the monkey's decision of perceived heading direction. In addition, psychophysical thresholds and MSTd neuronal thresholds showed parallel decreases for the combined condition compared to the cues presented independently suggesting increased sensitivity. Furthermore, microstimulation of area MST in monkeys during a heading discrimination task resulted in large biases of the monkey's perception of heading direction (Britten and van Wezel, 1998, 2002). Of particular interest here is the recent work of Logan and Duffy (2006) who have shown MSTd neurons in the monkey encode both self-motion from optic flow and object motion (81% of sampled neurons) and facilitate estimation of 3D heading direction. In this study, object motion was superimposed on optic flow simulating the same or opposite heading direction. Whereas in the
same direction condition MSTd neurons were more sensitive to optic flow, in the opposite condition they were more sensitive to object motion. Furthermore, population responses were particularly sensitive to object motion toward central vision. Therefore, relevant self- and object motion may be combined at this stage for later integration.
State estimation As part of the assessment of when to initiate the gait modification, it is imperative the animal estimates the relative position of its limbs and body. It has been proposed that this state estimation relies on a forward model that utilizes an efference copy of the motor command to predict future sensory consequences (Desmurget and Grafton, 2000; Miall and Wolpert, 1996; Wolpert and Ghahramani, 2000; Wolpert et al., 1995). This prediction is integrated with actual sensory feedback stemming from proprioceptors, the vestibular apparatus, and visual information of the limbs. During locomotion, the fact that either the animal and/or the object moves dictates a need for an estimate of self-motion of the whole body that must be related to obstacle characteristics such as the TTC and obstacle size/ dimensions. In the model, we suggest that there must be an integration process that combines the temporal information obtained from visual input (e.g., TTC and self-motion) with a spatial estimate of body and limb position to generate a signal providing information related to the gap closure of the animal and object. The gap closure between the obstacle and animal changes continuously as the animal and/or obstacle moves. Given the need to integrate multiple sources of sensory information and the fact that the incoming sensory feedback may originate in effectorspecific coordinates or reference frames (e.g., eye-centered for vision, body-centered for limb proprioception, head-centered for vestibular input), a critical step involved in state estimation is coordinate transformation (Andersen and
88
Buneo, 2002). Here, a single common reference frame must be derived from sensory input with different reference frames. Further, the frame of reference must be specific to the required gait modification. There is no doubt that multiple cortical and subcortical structures are involved in producing the appropriate motor plan for a given situation and many studies have tried to associate specific functions with specific parts of the brain (Andersen and Buneo, 2002; Buneo and Andersen, 2006; Drew et al., 2008; Kalaska, 1996; Kawato, 1999; Shadmehr and Krakauer, 2008). One structure that has consistently been implicated in both coordinate transformation and state estimation is the PPC. The idea that the PPC is involved in transforming between different coordinate reference frames is based, in large part, on work in the nonhuman primate. Early research demonstrated that neurons in different subregions of the PPC encode properties specific to the task (Andersen and Mountcastle, 1983; Batista et al., 1999; Colby et al., 1996; Gnadt and Andersen, 1988; Lacquaniti et al., 1995). In fact, Buneo et al. (2002, 2008) have recently shown that PRR neurons simultaneously encode target locations during reaches in eye and limb coordinates suggesting that PPC neurons may directly transform target locations for reaches between the two reference frames. Furthermore, neurons in LIP and area 7a carry head position information in their discharge activity as shown by changes in this activity depending on the head angle, which may also act to facilitate coordinate transformation (Brotchie et al., 1995; Buneo and Andersen, 2006; Snyder et al., 1998). Evidence for the PPC's involvement in state estimation of the limb stems largely from studies examining online control of reaching and grasping movements. In humans, disruption to the PPC from transcranial magnetic stimulation (Desmurget et al., 1999; Rice et al., 2006; Tunik et al., 2005) or lesions (Gréa et al., 2002; Pisella et al., 2000; Wolpert et al., 1998) results in impaired
online corrections of reaching trajectories and grasping aperture and orientation following target/object perturbations. In monkeys, LIP neurons show predictive updating during instructed delay saccade tasks such that these neurons increase their discharge in expectation of the appearance of a saccade target in their receptive field (Duhamel et al., 1992; Gnadt and Andersen, 1988). Recently, Mulliken et al. (2008) have provided some of the strongest evidence for the involvement of PPC neurons in forward modeling. In this study, monkeys maintained a central fixation while performing center-out and obstacle avoidance tasks using a joystick that controlled a cursor on a screen in front. Neurons in the PRR/area 5 were found to encode both the static goal angle of the target (representing static target direction) and dynamic movement angle of the cursor (representing the dynamic estimate of the state of the cursor). Furthermore, the temporal dynamic encoding of these neurons showed a large population ( 50%) with optimal lag times with the cursor of 0 30 ms. The lag times of these neurons are outside the range of encoding sensory feedback (i.e., at least <30 ms) and encoding the motor command (i.e., > 90 ms) and thus may best be explained as encoding the current movement angle. This implies a forward model that makes use of the efference copy to move the joystick to estimate the current position or state of the cursor.
Task selection and motor plan We propose that the result of the integration of the TTC information and the state estimation is a signal that relates the obstacle collision time (temporal aspect) to the position of the limbs (spatial aspect) in order to plan the necessary gait modifications. The way in which this signal affects the motor plan depends on context, and specifically on the magnitude of the gap between the animal and object. For example, when the cat is still distant from the
89
obstacle (as in the first diagram in Fig. 1a) the task for the animal is to place the paw in an appropriate manner to prepare for the step over the obstacle. As the cat approaches the obstacle this window in which the paw is placed becomes smaller (Lajoie and Drew, 2007) in a similar way to that observed in long jumpers (Lee et al., 1977; Montagne et al., 2000). When the distance between the cat and the obstacle decreases to a critical value (as in the second diagram in Fig. 1a) then the motor strategy is changed from one mode to another to plan the modified limb trajectory required to step over the obstacle. Effectively, we propose that the temporal/spatial integration signal is used to trigger the gait modification when a critical threshold is reached through the use of a task selection module. This same part of the circuit must also be used to modify the movements of the hindlimbs. When the cat steps over the obstacle using the standard strategy (Drew et al., 2008; Lajoie and Drew, 2007), in which each limb in turn passes over the obstacle, the animal must form an estimate of the spatial and temporal advance of the obstacle and store this in memory. From a neural perspective, the signals may be visual feedforward in nature in that object speed is calculated and held on to. It is also possible that the signals stem from forelimb proprioceptive feedback. Alternatively, the motor efference copy of the command for the step over obstacle of the forelimbs is preserved until the hindlimbs step over (see, e.g., McVea et al., 2009). However, in some situations, the order of the hindlimbs may be altered (e.g., the double step strategy of Lajoie and Drew, 2007). In this case, there is the need for a further modification of the motor plan so that the cat selects the relative order of the hindlimbs after the forelimbs have passed over the obstacle. There is no evidence in the literature to show which neural structures are involved in these processes during locomotion. However, as we detail in the following sections, we suggest that the PPC makes a major contribution to the determination of the motor plan that needs to be selected.
Contribution of the PPC to the control of visually guided locomotion As is evident from the focus of this volume, most studies of locomotion concentrate on the mechanisms responsible for the generation of the basic locomotor rhythm. Studies on supraspinal control mechanisms have normally placed an emphasis on either the initiation of locomotion or its step-by-step regulation (reviewed in Armstrong, 1986; Jordan et al., 2008; Rossignol, 1996; Rossignol et al., 2006). Although some studies, including many from this laboratory, have specifically examined the contribution of various structures to the control of visually guided locomotion (Amos et al., 1990; Armstrong and MarpleHorvat, 1996; Beloozerova and Sirota, 1993; Drew, 1993; Drew et al., 1996, 2008), only a few have placed any emphasis on the specific contribution of these structures to the planning of voluntary gait modifications (Drew et al., 2008; Lajoie and Drew, 2007; Marple-Horvat et al., 1998; McVea and Pearson, 2009). With respect to the specific contribution of the PPC to the control of locomotion, only Beloozerova and Sirota (2003) have published data on the activity of neurons in the PPC during locomotion. These authors showed that cells around the ansate sulcus (and probably mostly in area 5b) were modulated during locomotion and substantially increased their discharge when the cats stepped over barriers or walked from rung to rung of a horizontal ladder. However, while these experiments showed a contribution to the control of visually guided locomotion they were not designed to address the issue of a possible role in planning these locomotor movements. Two related questions then arise. First, does the PPC play a similar role in the planning of visually guided movements in the cat as in the primate? Second, during locomotion, does the PPC provide the specific information required for the planning of obstacle avoidance as proposed in our model? We have recently addressed this issue of the contribution of the PPC to the planning of visually
90
guided locomotion in our laboratory by using a task in which cats were trained to step over an obstacle that advanced toward them on a moving belt. The obstacle was visible to the cat for a minimum of five step cycles before the step over the obstacle ensuring that sufficient visual information was available to plan the appropriate motor response. In initial recordings, the speed of the obstacle and the treadmill on which the cat walked was the same; we refer to this as the matched task. After recording for 10 min, we modified the speed of advance of the obstacle so that it was slower than the speed of the treadmill on which the cat walked; we refer to this as the visual dissociation task. Slowing the speed of the obstacle had several effects. First, it dissociated the information available to the cat from its own self-motion and the visual information available from the advancing obstacle. This ensured that the cat had to integrate information from both sources to successfully complete the task. Second, the steps before the obstacle were more variable than when the treadmill and obstacle were matched thus facilitating temporal correlation of cell activity with behavior. Third, when the obstacle speed was substantially reduced, it resulted in a major change of strategy such that the cat was forced to modify the overall gait pattern and to make a stutter or double step with the hindlimbs to compensate for the increased time that the obstacle took to pass under the body (Drew et al., 2008; Lajoie and Drew, 2007). In terms of more natural locomotor movements, the challenges of the task may be compared to those involved in running up to hit a moving soccer ball.
Lesion studies In our initial studies, we examined the effects of lesion of either the medial or more lateral regions of the PPC, encompassing areas 5b and to a lesser extent, area 5a (Lajoie and Drew, 2007). The results from this study emphasized the importance of the PPC in locomotor tasks requiring visual guidance and provide support for the
PPC as a site of integration of exteroceptive visual information with estimates of the body and limb state. In all three cats used in this study, there was very little, if any, sign of a deficit during normal treadmill locomotion. However, during the steps over the obstacle, the cats frequently made contact with the obstacle, sometimes on the front edge and sometimes on the back edge as the limb was being placed on the other side. These deficits were larger in the visual dissociation task. An example from one cat is shown in Fig. 2. In this cat, the lesion was restricted to the caudal portion of the ansate sulcus and was centered on the lateral sulcus (Fig. 2a). After the lesion, the cat frequently hit the obstacle, especially in the visual dissociation condition (Fig. 2b). Deficits were still evident up to 3 weeks after the lesion. In our detailed analyses of these deficits, we found that a major determinant of whether the cats hit the obstacle, and whether they hit it on the front or the back, was the position in which the paws were placed in front of the obstacle in the step preceding the step over. In unlesioned cats, the location in which the paw was placed in front of the obstacles was very tightly localized. In trials in which the lesioned cats cleared the obstacle, the paws were placed at a distance that was comparable to that in the intact cats. This is shown in Fig. 2c (no hits) for the same cat as illustrated in Fig. 2a and b. In trials, in which they hit the front edge, the paw was placed closer than in the control and in trials in which they hit the backside, it was placed farther away. This speaks to an important function of the PPC in integrating the information about the approaching obstacle with information about the current position of the body and the limbs. With respect to Fig. 1, this might suggest that the PPC was using the available information to calculate gap closure. In the lesioned animal, we would suggest there was an incomplete integration of spatial and temporal information that resulted in inappropriate positioning of the paw in front of the advancing obstacle.
91 (a)
(b) Contralateral limb leads
LS
CoS SSS
% of deficits
AS
CS
M R
C L
100 90 80 70 60 50 40 30 20 10 0
0.5/0.5 0.5/0.35
1
2
3
(c)
7
8
9
10
15
20
Days post lesion
5 mm
No hits 0.5/0.5 Contralateral limb leads
27.0 cm
50
Hits front 20.7 cm coFL1
Hits back
Position of RFL1 (cm)
coFL1
45
Prelesion
40
Postlesion
N = 47
* N=9
35 30
N = 48 N=3
25 20 15 10 5
38.4 cm
0 coFL1
No hits
Hits front
Hits back
Fig. 2. Effects of lesions of the posterior parietal cortex on voluntary gait modifications. (a) Schematic representation of the lesion (shaded region) of the PPC in cat PCM3. (b) Histogram showing the percentage of times that the cat hit the obstacle at different times after the lesion. The graph shows the deficits in both the matched (0.5/0.5) and the visual dissociation (0.5/0.35) condition for the lead condition (limb contralateral to the lesion is the first to step over the obstacle). (c) Distance of the paw from the obstacle for the situation when the cat successfully steps over the lesion (no hits), hits the front of the obstacle, or hits the back (far side) of the obstacle. Data are shown schematically to the left. The histogram plots the distance of the paw from the obstacle in the step cycle before the step over the obstacle. The light gray bar and the dotted line indicates the prelesion control value (¼26.8 cm). Values on the schematic display provide the average distances for the paw following the lesion for each condition. Abbreviations: AS, ansate sulcus; CoS, coronal sulcus; CS, cruciate sulcus; LS, lateral sulcus; SSS, suprasylvian sulcus (modified from Lajoie and Drew, 2007).
A similar spatial deficit has recently been shown by McVea et al. (2009). These authors trained cats to stop with the forelimbs and hindlimbs straddling an obstacle; the obstacle was then lowered without the cat's knowledge. After a variable period in this maintained posture, the cats
were allowed to continue their forward progress. The authors found that even after periods of up to 10 min the cats still raised the hindlimb over the now absent obstacle (Mcvea and Pearson, 2006). They subsequently showed that lesion of area 5 (primarily area 5b) resulted in a loss of
92
the ability to perform this task (McVea et al., 2009). Instead, if the cats were prevented from progressing for periods of longer than 1.5 min, the hindlimb trajectory was always insufficient to clear the (now removed) obstacle. This result suggests a further role for the PPC in working memory.
Single-unit recording studies In a second series of experiments, we used single unit recordings from areas 5a and 5b of the PPC to determine the neuronal mechanisms contributing to the planning of these gait modifications. Cats were trained in the same tasks as used in our lesion experiments and a recording chamber was placed over the ansate sulcus. Neuronal recordings were made from neurons in layer V of the cortex, including some antidromically identified as projecting to the cerebral peduncle. As in the study of Beloozerova and Sirota (2003), we found that some cells were rhythmically modulated during unobstructed treadmill locomotion and that there was a substantial increase in discharge activity when the cats stepped over the obstacle (Andujar et al., 2010; Lajoie et al., 2010). We identified two major populations of cells. The first population showed a modification of their discharge activity compared to the unobstructed situation only with respect to the step over the obstacle by the forelimb. The second population had a component of the discharge related to the passage of the hindlimbs. Within the population of cells related to the forelimbs, we found that 102/121 (84%) of the neurons that we recorded in the PPC showed an increased level of activity when the cat stepped over the obstacle. We identified two subpopulations. The first subpopulation (41/102, 40%) of neurons showed a brief, phasic pattern of discharge and showed increased activity only during the step over the obstacle: we refer to these as Step-related neurons. An example of
one such cell is illustrated in Fig. 3a during the condition in which the limb contralateral to the recording site was the first to step over the obstacle (we refer to this as the lead condition). These Step-related neurons showed properties that are similar to those that we have previously described for motor cortical neurons during an identical task (Drew, 1993; Drew et al., 1996). In particular, they showed discrete phasic bursts of activity at different times during the swing phase of the modified step. We suggest that these cells contribute to the modulation of the activity of the groups of synergistic muscles that are responsible for defining the limb trajectory. These Step-related neurons were mostly recorded in the rostral ansate sulcus in area 5a. Many of them had cutaneous receptive fields and few of them responded to a looming stimulus. As such, these cells have properties very similar to those recorded in the dorsal region of area 5 in the primate (Chapman et al., 1984; Kalaska, 1996; Mountcastle et al., 1975). Although cells in this region have been suggested to have primarily a sensory function, it should be emphasized that the discharge pattern of these Step-related neurons was rarely explained simply on the basis of the peripheral receptive fields. This suggests that their discharge was determined on the basis of central or corollary inputs. It is possible that these cells may play a role in the online correction of the limb during the step over the obstacle. As such their major function in Fig. 1 would be state estimation of the moving limb, as proposed in the original models of limb control (i.e., the region within the dotted box in Fig. 1). The second subpopulation (61/102, 60%) discharged in advance of the step over the obstacle (Step-advanced cells). Cells in this population, by our definition (Andujar et al., 2010), began to discharge a minimum of 200 ms before the step over the obstacles, although many of them discharged substantially earlier than this. Indeed, many cells began to modify their discharge activity at least 1 step cycle (2 steps) before the obstacle. In averaged displays, synchronized to the onset of the
93
(a)
coBr N = 54
200
coBr N = 10
Spikes/s
200
Matched: lead condition
Control
Control + matched (lead condition) 200
0 0
0
Pcm2Tr33G2
coBr
iBr 200 ms
200 ms
(b)
Spikes/s
80
coBr N = 111
80
coBr N = 24
80
0
Pcm5Tr52D1
0
0
coBr
iBr 200 ms
200 ms
Fig. 3. Examples of a Step-related (a) and a Step-advanced (b) cell during the gait modification. For each cell, the two left-most traces illustrate a perievent histogram and raster display triggered on the onset of activity in the contralateral brachialis (coBr, straight vertical line) for both the unobstructed condition (left display) and the step over the obstacle (middle display). The staggered vertical lines indicate the end of the period of activity in the coBr. The right-most trace shows the EMG activity during the step over the obstacle (thick line) superimposed on the control activity (thin line). Each step cycle is synchronized to the onset of activity in the coBr and each step cycle is normalized to the averaged value. Data are shown for the two steps before and one step after the step over the obstacle (modified from Andujar et al., 2010).
flexor muscle activity during the modified step, these cells frequently showed a ramp increase of discharge activity prior to the modified step, as in the example in Fig. 3b. In some cells, this discharge activity continued as the limb passed over the obstacle, as in the cell in Fig. 3b, while in others it ceased at the time that the cat initiated the step over the obstacle. Such Step-advanced cells were rarely recorded in our previous studies in the motor cortex (referred to previously as early cells, Drew, 1993) where they formed only 16% of our database in that study. We suggest that these Step-advanced cells are implicated in the planning of the gait modification. Consideration of their discharge characteristics
with respect to the planning processes outlined in Fig. 1 allows us to make several suggestions as to which ones they may be involved in and others to which they are unlikely to contribute. First, it is unlikely that these cells function as a hazard detector as the modulation of the discharge only occurs relatively late (3–4 step cycles) after the obstacle becomes visible to the cat. Moreover, the discharge is time-locked to the step over the obstacle. It is also unlikely that the discharge is providing information on attributes of the obstacle such as its size and shape as cell discharge was not significantly different for obstacles of two different sizes (Andujar et al., 2010). As to what the discharge does signal, one intriguing possibility is that it
94
provides information on gap closure as a result of the integration of the temporal and spatial properties of the advancing obstacle. Cells in the cat PPC receive both proprioceptive and visual inputs (Andujar et al., 2010; Dubner, 1966; Garraghty et al., 1987; Thompson et al., 1963) and many respond preferentially to looming stimuli (Andujar et al., 2010; Beloozerova and Sirota, 2003). They might therefore integrate the information about self-motion and TTC together with the state of the limbs to provide the information necessary to determine either where to place the paws in front of the obstacle or when to initiate the gait modification. The latter role in particular would be compatible with some of the evidence from primate experiments. For instance, Maimon and Assad (2006a,b) have demonstrated that LIP and area 5 neurons increase their discharge activity prior to a reach and they attain a critical, constant, firing threshold at the onset of a self-timed, rather than triggered, hand movement. Furthermore, LIP neurons in the monkey have been found to represent elapsed time of remembered durations (Leon and Shadlen, 2003). With the Step-related and the Step-advanced cells taken together, virtually all cells that showed significantly modified activity in the lead condition also showed modified activity in the trail condition (94/102). In this latter condition, the contralateral limb is the second to pass over the obstacle. For some of these cells, the discharge activity was best related to activity in the contralateral limb during both the lead and the trail condition. We refer to these as limb-specific cells (Andujar et al., 2010). This is similar to cells recorded in the motor cortex (Drew, 1993; Drew et al., 1996). However, in many cells (57/94, 61%), the activity was always related to the lead limb, regardless of whether this limb was contralateral or ipsilateral to the recording site: we refer to these as limb-independent cells. One example of a limb-independent cell is shown in Fig. 4 (same cell as in Fig. 3b). As described above, in the lead condition, discharge frequency in this cell began to increase progressively in the step cycle before the step over the obstacle and then increased
again at the onset of the period of activity in the coBr, that is, just as the gait modification was initiated in the contralateral, lead, limb (Fig. 4a). In the trail condition, the cell discharge was phase advanced by 0.5 with respect to the activity in the contralateral limb and discharge ended prior to the onset of activity in the coBr (Fig. 4b). However, synchronizing the cell activity to the iBr in the trail condition (Fig. 4c) produced a pattern of activity that showed the same relationship to the iBr in the trail condition as to the coBr in the lead condition (compare Fig. 4c with a). All limb-independent cells showed a similar relationship to the activity of the coClB or coBr in the lead condition and to the iClB or iBr in the trail condition. The presence of so many limb-independent cells suggests strongly that the PPC in one hemisphere contributes to the control of each limb. This places the parietal cortex at a hierarchically higher level of control than the motor cortex, which is related principally to the control of the contralateral limb. Within the population of cells related to the hindlimbs, we also identified two subpopulations of cells. The major subpopulation discharged between the passage of the forelimb and the passage of the hindlimb. We suggest that these cells ensure the coordination between the forelimbs and the hindlimbs as the obstacle passes under the body. In this respect, it is important to note that the cat loses sight of the obstacle as the forelimbs step over it. The cat, therefore, has to maintain a neural representation of the spatial location of the obstacle with respect to the body as well as its temporal rate of progress. This is particularly important in the visual dissociation task in which the speed of the obstacle is reduced with respect to that of the treadmill on which the cat walks. In this task, the obstacle takes relatively longer to pass under the body than in the matched task. In many steps, the cats compensated for the increased time of passage of the obstacle by delaying swing onset or increasing its duration; however, at other times, they completely changed their strategy and adopted the double step strategy (Lajoie and Drew, 2007). In both conditions, the
95
(a)
(b)
Lead condition
Contralateral limb leads 95
(c)
Trail condition
Ipsilateral limb leads
F = −0.55
95
Ipsilateral limb leads
F = −1.03
coBr
95 Unit
Unit
Unit
0
0
N = 24
coBr
F = −0.55 iBr
coBr
0
Trail condition
N = 12
N = 12
coBr
coBr
iBr
iBr
Pcm5Tr52D1
iBr
Lead limb
Lead limb
Lead limb
Fig. 4. Example of a limb-independent cell. (a) Averaged discharge activity of a cell (same as in Fig. 3b) and selected EMG activity during the lead condition when the contralateral limb was the first to step over the obstacle. Note cell discharge ends at the onset of activity in the contralateral cleidobrachialis coBr. (b) the same cell during the trail condition when the contralateral limb is the second to pass over the obstacle. In this display, the activity is still synchronized to coBr. Note in this case that the cell discharge ends prior to the onset of activity in the coBr. (c) Discharge activity in the trail condition but synchronized to the onset of activity in the iBr. The shaded rectangle indicates the step over the obstacle with the contralateral limb in (a)–(c).
discharge activity of the cells in the PPC predicted the activity of the limb. In the example illustrated in Fig. 5a, the cell discharged between the passage of the contralateral forelimb (coFL) and the contralateral hindlimb (coHL) in the matched task (lead condition). Cell discharge began just after the onset of the activity in the coClB and continued until the passage of the coSt. In the trail condition, still in the matched task, the cell now discharged between the period of activity in the iClB and the iSt (Fig. 5b). It was, therefore, limb-independent, both with respect to the forelimbs and the
hindlimbs. In the trail condition, when the obstacle was slowed but the cat still adopted the standard strategy, the cell discharge was prolonged in parallel with the increased delay between the periods of activity of the iClB and the iSt (Fig. 5c). However, when the cat adopted the double step strategy during the visual dissociation task (still in the trail condition), the order of the hindlimbs was reversed so that after the passage of the iFL and the coFL, the iHL was placed in front of the advancing obstacle instead of being brought over it. As such, the coHL was the first hindlimb to be brought over the obstacles. In this situation,
96
Matched task Lead condition
(b)
Spikes/s
Spikes/s
(a) 90 0
coClB
1
Matched task Trail condition 90 0 2
coClB 2
iClB
iClB 3
coSt iSt
1 4
coSt 3
iSt
4
Ctrl (930 n = 12) Stim1 (n = 14)
(c)
(d)
Visual dissociation task
Visual dissociation task Double-step strategy (trail)
Spikes/s
Spikes/s
Standard strategy (trail) 90 0 2
coClB iClB
0 2
coClB
1
iClB 4
coSt iSt
90
3
coSt iSt
1 3 4
Fig. 5. Example of a forelimb–hindlimb cell. (a) Cell discharge in the matched task during the lead condition. (b) Cell discharge in the matched task during the trail condition. (c) Cell discharge in the visual dissociation condition when the cat adopts the standard strategy (trail condition). (d) Cell discharge activity in the visual dissociation condition when the cat adopts the double step strategy (trail condition). In all traces, we show averaged activity of a cell and selected EMGs for two steps before the step over the obstacle and one step after. In (a) and (b), the thin (black) line indicates discharge activity in the unobstructed condition and the thicker (red) line, cell discharge activity in the lead (a) or trail (b) condition. In (c) and (d), the thin (red) line illustrates the activity in the matched task, trail condition (same as in (b)), and the thicker (green) line indicates activity in the visual dissociation task. Numbers inside circles indicate the order in which the limbs step over the obstacle in the relative task and condition. Abbreviations: St, semitendinosus (modified from Lajoie et al., 2010).
many of the cells, as the example illustrated in Fig. 5d, changed their discharge activity in a similar manner. Cell discharge activity still began during the period of activity of the iClB but instead of continuing on until the passage of the iHL (as during the standard strategy) now continued only
until the period of activity in the coSt in the coHL. In other words, cell discharge signaled the period between the step over the obstacle by the lead forelimb and the lead hindlimb in all four situations illustrated in Fig. 5 despite the change in order in the double step strategy.
97
We suggest that the forelimb–hindlimb cells in the PPC are making an estimate of the location of the obstacle with respect to the body and the time at which the hindlimbs should initiate the step over the obstacle. This is conceptually similar to the functions we assigned to those PPC neurons related to the step over the obstacle with the forelimb (preceding paragraphs). However, during the passage of the forelimbs, the cat can use direct visual information to obtain information on gap closure and TTC. In the case of the hindlimbs, this estimate has to be internal and based on the prior visual information. The other subpopulation of cells discharged only between the passage of the two hindlimbs over the obstacle. These cells also predicted the increased delay between the hindlimbs in the visual dissociation task and likewise some showed limb-independent properties. We suggest that they perform an identical function of spatial and temporal estimation to that of the forelimb–hindlimb cells.
Conclusions In this chapter, we have presented a conceptual model defining the planning processes that are required to successfully negotiate a moving obstacle. The model is based on a consideration of obstacle avoidance in the cat but generalizes to other locomotor situations in which gait must be modified on the basis of visual input. There are also clearly aspects of this planning process which are equally applicable to the control of arm movements, particularly during self-motion. The major thesis in this review has been that the PPC contributes to an integration of information obtained from the moving obstacle (i.e., its location and speed of advance) together with information concerning the location of the body and the limbs with respect to that obstacle (obtained from vision and proprioception). This proposition is founded on the known convergence of visual and proprioceptive information in the PPC and particularly by the large body of information concerning the
contribution of the PPC to the control of visually guided arm movements in primates. The information obtained from the lesion and single unit recording studies clearly support this viewpoint. For example, the lesion experiments show that a major deficit following damage to the PPC is a loss of precision in paw placement that leads to an ability to successfully step over the obstacle in every trial. This modification of paw placement would be compatible with a compromised integration of the exteroceptive and proprioceptive components of the inputs to the PPC. The single unit recording studies equally provide support for a contribution of the PPC to the planning of visually guided locomotion and show neuronal properties consistent with some of the specific aspects of this integration process. For example, the ramp increase in discharge observed in our Step-advanced population of neurons is compatible with an integrative process defining gap closure (between subject and object). Such a ramp increase might conceivably provide either spatial or temporal information, or even both. Similarly, the sustained discharge between the passage of the forelimbs and hindlimbs during the step over the obstacle is equally compatible with a contribution of the PPC to an estimation of the spatial and temporal properties of the obstacle as it passes under the body. A more precise determination of the nature of the contribution of the PPC to such a complex process as obstacle avoidance is a challenge. However, the data presented in this chapter point to several issues that are open to further investigation, including the nature of the signal encoded by the Step-advanced neurons. Further investigations, differentially manipulating spatial and temporal variables, are needed to provide insight into these functions.
Acknowledgments D. S. Marigold is funded by NSERC, T. Drew is funded by CIHR, and K. Lajoie is funded by a studentship from FRSQ. Work supported by an
98
operating grant from the CIHR and an infrastructure grant from the FRSQ. References Amos, A., Armstrong, D. M., & Marple-Horvat, D. E. (1990). Changes in the discharge patterns of motor cortical neurons associated with volitional changes in stepping in the cat. Neuroscience Letters, 109, 107–112. Andersen, R. A., & Buneo, C. A. (2002). Intentional maps in posterior parietal cortex. Annual Review of Neuroscience, 25, 189–220. Andersen, R. A., & Mountcastle, V. B. (1983). The influence of the angle of gaze upon the excitability of the light-sensitive neurons of the posterior parietal cortex. The Journal of Neuroscience, 3, 532–548. Andujar, J.-E., Lajoie, K., & Drew, T. (2010). A contribution of area 5 of the posterior parietal cortex to the planning of visually-guided locomotion: Limb specific and limb-independent effects. Journal of Neurophysiology, 103, 986–1006. Armstrong, D. M. (1986). Supraspinal contributions to the initiation and control of locomotion in the cat. Progress in Neurobiology, 26, 273–361. Armstrong, D. M., & Marple-Horvat, D. E. (1996). Role of the cerebellum and motor cortex in the regulation of visually controlled locomotion. Canadian Journal of Physiology and Pharmacology, 73, 443–455. Batista, A. P., Buneo, C. A., Snyder, L. H., & Andersen, R. A. (1999). Reach plans in eye-centered coordinates. Science, 285, 257–260. Beloozerova, I. N., & Sirota, M. G. (1993). The role of the motor cortex in the control of accuracy of locomotor movements in the cat. Journal de Physiologie, 461, 1–25. Beloozerova, I. N., & Sirota, M. G. (2003). Integration of motor and visual information in the parietal area 5 during locomotion. Journal of Neurophysiology, 90, 961–971. Bhattacharyya, R., Musallam, S., & Andersen, R. A. (2009). Parietal reach region encodes reach depth using retinal disparity and vergence angle signals. Journal of Neurophysiology, 102, 805–816. Britten, K. H. (2008). Mechanisms of self-motion perception. Annual Review of Neuroscience, 31, 389–410. Britten, K. H., & van Wezel, R. J. (1998). Electrical microstimulation of cortical area MST biases heading perception in monkeys. Nature Neuroscience, 1, 59–63. Britten, K. H., & van Wezel, R. J. (2002). Area MST and heading perception in macaque monkeys. Cerebral Cortex, 12, 692–701. Brotchie, P. R., Andersen, R. A., Snyder, L. H., & Goodman, S. J. (1995). Head position signals used by parietal neurons to encode locations of visual stimuli. Nature, 375, 232–235.
Buneo, C. A., & Andersen, R. A. (2006). The posterior parietal cortex: Sensorimotor interface for the planning and online control of visually guided movements. Neuropsychologia, 44, 2594–2606. Buneo, C. A., Jarvis, M. R., Batista, A. P., & Andersen, R. A. (2002). Direct visuomotor transformations for reaching. Nature, 416, 632–636. Buneo, C. A., Batista, A. P., Jarvis, M. R., & Andersen, R. A. (2008). Time-invariant reference frames for parietal reach activity. Experimental Brain Research, 188, 77–89. Chapman, C. E., Spidalieri, G., & Lamarre, Y. (1984). Discharge properties of area 5 neurons during arm movements triggered by sensory stimuli in the monkey. Brain Research, 309, 63–77. Colby, C. L., Duhamel, J. R., & Goldberg, M. E. (1996). Visual, presaccadic, and cognitive activations of single neurons in monkey lateral intraparietal area. Journal of Neurophysiology, 76, 2841–2852. Desmurget, M., & Grafton, S. (2000). Forward modeling allows feedback control for fast reaching movements. Trends in Cognitive Sciences, 4, 423–431. Desmurget, M., Epstein, C. M., Turner, R. S., Prablanc, C., Alexander, G. E., & Grafton, S. T. (1999). Role of the posterior parietal cortex in updating reaching movements to a visual target. Nature Neuroscience, 2, 563–567. Drew, T. (1993). Motor cortical activity during voluntary gait modifications in the cat. I. Cells related to forelimbs. Journal of Neurophysiology, 70, 179–199. Drew, T., Jiang, W., Kably, B., & Lavoie, S. (1996). Role of the motor cortex in the control of visually triggered gait modifications. Canadian Journal of Physiology and Pharmacology, 74, 426–442. Drew, T., Andujar, J.-E., Lajoie, K., & Yakovenko, S. (2008). Cortical mechanisms involved in visuomotor coordination during precision walking. Brain Research Reviews, 57, 199–211. Dubner, R. (1966). Single cell analysis of sensory interaction in anterior lateral and suprasylvian gyri of the cat cerebral cortex. Experimental Neurology, 15, 255–273. Duffy, C. J., & Wurtz, R. H. (1991). Sensitivity of MST neurons to optic flow stimuli. I. A continuum of response selectivity to large-field stimuli. Journal of Neurophysiology, 65, 1329–1345. Duffy, C. J., & Wurtz, R. H. (1995). Response of monkey MST neurons to optic flow stimuli with shifted centers of motion. The Journal of Neuroscience, 15, 5192–5208. Duffy, C. J., & Wurtz, R. H. (1997). Medial superior temporal area neurons respond to speed patterns in optic flow. The Journal of Neuroscience, 17, 2839–2851. Duhamel, J.-R., Colby, C. L., & Goldberg, M. E. (1992). The updating of the representation of visual space in parietal cortex by intended eye movements. Science, 255, 90–92.
99 Garraghty, P. E., Pons, T. P., Huerta, M. F., & Kaas, J. H. (1987). Somatotopic organization of the third somatosensory area (SIII) in cats. Somatosensory Research, 4, 333–357. Genovesio, A., & Ferraina, S. (2004). Integration of retinal disparity and fixation-distance related signals toward an egocentric coding of distance in the posterior parietal cortex of primates. Journal of Neurophysiology, 91, 2670–2684. Gibson, J. J. (1958). Visually controlled locomotion and visual orientation in animals. British Journal of Psychology, 49, 182–194. Gnadt, J. W., & Andersen, R. A. (1988). Memory related motor planning activity in posterior parietal cortex of macaque. Experimental Brain Research, 70, 216–220. Gréa, H., Pisella, L., Rossetti, Y., Desmurget, M., Tilikete, C., Grafton, S., et al. (2002). A lesion of the posterior parietal cortex disrupts on-line adjustments during aiming movements. Neuropsychologia, 40, 2471–2480. Gritsenko, V., Yakovenko, S., & Kalaska, J. F. (2009). Integration of predictive feedforward and sensory feedback signals for online control of visually guided movement. Journal of Neurophysiology, 102, 914–930. Gu, Y., Watkins, P. V., Angelaki, D. E., & DeAngelis, G. C. (2006). Visual and nonvisual contributions to three-dimensional heading selectivity in the medial superior temporal area. The Journal of Neuroscience, 26, 73–85. Gu, Y., Angelaki, D. E., & DeAngelis, G. C. (2008). Neural correlates of multisensory cue integration in macaque MSTd. Nature Neuroscience, 11, 1201–1210. Hollands, M. A., Marple-Horvat, D. E., Henkes, S., & Rowan, A. K. (1995). Human eye movements during visually guided stepping. Journal of Motor Behavior, 27, 155–163. Jordan, L. M., Liu, J., Hedlund, P. B., Akay, T., & Pearson, K. G. (2008). Descending command systems for the initiation of locomotion in mammals. Brain Research Reviews, 57, 183–191. Kalaska, J. F. (1996). Parietal cortex area 5 and visuomotor behavior. Canadian Journal of Physiology and Pharmacology, 74, 483–498. Kawato, M. (1999). Internal models for motor control and trajectory planning. Current Opinion in Neurobiology, 9, 718–727. Lacquaniti, F., Guigon, E., Bianchi, L., Ferraina, S., & Caminiti, R. (1995). Representing spatial information for limb movement: Role of area 5 in the monkey. Cerebral Cortex, 5, 391–409. Lajoie, K., & Drew, T. (2007). Lesions of area 5 of the posterior parietal cortex in the cat produce errors in the accuracy of paw placement during visually guided locomotion. Journal of Neurophysiology, 97, 2339–2354. Lajoie, K., Andujar, J.-E., Pearson, K., & Drew, T. (2010). Neurones in area 5 of the posterior parietal cortex in the cat contribute to interlimb coordination during visuallyguided locomotion: A role in working memory. Journal of Neurophysiology, 103, 2234–2254. Lee, D. N. (1976). A theory of visual control of braking based on information about time-to-collision. Perception, 5, 437–459.
Lee, D. N., Lishman, J. R., & Thomson, J. (1977). Visual guidance in the long jump. Athletics Coach, 11, 26–30. Leon, M. I., & Shadlen, M. N. (2003). Representation of time by neurons in the posterior parietal cortex of the macaque. Neuron, 38, 317–327. Logan, D. J., & Duffy, C. J. (2006). Cortical area MSTd combines visual cues to represent 3-D self-movement. Cerebral Cortex, 16, 1494–1507. Maimon, G., & Assad, J. A. (2006a). A cognitive signal for the proactive timing of action in macaque LIP. Nature Neuroscience, 9, 948–955. Maimon, G., & Assad, J. A. (2006b). Parietal area 5 and the initiation of self-timed movements versus simple reactions. The Journal of Neuroscience, 26, 2487–2498. Marigold, D. S. (2008). Role of peripheral visual cues in online visual guidance of locomotion. Exercise and Sport Sciences Reviews, 36, 145–151. Marigold, D. S., & Patla, A. E. (2007). Gaze fixation patterns for negotiating complex ground terrain. Neuroscience, 144, 302–313. Marigold, D. S., & Patla, A. E. (2008). Visual information from the lower visual field is important for walking across multisurface terrain. Experimental Brain Research, 188, 23–31. Marigold, D. S., Weerdesteyn, V., Patla, A. E., & Duysens, J. (2007). Keep looking ahead? Re-direction of visual fixation does not always occur during an unpredictable obstacle avoidance task. Experimental Brain Research, 176, 32–42. Marple-Horvat, D. E., Criado, J. M., & Armstrong, D. M. (1998). Neuronal activity in the lateral cerebellum of the cat related to visual stimuli at rest, visually guided step modification, and saccadic eye movements. Journal de Physiologie, 506, 489–514. Maunsell, J. H., & van Essen, D. C. (1983). Functional properties of neurons in middle temporal visual area of the macaque monkey. I. Selectivity for stimulus direction, speed, and orientation. Journal of Neurophysiology, 49, 1127–1147. McVea, D. A., & Pearson, K. G. (2006). Long-lasting memories of obstacles guide leg movements in the walking cat. The Journal of Neuroscience, 26, 1175–1178. McVea, D. A., & Pearson, K. G. (2009). Object avoidance during locomotion. Advances in Experimental Medicine and Biology, 629, 293–315. McVea, D. A., Taylor, A. J., & Pearson, K. G. (2009). Longlasting working memories of obstacles established by foreleg stepping in walking cats require area 5 of the posterior parietal cortex. The Journal of Neuroscience, 29, 9396–9404. Merchant, H., Battaglia-Mayer, A., & Georgopoulos, A. P. (2001). Effects of optic flow in motor cortex and area 7a. Journal of Neurophysiology, 86, 1937–1954. Merchant, H., Battaglia-Mayer, A., & Georgopoulos, A. P. (2004). Neural responses during interception of real and apparent circularly moving stimuli in motor cortex and area 7a. Cerebral Cortex, 14, 314–331.
100 Miall, R. C., & Wolpert, D. M. (1996). Forward models for physiological motor control. Neural Networks, 9, 1265–1279. Montagne, G., Cornus, S., Glize, D., Quaine, F., & Laurent, M. (2000). A perception–action coupling type of control in long jump. Journal of Motor Behavior, 32, 37–43. Mountcastle, V. B., Lynch, J. C., Georgopoulos, A., Sakata, H., & Acuna, C. (1975). Posterior parietal association cortex of the monkey: Command functions for operation within extrapersonal space. Journal of Neurophysiology, 38, 871–908. Mulliken, G. H., Musallam, S., & Andersen, R. A. (2008). Forward estimation of movement state in posterior parietal cortex. Proceedings of the National Academy of Sciences of the United States of America, 105, 8170–8177. Patla, A. E. (1997). Understanding the roles of vision in the control of human locomotion. Gait & Posture, 5, 54–69. Patla, A. E., & Greig, M. A. (2006). Any way you look at it, successful obstacle negotiation needs visually guided on-line foot placement regulation during the approach phase. Neuroscience Letters, 397, 110–114. Pisella, L., Gréa, H., Tilikete, C., Vighetto, A., Desmurget, M., Rode, G., et al. (2000). An ‘automatic pilot’ for the hand in human posterior parietal cortex: Toward reinterpretation of optic ataxia. Nature Neuroscience, 3, 729–736. Prokop, T., Schubert, M., & Berger, W. (1997). Visual influence on human locomotion. Experimental Brain Research, 114, 63–70. Rauschecker, J. P., von Grünau, M. W., & Poulin, C. (1987). Centrifugal organization of direction preferences in the cat's lateral suprasylvian visual cortex and its relation to flow field processing. The Journal of Neuroscience, 7, 943–958. Regan, D. (2002). Binocular information about time to collision and time to passage. Vision Research, 42, 2479–2484. Rice, N. J., Tunik, E., & Grafton, S. T. (2006). The anterior intraparietal sulcus mediates grasp execution, independent of requirements to update: New insights from transcranial magnetic stimulation. The Journal of Neuroscience, 26, 8176–8182. Rietdyk, S., & Rhea, C. K. (2006). Control of adaptive locomotion: Effect of visual obstruction and visual cues in the environment. Experimental Brain Research, 169, 272–278. Rossignol, S. (1996). Neural control of stereotypic limb movements. In L. B. Rowell & J. T. Sheperd (Eds.), Handbook of physiology. Section 12. Regulation and integration of multiple systems (pp. 173–216). Bethesda, MD: American Physiological society. Rossignol, S., Dubuc, R., & Gossard, J. P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86, 89–154. Shadmehr, R., & Krakauer, J. W. (2008). A computational neuroanatomy for motor control. Experimental Brain Research, 185, 359–381. Sherk, H., & Fowler, G. A. (2001a). Neural analysis of visual information during locomotion. Progress in Brain Research, 134, 247–264.
Sherk, H., & Fowler, G. A. (2001b). Visual analysis and image motion in locomoting cats. The European Journal of Neuroscience, 13, 1239–1248. Sherk, H., & Kim, J.-N. (2002). Responses to extrastriate cortex to optic flow during simulated turns. Visual Neuroscience, 19, 409–419. Siegel, R. M., & Read, H. I. (1997). Analysis of optic flow in the monkey parietal area 7a. Cerebral Cortex, 7, 327–346. Snyder, L. H., Grieve, K. L., Brotchie, P., & Andersen, R. A. (1998). Separate bodyand world-referenced representations of visual space in parietal cortex. Nature, 394, 887–891. Sun, H., & Frost, B. J. (1998). Computation of different optical variables of looming objects in pigeon nucleus rotundus neurons. Nature Neuroscience, 1, 296–303. Sun, H.-J., Carey, D. P., & Goodale, M. A. (1992). A mammalian model of optic flow utilization in the control of locomotion. Experimental Brain Research, 91, 171–175. Thompson, R. F., Johnson, R. H., & Hoopes, J. J. (1963). Organization of auditory, somatic sensory, and visual projection to association fields of cerebral cortex in the cat. Journal of Neurophysiology, 26, 343–364. Tunik, E., Frey, S. H., & Grafton, S. T. (2005). Virtual lesions of the anterior intraparietal area disrupt goal-dependent on-line adjustments of grasp. Nature Neuroscience, 8, 505–511. Wang, Y., & Frost, B. J. (1992). Time to collision is signaled by neurons in the nucleus rotundus of pigeons. Nature, 356, 236–238. Wann, J. P. (1996). Anticipating arrival: Is the tau margin a specious theory? Journal of Experimental Psychology: Human Perception and Performance, 22, 1031–1048. Warren, W. H., Young, D. S., & Lee, D. N. (1986). Visual control of step length during running over irregular terrain. Journal of Experimental Psychology: Human Perception and Performance, 12, 259–266. Warren, W. H., Kay, B. A., Zosh, W. D., Duchon, A. P., & Sahuc, S. (2001). Optic flow is used to control human walking. Nature Neuroscience, 4, 213–216. Wolpert, D. M., & Ghahramani, Z. (2000). Computational principles of movement neuroscience. Nature Neuroscience, 3(Suppl.), 1212–1217. Wolpert, D. M., Ghahramani, Z., & Jordan, M. I. (1995). An internal model for sensorimotor integration. Science, 269, 1880–1882. Wolpert, D. M., Goodbody, S. J., & Husain, M. (1998). Maintaining internal representations: The role of the human superior parietal lobe. Nature Neuroscience, 1, 529–533. Wu, L.-Q., Niu, Y.-Q., Yang, J., & Wang, S.-R. (2005). Tectal neurons signal impending collision of looming objects in the pigeon. The European Journal of Neuroscience, 22, 2325–2331.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 7
Interindividual variability and its implications for locomotor adaptation following peripheral nerve and/or spinal cord injury Alain Frigon* Département de physiologie et biophysique, Université de Sherbrooke, Sherbrooke, Quebec, Canada
Abstract: Following injury to the nervous system, there is a range of possible functional outcomes that can only be partly explained by the extent of injury. Moreover, treatments effective in certain individuals might not work in others. Why such variability from one individual to another, in terms of functional outcomes and responsiveness to a given treatment following a similar injury? The answer to that question is not simple, and to begin to answer we must first consider that individuals of the same species can be quite variable in terms of neuronal circuit parameters involved in performing a given task. Interindividual variability can be subtle but the term “variability” in this chapter will be used to denote marked differences between individuals at the systems level (e.g., spinal reflexes, bursts of muscle activity, kinematics) during the same motor behavior, with an emphasis on locomotion. Injury to any level of the nervous system, in turn, can further compound this variability by altering spared neuronal connections. The aim of the present chapter is to (1) review studies that have investigated interindividual variability, (2) review studies that have described variable adaptive mechanisms following spinal and/or peripheral nerve lesions during locomotion, and (3) discuss the implications of intersubject variability for locomotor adaptation. Keywords: variability; locomotion; spinal cord injury; peripheral nerve injury; reflex; adaptation.
called a central pattern generator (CPG) (reviewed in Delcomyn, 1980; Grillner, 1981; Lundberg, 1981; McCrea and Rybak, 2008; Rossignol et al., 2006). A key feature of the spinal locomotor CPG is its enormous flexibility in adapting to changing conditions in the short and long terms (Pearson, 2000). One consequence of this flexibility is that different solutions can
Introduction In vertebrates, the basic pattern of walking is produced by neuronal circuitry within the spinal cord
*Corresponding author. Tel.: þ1-312-503-1323 E-mail:
[email protected] DOI: 10.1016/B978-0-444-53825-3.00012-7
101
102
achieve the same goal. As a result, interindividual variability can arise in neuronal connections. Before delving into the issue of interindividual variability, it is important to discuss how the locomotor program is configured in the first place via a predetermined genetic program and through interactions with the environment during development. It is possible that the spinal network acquires the ability to produce locomotion through developmental processes, as the animal learns to walk in the early stages of life. However, several animals walk and run almost immediately after birth and studies have shown that completely transecting the spinal cord soon after birth does not prevent the expression of hindlimb locomotion (i.e., spinal locomotion), when the legs are placed on a motorized treadmill (Rossignol, 1996). In fact, hindlimb locomotion is often better in animals spinalized soon after birth compared to the same lesion performed in adults. Well-developed bipedal walking can also be observed in human infants within the first year of life, well before learning to walk voluntarily, when the legs are placed on a moving treadmill (reviewed in Yang et al., 2004). In human infants, the corticospinal tract is extremely immature at birth and locomotor-like movements are largely independent of volitional descending control (Forssberg, 1985; Yang et al., 2004). However, descending pathways from the brainstem are functioning and capable of configuring the spinal network to produce locomotion if appropriate sensory cues are provided. Therefore, it is likely that the locomotor program has a very strong genetically predetermined component. Additional support for a genetically predetermined program comes from tendon transfers in adult cats and kittens, in which crossed or transferred muscles largely retain their normal activation pattern (Forssberg and Svartengren, 1983; Loeb, 1999; O'Donovan et al., 1985). For example, transposing the tendons of the lateral (LG) and medial gastrocnemii (MG), both ankle extensors, on to the cut tendon of the tibialis
anterior (TA), an ankle flexor, in adult cats and kittens, did not disrupt the normal activation of gastrocnemii muscles during the stance phase of locomotion, even though they now functioned as anatomical ankle flexors, which resulted in deficits (Forssberg and Svartengren, 1983; Loeb, 1999). However, although transferred or crossed muscles retained their normal locomotor activity pattern, cutaneous reflex responses could display considerable left/right asymmetry, suggesting some plasticity in the interactions between the spinal CPG and reflex pathways (Loeb, 1999). Evidently, some of these changes could be mediated by altered interactions between sensory feedback and supraspinal and spinal structures, as the animal learns to cope with an altered anatomical organization. Therefore, although it is clear that the basic locomotor program is in large part genetically determined, some components of the spinal CPG, and in the interactions between spinal, peripheral, and supraspinal structures, are malleable. During development and through repeated practice (i.e., trial and error), some connections within the central nervous system become stronger, or weaker, than others (recently reviewed in Butz et al., 2009; Holtmaat and Svoboda, 2009). Neuronal connections within the spinal locomotor CPG are most likely not exempt from this phenomenon. For example, in spinalized kittens (Forssberg et al., 1980) and human infants, the walking pattern of the legs adapts to different treadmill speeds, to independent speeds for each leg on a split-belt treadmill, and to load-related signals (Musselman and Yang, 2007; Yang et al., 1998, 2005), thus showing that sensory feedback interacts with the spinal locomotor CPG in the earliest stages of life. The consequences of the interactions between sensory feedback and the spinal CPG during development and over a lifetime is largely unknown but there is evidence that some intrinsic properties of the locomotor circuitry are permanently altered. For instance, we recently showed that the stance and swing phases of the cycle period (i.e., the time between
103
successive bursts) were asymmetrically controlled during fictive locomotion in adult cats in which supraspinal signals and phasic sensory feedback were abolished, as they were curarized, decerebrated, and spinalized (Frigon and Gossard, 2009). Specifically, cycle period varied more with the duration of the extension phase (i.e., extensor-dominated), while the flexion phase remained relatively invariant, similar to what is observed during normal walking where cycle period changes as a function of the stance phase with the swing phase remaining invariant (Grillner et al., 1979; Halbertsma, 1983). However, in neonatal rats during fictive locomotion (Juvin et al., 2007), and human infants during air-stepping (Musselman and Yang, 2007), extensor-dominant asymmetry in regulating cycle period is only observed if phasic sensory inputs consistent with locomotion are provided. It might be that the locomotor spinal CPG, over several years, develops an intrinsic extensor-dominant asymmetry due to repetitive interactions with stance-related sensory feedback (Frigon and Gossard, 2009). The pattern of walking in human infants also shows marked differences with the adult pattern in terms of foot placement and intralimb coordination (Forssberg, 1985; Yang et al., 2004). Therefore, as the interactions between spinal, supraspinal, and peripheral systems mature and strengthen over several years, it is likely that “hard-wired” changes occur within some components of the locomotor circuitry. Consequently, because each animal is confronted with a different set of circumstances over a lifetime, interindividual differences within the locomotor circuitry can arise. Changes in reflex pathways during development are well demonstrated in the nociceptive withdrawal reflex of the rat (Levinsson et al., 2002; Schouenborg, 2002, 2008). At birth, withdrawal reflexes are maladaptive, often causing movements toward the noxious stimuli. In the first 3 weeks after birth, the reflex circuitry is shaped by experience-dependent mechanisms, whereby erroneous connections are eliminated or reduced and appropriate
connections are strengthened, becoming proportional to withdrawal efficiency. Changes in reflex pathways also occur in the adult system following lesions or using learning paradigms. In the past 20 years, Wolpaw, Chen, and colleagues have clearly established that operant conditioning of the soleus H-reflex in adult rats induces persistent changes in the spinal circuitry and in how these pathways interact with supraspinal signals (reviewed in Wolpaw, 2007; Wolpaw and Tennissen, 2001). In invertebrate motor systems, recent experimental and modeling studies have shown considerable variability in the production of motor patterns between animals, in a relatively simple motor system, the pyloric rhythm of the lobster (Bucher et al., 2005; Marder and Goaillard, 2006; Prinz et al., 2004). What emerges from these studies is that widely disparate circuit parameters can produce similar network activity. Moreover, eliminating one parameter (e.g., a specific conductance) often does not alter network activity because of different combinations of compensatory mechanisms. During development, as the animal grows in size, parameters of the network change but network performance remains essentially the same (Bucher et al., 2005; Marder and Goaillard, 2006; Schulz et al., 2006). As a consequence, because there are many ways to generate a desired outcome, interindividual variability can arise in how the neuronal network is configured. Therefore, although the act of walking is similar between individuals of the same species, including humans, substantial differences can occur in the locomotor circuitry due to activitydependent mechanisms. In fact, humans might be the most “variable” animal of all, because in humans, physical activity is influenced by myriad factors, such as gender, cultural background, socioeconomic status, genetic predispositions (e.g., height, weight), age, climate, geography, etc. (reviewed in Caperchione et al., 2009; Kumanyika, 2008; Seefeldt et al., 2002). As such, it is not surprising that neuronal connections at the systems level can differ so dramatically from
104
Interindividual variability during locomotion Gait transitions and walking speed Walking, whether tested on a treadmill or on the ground, can vary from one animal to another. For instance, during treadmill locomotion at matched speeds some cats will walk, whereas others will trot or gallop (Vilensky and Patrick, 1984; Vilensky et al., 1990). The transition speed from one gait pattern to another, such as walking, trotting, and galloping varies between animals. Additionally, some animals will even trot or gallop at the same speed on different days (Vilensky and Patrick, 1984). The coordination between the four limbs can also display intra- and interanimal variability during walking in the cat (English, 1979). For example, some cats prefer a trot-like coordination (i.e., homolateral limbs are out-ofphase), whereas some cats adopt a pacing pattern (i.e., homolateral limbs are in-phase) during overground or treadmill walking (English, 1979; Stuart et al., 1973; Vilensky and Patrick, 1984; Wetzel et al., 1975). Although the four limbs are loosely coupled, individual cats will primarily adopt one type of strategy (English, 1979). Walking speed is also interesting in the context of interindividual variability. Individuals have a preferred walking speed and any modification requires conscious voluntary control. For example, we alter our speed when walking with someone else, or increase our speed when in a hurry, both of which requires volitional control. Cats are no different. When walking on a treadmill they usually walk most consistently at a specific speed, which does not necessarily depend on the size of the animal. Some cats will prefer a slow treadmill speed (e.g., 0.3 m/s), whereas other cats will walk more consistently at higher speeds (e.g., 0.6 m/s). The majority of cats prefer a treadmill
speed of 0.4–0.5 m/s. During fictive locomotion the rhythm also displays a range of “speeds,” which can be inferred by measuring cycle period (i.e., the time between successive bursts of activity in a given nerve). Figure 1 shows cycle periods from 34 episodes of spontaneous fictive locomotion recorded in 27 adult cats that had undergone the same experimental procedure (Frigon and Gossard, 2009), and from 21 episodes of treadmill locomotion in 21 intact cats walking at their “preferred speed” (i.e., 0.4 or 0.5 m/s). During fictive locomotion, the average locomotor cycle period was 1065 466 ms, which approximately corresponds to a treadmill speed of 0.3–0.35 m/s during intact
35 30 25
Episode number
one individual to another, even though the general goal and pattern of walking is essentially the same.
20 15 10 5 0 0
500
1000 1500 2000 Cycle period (ms)
2500
3000
Treadmill locomotion Spontaneous fictive locomotion Fig. 1. Cycle periods during spontaneous fictive locomotion in adult decerebrate cats and during treadmill locomotion in intact cats. Cycle period was measured from successive burst onsets in selected extensor nerves during spontaneous fictive locomotion and from successive left foot contacts during treadmill locomotion. The data set consists of 34 episodes of fictive locomotion from 27 adult decerebrate cats (Frigon and Gossard, 2009), and from 21 episodes of treadmill locomotion in 21 intact cats. Each data point is the average of 15 cycles.
105
treadmill locomotion in the cat (Halbertsma, 1983). Of the 34 episodes only 3 fell outside one standard deviation of the mean (2 above and 1 below). Thus, if we omit the few outliers, the locomotor CPG, without phasic sensory feedback and volitional control, operates within a narrow range and the cycle period from these 31 episodes now becomes 983 241 ms. This is remarkable considering that, in such preparations, there are some interindividual differences in supraspinal drive, overall excitability, static limb posture, etc. In intact cats walking on a treadmill, the cycle period was 941 98 ms, which is strikingly similar to the average cycle period during spontaneous fictive locomotion. Moreover, even though cycle period is constrained by the speed of the treadmill there is still some interanimal variability. The small range of speeds during spontaneous fictive locomotion, inferred by measuring cycle period, could partly reflect some interindividual variability in the intrinsic organization of the CPG, which might play a role in determining the animal's preferred walking speed. To change this “default” speed would require voluntary control or peripheral sensory feedback, as observed during treadmill locomotion, where the imposed speed narrows the range of cycle periods. Although interindividual variability appears to be an inescapable component of motor systems, few studies have specifically assessed it, directly or indirectly, in vertebrate preparations. Improved chronic recording and stimulating procedures have made quantifiable measures possible in animal studies for prolonged periods, while reducing variability related to methodological sources (Loeb, 1993). That is not to say that interindividual variability is not partly related to experimental procedures; only that methodological issues cannot account fully for observed interindividual variability. What emerges from such studies is that activation patterns and reflexes during locomotion, or other natural behaviors, are very consistent from day to day in the same animal but can be quite variable from one animal to another.
Recruitment patterns and spinal reflexes Loeb (1993) investigated interanimal variability in the activation profile of certain muscles and in electrically evoked cutaneous reflexes during treadmill locomotion in the cat and found that most animals showed idiosyncratic patterns of activation and reflex responses in at least some hindlimb muscles. For instance, flexor digitorum longus (FDL, ankle adductor/digit flexor) was consistently activated at the time of foot lift but in some animals it could also be recruited during the stance phase (Loeb, 1993). Its close synergist, flexor hallucis longus (FHL, ankle extensor/digit flexor), which shares origins and insertions with FDL (Abraham and Loeb, 1985; Fleshman et al., 1984; O'Donovan et al., 1982; Schmidt et al., 1988), was consistently recruited during the stance phase, along with other extensors, in all cats. Interestingly, cutaneous reflex responses evoked by stimulating the superficial peroneal (SP) nerve differed between animals (i.e., showed idiosyncratic responses) in FHL but not FDL. Thus in the muscle that was consistently recruited during stance in all animals (i.e., FHL), cutaneous reflex responses differed from one animal to the other, whereas in a muscle that showed interanimal variability in its recruitment pattern (i.e., FDL), reflex responses were very consistent. The peroneus longus (ankle abductor/flexor) was another muscle that showed considerable variability in its recruitment pattern but consistent excitatory reflexes during swing and late stance. It could be coactive with TA during swing, active during both stance and swing, and in other cats it could be relatively silent during the entire locomotor cycle (Loeb, 1993). Interindividual variability in cutaneous reflexes is also present in humans at rest and during walking. During human walking, in some subjects, instead of a middle latency excitatory response, there is an inhibitory response, particularly with stimulation of cutaneous afferents that supply the plantar surface of the foot (Haridas et al., 2005; Zehr et al., 1997). Therefore, variability in recruitment
106
patterns and/or cutaneous reflexes is observed in multiple muscles during walking in the cat and humans. Figure 2 shows reflex responses in the left semitendinosus (St) muscle, evoked by stimulating the tibial (Tib) nerve with a cuff electrode just distal to the medial malleolus, in two intact cats during treadmill locomotion (Frigon and Rossignol, 2008a). Stimulating the Tib nerve during locomotion usually evokes short- (P1) and longer latency (P2) excitatory responses, which are modulated according to the phase of the cycle (Abraham et al., 1985; Drew and Rossignol, 1985; Duysens and Stein, 1978; Loeb, 1993). In the example depicted in Fig. 2, P1 and P2 responses were modulated throughout the locomotor cycle in cat 1, with large responses during the swing phase and the stance-to-swing transition (Fig. 2a). In cat 2, however, there were virtually no P1 responses throughout the locomotor cycle but there were large P2 responses during swing, the swing-tostance transition, early stance, and the stance-toswing transition. The rectified electromyography (EMG) of the left St pooled from nonstimulated cycles is shown on the far right for both cats, and clearly, differences in reflex responses between cats cannot be explained by large differences in the timing of bursting activity of the muscle. As we will see later, this interindividual variability in reflex response can be altered following a complete spinal transection, which abolishes all supraspinal inputs to the spinal locomotor circuitry.
Muscle synergies It is thought that the central nervous system simplifies the task of controlling movement by combining multiple muscles into synergies or modules (Bizzi et al., 2000, 2008; Grillner, 1981; Grillner and Wallen, 1985; Ivanenko et al., 2007; Jordan, 1991; Krouchev et al., 2006; Loeb et al., 2000; Schouenborg, 2003; Stein and Smith,
1997). Quantifying muscle synergies during normal motor behaviors offers another means of testing recruitment patterns/activation profiles. There is good evidence that most, but not all, muscle synergies are organized centrally and that sensory feedback from the periphery and/or from supraspinal signals adjusts the temporal pattern of synergy activation (Cheung et al., 2005). Interestingly, the number and type of muscle synergies involved in producing a given movement can differ from one animal to another (Cheung et al., 2005; Krouchev et al., 2006). For instance, in the bullfrog, it was shown that hindlimb swimming was produced by four different muscle synergies in some frogs, whereas in other frogs up to six different synergies were used (Cheung et al., 2005). Although many of the synergies were shared between animals, some frogs displayed unique synergies. How can the system give rise to different muscle synergies between animals? One obvious answer is that the interindividual variability in sensory feedback from the periphery and/or supraspinal signals shape the activation of different muscle synergies. Cheung et al. (2005) performed a unilateral deafferentation of dorsal roots 7–9 in bullfrogs to abolish some sensory feedback from the hindlimbs, and although most synergies were preserved, the total number of synergies changed slightly in three of four frogs. The interindividual difference in the number of synergies also persisted. Moreover, deafferentation appeared to compound the interindividual variability in the temporal pattern of synergy activation with some frogs showing an increase, a decrease, or no change in specific synergies. Therefore, it is likely that the emergence of synergies within the spinal cord is a complex process that involves interactions between genetically determined motor programs and experiencedependent processes. The use of slightly different strategies from one animal to another to accomplish the same movement is the resultant of these complex interactions.
107 (a)
Phase of locomotor cycle (normalized to left St onset)
Cat 1
300 mV
P1
P2
0.0
Swing
Swing-tostance transition
{
Stance
Stance-toswing transition
{ –5
(b)
15 35 Tib nerve stimulation
55
{ { { { { 0.5 { { { { { 1.0
75
28 mV
150 mV
Cat 2
0.0
Swing
Swing-tostance transition
{
Stance
Stance-toswing transition
{ –5
15 35 55 Tib nerve stimulation Time relative to stimulation (ms)
{ { { { { 0.5 { { { { { 1.0
75
48 mV
Fig. 2. Reflex responses of the left St evoked by stimulating the tibial (Tib) nerve at 1.5 times the motor threshold in two intact cats, (a, Cat 1; b, Cat 2) during treadmill locomotion. Values for each horizontal trace in a single graph are at the same scale in mV. Each horizontal trace is the average of approximately 10 cycles with stimulation superimposed on the background level of EMG derived from control cycles. The first horizontal trace in each figure is phase 0.05 (i.e., from 0.0 to 0.10) of the locomotor cycle synchronized to left St burst onset followed by 0.10, 0.15, etc. The rectified activity of the left St from control cycles is shown on the far right at 90 . The dashed vertical line indicates the time of the stimulation, whereas solid vertical lines indicate the time windows used to delineate P1 (black area, 10–25 ms) and P2 (gray area, 25–55 ms) responses.
108
Factors inducing interindividual variability What gives rise to this variability in reflexes and activation patterns during locomotion between animals? As pointed out before, experience and genetic factors (e.g., genetic polymorphisms) probably play a large role in inducing interindividual variability but discussing those factors is outside the scope of this chapter. There are, however, more tangible factors that can cause disparities in the locomotor circuitry between individuals. For example, as discussed by Loeb (1993), the presence of consistencies in recruitment patterns between animals is not surprising because the musculoskeletal system imposes mechanical constraints on the control system, which are primarily genetically determined. However, there is anatomical variability between animals of the same species and differences in size, mass, and structure of the mechanical system impose different challenges on the neural control of locomotion. Interestingly, despite a wide range of sizes within a given species there is little relation between body size and specific gait parameters. For instance, in genetically related vervet monkeys, body mass and segment length did not correlate with gait transition speed (i.e., from trotting to galloping) because of considerable interanimal variability (Vilensky and Gankiewicz, 1990; Vilensky et al., 1988, 1990). At a particular body mass, there was a substantial range of transition speeds between animals. Moreover, within a group, changes in gait parameters did not change consistently with age between animals, although as the animal aged and grew in size it tended to maintain a similar gait pattern at match speeds (Vilensky and Gankiewicz, 1990). Therefore, each animal appears to use a unique strategy to produce the propulsive forces necessary for locomotion across speeds and body size. Vilensky et al. (1990) postulated that the underlying factors of this interanimal variability could be physiological, psychological, and/or morphological. What is clear is that the neural locomotor program is extremely flexible.
Variable adaptive mechanisms following peripheral nerve lesions Loeb (1993) made an important point: if neural circuits for a given behavior can differ between animals, this raises important methodological problems for studies that pool data among animals. One way to circumvent this methodological problem is to evaluate changes following a given treatment on a case-by-case basis to determine consistencies between individuals, while concomitantly highlighting variable or unique adaptive strategies. In recent studies, we used this approach to quantify changes in the locomotor pattern and cutaneous reflexes following lesions to the spinal cord and/or peripheral nerves (Frigon and Rossignol, 2007, 2008a,b, 2009; Frigon et al., 2009). This section discusses changes following peripheral nerve lesions in intact (i.e., with an intact spinal cord) and chronic spinal cats while the next section discusses changes following spinal cord lesions. What is clear from these studies is that a given lesion can produce different outcomes and compensatory mechanisms from one animal to another. To evaluate locomotor and reflex adaptation following a peripheral denervation, we performed a lesion of the left lateral gastrocnemius-soleus (LGS) nerve in otherwise intact cats (Frigon and Rossignol, 2007), or in cats that had undergone a complete spinal transection (i.e., spinalization) several weeks before the denervation (Frigon and Rossignol, 2008b). Cats were implanted with recording and stimulating electrodes and hindlimb kinematics, patterns of muscle activation, and cutaneous reflexes evoked by stimulating the Tib nerve, were quantified during treadmill locomotion in the same animal before and after denervating the LGS. Following denervation of the LGS, the most consistent deficit was an increase in ankle flexion, or yield, at the onset of the stance phase in intact (Frigon and Rossignol, 2007; Pearson et al., 1999) and chronic spinal (Bouyer et al., 2001; Frigon and Rossignol, 2008b) cats. However, the magnitude of this
109
deficit and the time course of recovery could substantially differ between animals. Figure 3 shows changes in joint excursions in two intact cats following a lesion of the left LGS nerve (Frigon and Rossignol, 2007). For instance, in cat 3, ankle yield increased considerably during early stance following LGS denervation (Fig. 3a), whereas in cat 4 there was virtually no change in ankle yield (Fig. 3b). At the hip and metatarsophalangeal (MTP) joints, changes were consistent between cats 3 and 4, but at the knee there was a marked difference. The left knee joint was more extended throughout the locomotor cycle in cat 3 (Fig. 3a), while in cat 4 it was more flexed (Fig. 3b) following denervation. There were also similarities and marked differences in muscle activation profiles after denervation (Figs. 3c and d). For example, activity increased in the left MG, left VL, and right VL, but remained unaltered in the left Srt in both cats. However, in cat 4, the left MG occupied a greater percentage of the cycle, which was associated with a marked delay in the onset and magnitude of activity in the right St (Fig. 3d). It thus appears that cat 4 had a smaller locomotor deficit (i.e., ankle yield) by using a more prominent bilateral adjustment between hindlimbs. The comparison of these 2 cats illustrates that although some adaptive changes can be similar between cats, there are still unique compensatory strategies that emerge following injury.
Variable adaptive mechanisms following spinal cord lesions Spinal cord injury (SCI) severely disrupts interactions between supraspinal, spinal, and peripheral structures (Cai et al., 2006; Frigon and Rossignol, 2006; Rossignol, 2006; Rossignol et al., 2008, 2009). Due to the inherent interindividual variability in neuronal connections and interactions between structures, it is not surprising that SCI can induce variable changes and adaptive strategies between animals. For
instance, following a complete SCI, changes in reflex responses during locomotion can differ quite dramatically between cats. During locomotion in intact cats, stimulating a cutaneous nerve evokes a pattern of short-latency inhibition (N1) followed by longer latency excitatory responses (P2) in ipsilateral (i.e., on the same side as the stimulation) extensors during stance. This pattern shows little variability between cats. However, following spinalization, there can be considerable interanimal variability. In some cats, following spinalization, P1 responses appear in LG and MG muscles during the stance phase, instead of the more common short-latency inhibition (Frigon and Rossignol, 2008a). In other cats, however, N1 responses persist. In an interesting case, in one cat P1 responses appeared in the left MG but not the left LG. P2 responses could increase, decrease, or remain unaltered from one extensor muscle and cat to another. This intra- and interanimal variability undoubtedly reflects some differences in cutaneous reflex pathways to specific motor pools and/or in independent sources of modulation to close synergists (Degtyarenko et al., 1996). It also indicates that supraspinal signals are important in shaping spinal reflex excitability, because a change from N1 to P1 responses after spinalization suggests activation of an alternate reflex pathway (i.e., a different spinal interneuronal route). Differences in reflex changes between animals can also be more subtle following spinalization. In Fig. 2, we showed that in cat 1 there were prominent P1 and P2 responses that were phasemodulated during locomotion, whereas in cat 2 primarily only P2 responses were found. Figure 4 shows the same nerve stimulation in the same two cats following a complete spinal transection. Note that the scales for reflex responses are larger in Fig. 4, while those for the activity of the St are the same as in control cycles. The bursting behavior of the muscle is shown on the far right and there was little change after spinalization. As can be seen, P2 responses were reduced in both cats, probably because the pathway responsible
110 Cat 3 Swing
(a) Stance F
E1
Stance F
E2 E3
150
80 160
80 170
90 150
100 160
100 260
100 270
180
180
0.0
1.5 0.5 1.0 Phase of locomotor cycle (normalized to left foot contact)
2.0
0.0
2 days
E2 E3
0.5 1.0 1.5 Phase of locomotor cycle (normalized to left foot contact)
2.0
0.5 1.0 1.5 Phase of locomotor cycle (normalized to left MG onset)
2.0
(d)
15 mV
Right St
Right VL 0.0
45 mV
35 mV
Left VL
17 mV
60 mV
Left MG
75 mV
37 mV
Left Srt
70 mV
55 mV
(c)
Intact
E1
35 mV
Left hindlimb joint angles (°) MTP Ankle Knee Hip
140
Cat 4 Swing
(b)
1.5 0.5 1.0 Phase of locomotor cycle (normalized to left MG onset)
2.0
0.0
Fig. 3. Angular excursions and patterns of muscular activation before and 2 days after sectioning the left LGS nerve in two cats during locomotion. (a, b) Changes in the angle of the hip, knee, ankle, and metatarsophalangeal (MTP) joints of the left hindlimb during locomotion. The locomotor cycle is normalized to contact of the left foot. Phases and subdivisions (F, E1, E2, E3) are indicated at the top. (c, d) Rectified EMG bursts of selected muscles during the same episodes as in (a) and (b). The locomotor cycle is normalized to the onset of the left MG burst. Srt, anterior part of sartorius (hip flexor/knee extensor); MG, medial gastrocnemius (ankle extensor/knee flexor); VL, vastus lateralis (knee extensor); and St, semitendinosus (knee flexor/hip extensor). Each waveform is the average of approximately 20 cycles.
111
600 mV
(a)
Swing
{
Swing-tostance transition
{
Stance
Stance-toswing transition
P1
Phase of locomotor cycle (normalized to left St onset)
Cat 1 P2
0.0
{ { −5
(b)
15 35 Tib nerve stimulation
55
{ { { { { 0.5 { { { { { 1.0
75
28 mV
400 mV
Cat 2
Swing
Swing-tostance transition
Stance
0.0
{ {
{
Stance-toswing { transition −5
15 35 55 Tib nerve stimulation Time relative to stimulation (ms)
{ { { { { 0.5 { { { { { 1.0
75
48 mV
Fig. 4. Reflex responses of the left St evoked by stimulating the tibial (Tib) nerve at 1.5 times the motor threshold in the same two cats as Fig. 2 during treadmill locomotion but following a complete spinal transection at T13. The responses shown in (a) and (b) are more than 40 days following spinalization. Values for each horizontal trace in a single graph are at the same scale in mV. However, the scales are different than in Fig. 2. Each horizontal trace is the average of approximately 10 cycles with stimulation superimposed on the background level of EMG derived from control cycles. The first horizontal trace in each figure is phase 0.05 (i.e., from 0.0 to 0.10) of the locomotor cycle synchronized to left St burst onset followed by 0.10, 0.15, etc. The rectified activity of the left St from control cycles is shown on the far right (90 ) at the same scales as in Fig. 2. The dashed vertical line indicates the time of the stimulation, whereas solid vertical lines indicate the time windows used to delineate P1 (black area, 10–25 ms) and P2 (gray area, 25–55 ms) responses.
112
depends on a supraspinal contribution. In cat 1, P1 responses increased throughout the locomotor cycle but the pattern of phase-dependent modulation was similar (compare Fig. 4a with Fig. 2a). In cat 2, however, large P1 responses appeared during the swing phase and the swing-to-stance transition, a pattern that now more closely resembles cat 1. Such changes are most likely due to differences in the interactions between supraspinal signals and reflex pathways (Eccles and Lundberg, 1959; Holmqvist and Lundberg, 1959; Lundberg, 1967; Schomburg, 1990). There is a strong supraspinal contribution to reflex responses in flexors, which does not appear to be the case for extensors (Frigon and Rossignol, 2008a; Shimamura et al., 1991). Overall, following spinalization there are not only marked consistencies but also variable changes between animals during spinal locomotion.
Dual-lesion paradigms What happens if a lesion is followed by a complete spinal lesion? In such “dual-lesion paradigms,” the capacity to adapt to the new state of the system following spinalization is critically dependent on the first lesion. For instance, performing a peripheral nerve lesion before a spinalization adversely influences the ability to express spinal locomotion (Bouyer and Rossignol, 2003; Carrier et al., 1997; Frigon and Rossignol, 2009), whereas performing a partial spinal lesion before spinalization facilitates the expression of spinal locomotion (Barriere et al., 2008; Frigon et al., 2009). Recently, we showed that sectioning the left LGS nerve in three cats before spinalization introduced considerable interanimal variability in the ability to express spinal locomotion, which ranged from an inability to express spinal locomotion in one cat to an abnormal form of spinal locomotion in another (Frigon and Rossignol, 2009). Why such variability between animals following the same experimental protocol? It should be noted that no concerted
effort was made to provide a consistent behavioral context (e.g., daily locomotor training) in all cats after the initial peripheral denervation. Consequently, we surmise that activity-dependent processes following the peripheral denervation introduced different compensatory mechanisms, particularly in the interactions between supraspinal and peripheral inputs within the spinal locomotor circuitry. This was reflected by variable changes in reflex responses between cats (Frigon and Rossignol, 2007). As a result, the configuration of the locomotor circuitry was different from one cat to another at the time of spinalization, which produced variable changes following complete SCI. Further experiments are required to confirm this hypothesis. In contrast to a peripheral denervation, if a partial spinal lesion is followed by spinalization, the ability to express spinal locomotion is greatly facilitated (Barriere et al., 2008; Frigon et al., 2009). In some cats, a full weight-bearing hindlimb locomotion that can adjust to very high treadmill speeds is expressed within 24 h of the spinalization (Barriere et al., 2008). Treadmill training following the partial spinal lesion appears particularly helpful for the expression of hindlimb locomotion following spinalization because all trained cats expressed spinal locomotion within 24 h of spinalization. However, cats that were not trained following the partial lesion only showed some unilateral movements on the side of the partial lesion following spinalization. Thus, locomotor training appeared to increase consistency between animals after incomplete SCI. Treadmill training in complete spinal cats was shown to normalize the interactions between the spinal locomotor program and sensory pathways from load and cutaneous receptors (Cote and Gossard, 2004; Cote et al., 2003). A similar phenomenon might occur following incomplete SCI, whereby locomotor training standardizes interactions between the supraspinal, spinal, and peripheral structures, thus reducing interanimal variability following spinalization.
113
Thus, as discussed in more detail in the following section, the “state” of the system at the time of spinalization can profoundly influence the recovery of locomotion following complete SCI and induce dramatic differences between individuals. Although this cannot directly extend to human studies it does suggest that the level of motor skill at the time of injury might be an important factor in how well SCI patients can recover.
Implications for locomotor adaptation Interindividual variability within motor systems undoubtedly has functional implications. If circuits were rigidly hard-wired it would prevent or greatly diminish our ability to learn new motor tasks and to recover functions following injury. There might also be an evolutionary function of interindividual variability, which permits new outcomes or motor behaviors to emerge within a population when confronted with a different set of circumstances. If variability was absent, the same solution to a given problem would occur over and over again. That animals use different strategies for learning and in the context of recovery simply reflects the inherent variability within sensorimotor systems, and also highlights the remarkable flexibility of the central nervous system. Consequently, it is not surprising that lesioning a given structure produces different results between animals during the same motor behavior. However, interindividual variability does pose a problem for therapeutic interventions where wishful thinking assumes that a given treatment will produce the desired outcome in all patients. In spinal cord-injured humans, treadmill training has been shown to be an effective treatment to promote the recovery of walking (Barbeau and Fung, 2001; Dietz et al., 1995; Edgerton et al., 2001; Harkema, 2001, 2008; Wernig et al., 1995). However, not all patients respond to treadmill training (Gorassini et al., 2009; Norton and
Gorassini, 2006). For instance, in 17 SCI patients subjected to the same treadmill training protocol, 9 responded positively, whereas 8 SCI showed no significant improvements (Gorassini et al., 2009), a nonnegligible proportion. In responders, locomotor training induced several changes in the magnitude and timing of muscular activity during walking, which were not observed in nonresponders. Specifically, Gorassini et al. (2009) found that responders had greater EMG activity in leg muscles prior to locomotor training and greater voluntary muscle strength, compared to nonresponders, indicating that the efficacy of spared descending pathways might be an important contributing factor in the recovery of walking. Interestingly, locomotor training increased the regularity of walking in responders and nonresponders (i.e., the timing of EMG bursts was more consistent from one walk cycle to another). In another study, the trajectory of the foot was also more consistent following training in SCI subjects (Grasso et al., 2004). Treadmill training is based on the principle that providing sensory cues consistent with normal walking facilitates locomotor recovery (Harkema, 2001). As suggested by others, this might be accomplished by modifying, or “normalizing” interactions between peripheral sensory feedback and the spinal locomotor CPG (Cote and Gossard, 2004; Cote et al., 2003; Frigon and Rossignol, 2006), a process that normally depends strongly on supraspinal influences. That the walking pattern is more regular after locomotor training in SCI patients is consistent with the hypothesis that training might stabilize interactions in spared descending pathways and peripheral sensory inputs with the spinal locomotor circuitry. It should also be noted that changes in the pattern of muscular activation in responders did not evolve or revert to a pattern normally observed in neurologically intact individuals. Instead, different adaptive strategies, such as increased extensor activity and cocontraction of antagonist muscles, were used (Gorassini et al., 2009). Adaptive strategies can also differ from one SCI
114
patient to another. For instance, complete paraplegics make greater use of their arms and body to assist leg movements compared to incomplete paraplegics (Grasso et al., 2004). In effect, SCI patients attempt to maximize their locomotor performance by optimizing the function of remnant pathways and structures. This does not imply that the intrinsic spinal locomotor circuitry is rebuilt anew but that substantial modifications take place in how different inputs interact with the CPG. Consequently, the goal of rehabilitation should not be to alter the locomotor circuitry in one specific way but to shape the circuitry so it can accomplish the task more effectively. As a result, there probably are several optimal solutions to the same problem. Pretend that two very different configurations of the locomotor network can produce adequate walking. Some patients might respond better to one particular treatment because the remaining circuitry (i.e., the current state) can more easily be directed to one type of configuration than the other. For instance, if very little volitional control remains, protocols aimed at enhancing sensory feedback from the legs might be more effective than protocols geared primarily at voluntarily activating the legs. Moreover, some neurologically intact human subjects and animals are less responsive to a given input (e.g., sensory feedback), and it is likely that this phenomenon persists following injury. Therefore, it is important to consider the current state of the system, and hence interindividual variability, before initiating or pursuing costly training protocols.
Concluding remarks If intersubject variability is so ubiquitous, how is it that so few studies have specifically addressed it? Well for one, variability between individuals was often considered to result from experimental error. However, methodological improvements over the years have revealed that much of this variability cannot be explained by experimental
shortcomings. For example, to uncover the underlying mechanisms of locomotor recovery after injury, it is imperative that recordings be made in the same animal before and after a given lesion in order to minimize and evaluate interindividual variability. Interindividual variability in motor responses, such as spinal reflexes, activation patterns, or adaptive strategies should not be dismissed as experimental error, or eliminated by pooling data so that statistical analyses can be performed more easily. Although having a general idea of trends across individuals is necessary, so are changes that are specific to a smaller subset of the group, even at the individual level. This becomes of critical importance for rehabilitation following injury. If another treatment is more conducive to the recovery of motor functions in a small subset of the population than the “gold-standard,” then this treatment should be used. One of the great challenges will be in identifying a priori what treatment will be most effective for a particular individual or to design a method of quickly switching from one form of treatment to another.
Acknowledgments The present work was funded by a postdoctoral fellowship from the Christopher and Dana Reeve Foundation and by a postdoctoral fellowship from the Canadian Institutes of Health Research.
Abbreviations CPG FDL FHL LG MG MTP SCI TA Tib VL
central pattern generator flexor digitorum longus flexor hallucis longus lateral gastrocnemius medial gastrocnemius metatarsophalangeal spinal cord injury tibialis anterior tibial vastus lateralis
115
References Abraham, L. D., & Loeb, G. E. (1985). The distal hindlimb musculature of the cat. Patterns of normal use. Experimental Brain Research, 58, 580–593. Abraham, L. D., Marks, W. B., & Loeb, G. E. (1985). The distal hindlimb musculature of the cat. Cutaneous reflexes during locomotion. Experimental Brain Research, 58, 594–603. Barbeau, H., & Fung, J. (2001). The role of rehabilitation in the recovery of walking in the neurological population. Current Opinion in Neurology, 14, 735–740. Barriere, G., Leblond, H., Provencher, J., & Rossignol, S. (2008). Prominent role of the spinal central pattern generator in the recovery of locomotion after partial spinal cord injuries. The Journal of Neuroscience, 28, 3976–3987. Bizzi, E., Tresch, M. C., Saltiel, P., & d'Avella, A. (2000). New perspectives on spinal motor systems. Nature Reviews. Neuroscience, 1, 101–108. Bizzi, E., Cheung, V. C., d'Avella, A., Saltiel, P., & Tresch, M. (2008). Combining modules for movement. Brain Research Reviews, 57, 125–133. Bouyer, L. J. G., & Rossignol, S. (2003). Contribution of cutaneous inputs from the hindpaw to the control of locomotion: 2. Spinal cats. Journal of Neurophysiology, 90, 3640–3653. Bouyer, L. J. G., Whelan, P., Pearson, K. G., & Rossignol, S. (2001). Adaptive locomotor plasticity in chronic spinal cats after ankle extensors neurectomy. The Journal of Neuroscience, 21, 3531–3541. Bucher, D., Prinz, A. A., & Marder, E. (2005). Animal-to-animal variability in motor pattern production in adults and during growth. The Journal of Neuroscience, 25, 1611–1619. Butz, M., Worgotter, F., & van, O. A. (2009). Activity-dependent structural plasticity. Brain Research Reviews, 60, 287–305. Cai, L. L., Courtine, G., Fong, A. J., Burdick, J. W., Roy, R. R., & Edgerton, V. R. (2006). Plasticity of functional connectivity in the adult spinal cord. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 361, 1635–1646. Caperchione, C. M., Kolt, G. S., & Mummery, W. K. (2009). Physical activity in culturally and linguistically diverse migrant groups to Western society: A review of barriers, enablers and experiences. Sports Medicine, 39, 167–177. Carrier, L., Brustein, L., & Rossignol, S. (1997). Locomotion of the hindlimbs after neurectomy of ankle flexors in intact and spinal cats: Model for the study of locomotor plasticity. Journal of Neurophysiology, 77, 1979–1993. Cheung, V. C., d'Avella, A., Tresch, M. C., & Bizzi, E. (2005). Central and sensory contributions to the activation and organization of muscle synergies during natural motor behaviors. The Journal of Neuroscience, 25, 6419–6434.
Cote, M. P., & Gossard, J. P. (2004). Step training-dependent plasticity in spinal cutaneous pathways. The Journal of Neuroscience, 24, 11317–11327. Cote, M. P., Menard, A., & Gossard, J. P. (2003). Spinal cats on the treadmill: Changes in load pathways. The Journal of Neuroscience, 23, 2789–2796. Degtyarenko, A. M., Simon, E. S., & Burke, R. E. (1996). Differential modulation of disynaptic cutaneous inhibition and excitation in ankle flexor motoneurons during fictive locomotion. Journal of Neurophysiology, 76, 2972–2985. Delcomyn, F. (1980). Neural basis of rhythmic behavior in animals. Science, 210, 492–498. Dietz, V., Colombo, G., Jensen, L., & Baumgartner, L. (1995). Locomotor capacity of spinal cord in paraplegic patients. Annals of Neurology, 37, 574–582. Drew, T., & Rossignol, S. (1985). Forelimb responses to cutaneous nerve stimulation during locomotion in intact cats. Brain Research, 329, 323–328. Duysens, J., & Stein, R. B. (1978). Reflexes induced by nerve stimulation in walking cats with implanted cuff electrodes. Experimental Brain Research, 32, 213–224. Eccles, R. M., & Lundberg, A. (1959). Supraspinal control of interneurons mediating spinal reflexes. Journal de Physiologie, 147, 565–584. Edgerton, V. R., Leon, R. D., Harkema, S. J., et al. (2001). Retraining the injured spinal cord. Journal de Physiologie, 533, 15–22. English, A. W. (1979). Interlimb coordination during stepping in the cat: An electromyographic analysis. Journal of Neurophysiology, 42, 229–243. Fleshman, J. W., Lev-Tov, A., & Burke, R. E. (1984). Peripheral and central control of flexor digitorium longus and flexor hallucis longus motoneurons: The synaptic basis of functional diversity. Experimental Brain Research, 54, 133–149. Forssberg, H. (1985). Ontogeny of human locomotor control: I. Infant stepping, supported locomotion and transition to independent locomotion. Experimental Brain Research, 57, 480–493. Forssberg, H., & Svartengren, G. (1983). Hardwired locomotor network in cat revealed by a retained motor pattern to gastrocnemius after muscle transposition. Neuroscience Letters, 41, 283–288. Forssberg, H., Grillner, S., Halbertsma, J., & Rossignol, S. (1980). The locomotion of the low spinal cat: II. Interlimb coordination. Acta Physiologica Scandinavica, 108, 283–295. Frigon, A., & Gossard, J. P. (2009). Asymmetric control of cycle period by the spinal locomotor rhythm generator in the adult cat. Journal de Physiologie, 587, 4617–4628. Frigon, A., & Rossignol, S. (2006). Functional plasticity following spinal cord lesions. Progress in Brain Research, 157, 231–260.
116 Frigon, A., & Rossignol, S. (2007). Plasticity of reflexes from the foot during locomotion after denervating ankle extensors in intact cats. Journal of Neurophysiology, 98, 2122–2132. Frigon, A., & Rossignol, S. (2008a). Adaptive changes of the locomotor pattern and cutaneous reflexes during locomotion studied in the same cats before and after spinalization. Journal de Physiologie, 586, 2927–2945. Frigon, A., & Rossignol, S. (2008b). Locomotor and reflex adaptation after partial denervation of ankle extensors in chronic spinal cats. Journal of Neurophysiology, 100, 1513–1522. Frigon, A., & Rossignol, S. (2009). Partial denervation of ankle extensors prior to spinalization in cats impacts the expression of locomotion and the phasic modulation of reflexes. Neuroscience, 158, 1675–1690. Frigon, A., Barriere, G., Leblond, H., & Rossignol, S. (2009). Asymmetric changes in cutaneous reflexes after a partial spinal lesion and retention following spinalization during locomotion in the cat. Journal of Neurophysiology, 102, 2667–2680. Gorassini, M. A., Norton, J. A., Nevett-Duchcherer, J., Roy, F. D., & Yang, J. F. (2009). Changes in locomotor muscle activity after treadmill training in subjects with incomplete spinal cord injury. Journal of Neurophysiology, 101, 969–979. Grasso, R., Ivanenko, Y. P., Zago, M., et al. (2004). Distributed plasticity of locomotor pattern generators in spinal cord injured patients. Brain, 127, 1019–1034. Grillner, S. (1981). Control of locomotion in bipeds, tetrapods, and fish. In J. M. Brookhart & V. B. Mountcastle (Eds.), Handbook of physiology. The nervous system II (pp. 1179–1236). Bethesda, MD: The American Physiological Society. Grillner, S., & Wallen, P. (1985). Central pattern generators for locomotion, with special reference to vertebrates. Annual Review of Neuroscience, 8, 233–261. Grillner, S., Halbertsma, J., Nillsson, J., & Thorstensson, A. (1979). The adaptation to speed in human locomotion. Brain Research, 165, 177–182. Halbertsma, J. M. (1983). The stride cycle of the cat: The modelling of locomotion by computerized analysis of automatic recordings. Acta Physiologica Scandinavica, 521 (Suppl.), 1–75. Haridas, C., Zehr, E. P., & Misiaszek, J. E. (2005). Postural uncertainty leads to dynamic control of cutaneous reflexes from the foot during human walking. Brain Research, 1062, 48–62. Harkema, S. J. (2001). Neural plasticity after human spinal cord injury: Application of locomotor training to the rehabilitation of walking. The Neuroscientist, 7, 455–468. Harkema, S. J. (2008). Plasticity of interneuronal networks of the functionally isolated human spinal cord. Brain Research Reviews, 57, 255–264.
Holmqvist, B., & Lundberg, A. (1959). On the organization of the supraspinal inhibitory control of interneurons of various reflex arcs. Archives Italiennes de Biologie, 97, 340–356. Holtmaat, A., & Svoboda, K. (2009). Experience-dependent structural synaptic plasticity in the mammalian brain. Nature Reviews. Neuroscience, 10, 647–658. Ivanenko, Y. P., Cappellini, G., Dominici, N., Poppele, R. E., & Lacquaniti, F. (2007). Modular control of limb movements during human locomotion. The Journal of Neuroscience, 27, 11149–11161. Jordan, L. M. (1991). Brainstem and spinal cord mechanisms for the initiation of locomotion. In M. Shimamura, S. Grillner & V. R. Edgerton (Eds.), Neurobiological basis of human locomotion (pp. 3–20). Tokyo: Japan Scientific Societies Press. Juvin, L., Simmers, J., & Morin, D. (2007). Locomotor rhythmogenesis in the isolated rat spinal cord: A phase-coupled set of symmetrical flexion extension oscillators. Journal de Physiologie, 583, 115–128. Krouchev, N., Kalaska, J. F., & Drew, T. (2006). Sequential activation of muscle synergies during locomotion in the intact cat as revealed by cluster analysis and direct decomposition. Journal of Neurophysiology, 96, 1991–2010. Kumanyika, S. K. (2008). Environmental influences on childhood obesity: Ethnic and cultural influences in context. Physiology & Behavior, 94, 61–70. Levinsson, A., Holmberg, H., Broman, J., Zhang, M., & Schouenborg, J. (2002). Spinal sensorimotor transformation: Relation between cutaneous somatotopy and a reflex network. The Journal of Neuroscience, 22, 8170–8182. Loeb, G. E. (1993). The distal hindlimb musculature of the cat: Interanimal variability of locomotor activity and cutaneous reflexes. Experimental Brain Research, 96, 125–140. Loeb, G. E. (1999). Asymmetry of hindlimb muscle activity and cutaneous reflexes after tendon transfers in kittens. Journal of Neurophysiology, 82, 3392–3405. Loeb, E. P., Giszter, S. F., Saltiel, P., Bizzi, E., & MussaIvaldi, F. A. (2000). Output units of motor behavior: An experimental and modeling study. Journal of Cognitive Neuroscience, 12, 78–97. Lundberg, A. (1967). The supraspinal control of transmission in spinal reflex pathways. Electroencephalography and Clinical Neurophysiology, Suppl, 25, 35–46. Lundberg, A. (1981). Half-centres revisited. In J. Szentagothai, M. Palkovits & J. Hamori (Eds.), Regulatory functions of the CNS. Principles of motion and organization. Advances Physiology Sciences, (Vol. 1, pp. 155–167). Budapest: Pergamon Press. Marder, E., & Goaillard, J. M. (2006). Variability, compensation and homeostasis in neuron and network function. Nature Reviews. Neuroscience, 7, 563–574. McCrea, D. A., & Rybak, I. A. (2008). Organization of mammalian locomotor rhythm and pattern generation. Brain Research Reviews, 57, 134–146.
117 Musselman, K. E., & Yang, J. F. (2007). Loading the limb during rhythmic leg movements lengthens the duration of both flexion and extension in human infants. Journal of Neurophysiology, 97, 1247–1257. Norton, J. A., & Gorassini, M. A. (2006). Changes in cortically related intermuscular coherence accompanying improvements in locomotor skills in incomplete spinal cord injury. Journal of Neurophysiology, 95, 2580–2589. O'Donovan, M. J., Pinter, M. J., Dum, R. P., & Burke, R. E. (1982). Actions of FDL and FHL muscles in intact cats: Functional dissociation between anatomical synergists. Journal of Neurophysiology, 47, 1126–1143. O'Donovan, M. J., Pinter, M. J., Dum, R. P., & Burke, R. E. (1985). Kinesiological studies of self- and cross-reinnervated FDL and soleus muscles in freely moving cats. Journal of Neurophysiology, 54, 852–866. Pearson, K. G. (2000). Neural adaptation in the generation of rhythmic behavior. Annual Review of Physiology, 62, 723–753. Pearson, K. G., Fouad, K., & Misiaszek, J. E. (1999). Adaptive changes in motor activity associated with functional recovery following muscle denervation in walking cats. Journal of Neurophysiology, 82, 370–381. Prinz, A. A., Bucher, D., & Marder, E. (2004). Similar network activity from disparate circuit parameters. Nature Neuroscience, 7, 1345–1352. Rossignol, S. (1996). Neural control of stereotypic limb movements. In L. B. Rowell & J. T. Sheperd (Eds.), Handbook of physiology, Section 12. Exercise: Regulation and integration of multiple systems (pp. 173–216). New York: Oxford University Press. Rossignol, S. (2006). Plasticity of connections underlying locomotor recovery after central and/ or peripheral lesions in the adult mammals. Philosophical transactions of the Royal Society of London. Series B, Biological Sciences, 361, 1647–1671. Rossignol, S., Dubuc, R., & Gossard, J. P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86, 89–154. Rossignol, S., Barriere, G., Frigon, A., et al. (2008). Plasticity of locomotor sensorimotor interactions after peripheral and/or spinal lesions. Brain Research Reviews, 57, 228–240. Rossignol, S., Barriere, G., Alluin, O., & Frigon, A. (2009). Re-expression of locomotor function after partial spinal cord injury. Physiology (Bethesda), 24, 127–139. Schmidt, B. J., Meyers, D. E. R., Fleshman, J. L., Tokuriki, M., & Burke, R. E. (1988). Phasic modulation of short latency cutaneous excitation in flexor digitorum longus motoneurons during fictive locomotion. Experimental Brain Research, 71, 568–578. Schomburg, E. D. (1990). Spinal sensorimotor systems and their supraspinal control. Neuroscience Research, 7, 265–340.
Schouenborg, J. (2002). Modular organisation and spinal somatosensory imprinting. Brain Research. Brain Research Reviews, 40, 80–91. Schouenborg, J. (2003). Modular organisation and spinal somatosensory imprinting. Brain Research Reviews, 40, 80–91. Schouenborg, J. (2008). Action-based sensory encoding in spinal sensorimotor circuits. Brain Research Reviews, 57, 111–117. Schulz, D. J., Goaillard, J. M., & Marder, E. (2006). Variable channel expression in identified single and electrically coupled neurons in different animals. Nature Neuroscience, 9, 356–362. Seefeldt, V., Malina, R. M., & Clark, M. A. (2002). Factors affecting levels of physical activity in adults. Sports Medicine, 32, 143–168. Shimamura, M., Tanaka, I., & Livingston, R. B. (1991). Longitudinal conduction systems serving spinal and brainstem coordination (spino-bulbo-spinal reflex). In M. Shimamura, S. Grillner & V. R. Edgerton (Eds.), Neurobiological basis of human locomotion (pp. 241–255). Tokyo: Japan Scientific Societies Press. Stein, P. S. G., & Smith, J. L. (1997). Neural and biochemical control strategies for different forms of vertebrate hindlimb locomotor tasks. In P. S. G. Stein, S. Grillner, A. I. Selverston & D. G. Stuart (Eds.), Neurons, networks and motor behavior (pp. 61–73). Computational neurosciences series. Cambridge, MA: MIT Press. Stuart, D. G., Withey, T. P., Wetzel, M. C., & Goslow, G. E. J. (1973). Time contraints for inter-limb co-ordination in the cat during unrestrained locomotion. In R. B. Stein, K. G. Pearson, R. S. Smith & J. B. Redford (Eds.), Control of posture and locomotion (pp. 537–560). New York: Plenum Press. Vilensky, J. A., & Gankiewicz, E. (1990). Effects of growth and speed on hindlimb joint angular displacement patterns in vervet monkeys (Cercopithecus aethiops). American Journal of Physical Anthropology, 81, 441–449. Vilensky, J. A., & Patrick, M. C. (1984). Inter and intratrial variation in cat locomotor behavior. Physiology & Behavior, 33, 733–743. Vilensky, J. A., Gankiewicz, E., & Townsend, D. W. (1988). Effects of size on vervet (Cercopithecus aethips) gait parameters: A cross-sectional approach. American Journal of Physical Anthropology, 76, 463–480. Vilensky, J. A., Gankiewicz, E., & Townsend, D. W. (1990). Effects of size on vervet (Cercopithecus aethiops) gait parameters: A longitudinal approach. American Journal of Physical Anthropology, 81, 429–439. Wernig, A., Muller, S., Nanassy, A., & Cagol, E. (1995). Laufband therapy based on ‘rules of spinal locomotion’ is effective in spinal cord injured persons. The European Journal of Neuroscience, 7, 823–829.
118 Wetzel, M. C., Atwater, A. E., Wait, J. V., & Stuart, G. G. (1975). Neural implications of different profiles between treadmill and overground locomotion timings in cats. Journal of Neurophysiology, 38, 492–501. Wolpaw, J. R. (2007). Spinal cord plasticity in acquisition and maintenance of motor skills. Acta Physiologica (Oxford), 189, 155–169. Wolpaw, J. R., & Tennissen, A. M. (2001). Activity-dependent spinal cord plasticity in health and disease. Annual Review of Neuroscience, 24, 807–843. Yang, J. F., Stephens, M. J., & Vishram, R. (1998). Infant stepping: A method to study the sensory control of human walking. Journal de Physiologie, 507(Pt. 3), 927–937.
Yang, J. F., Lam, T., Pang, M. Y., Lamont, E., Musselman, K., & Seinen, E. (2004). Infant stepping: A window to the behaviour of the human pattern generator for walking. Canadian Journal of Physiology and Pharmacology, 82, 662–674. Yang, J. F., Lamont, E. V., & Pang, M. Y. (2005). Split-belt treadmill stepping in infants suggests autonomous pattern generators for the left and right leg in humans. The Journal of Neuroscience, 25, 6869–6876. Zehr, E. P., Komiyama, T., & Stein, R. B. (1997). Cutaneous reflexes during human gait: Electromyographic and kinematic responses to electrical stimulation. Journal of Neurophysiology, 77, 3311–3325.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 8
Challenging the adaptive capacity of rhythmic movement control: From denervation to force field adaptation Laurent J. Bouyer{,{,* {
Center for Interdisciplinary Research in Rehabilitation and Social Integration (CIRRIS), Department of Rehabilitation, Université Laval, Quebec, Canada { Member of the Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Quebec City, Quebec, Canada
Abstract: The neural control of walking involves voluntary descending drive, automatic rhythm and patterngenerating circuits, and sensory feedback to produce appropriate motor output. This control system has to be both robust and adaptable to remain appropriately calibrated to the changes in body size and in environmental demands that occur throughout life. In this chapter, current experimental models that are used to study the adaptive capacity of rhythmic movement control will be presented. Overall, while walking is a complex movement requiring extremely well-timed muscle activation sequences, and considering the presence of automatic rhythm generating circuits, its neural control nevertheless shows a large potential for adaptive modification. Regardless if the need for motor output modification is of internal (e.g., denervations) or external (e.g., changes in environment dynamics) origin, the system copes with the challenge rapidly and efficiently. Neural structures involved in adaptation are distributed, and even reduced preparations such as low spinal cats show extensive adaptive capacity. The degree of adaptive capacity is not unlimited, however. Functional flexors cannot be turned into extensors, and vice versa. In addition, recent evidence suggest that adaptive capacity may be dependent on the timing in the movement where adaptation is required (phase dependency), some phases being more amendable to change than others. Clearly, while important progress has been achieved using denervations and motor adaptation protocols, many questions remain to be answered regarding the mechanisms underlying adaptation and retention of adapted motor output, as well as regarding how sensory inputs are used to trigger adaptation. Recent advances in robotics, together with the design of simple, yet clever protocols such as catch trials are very promising tools to provide more answers. Keywords: adaptation; locomotion; force field; denervation; neural control; robotics. *Corresponding author. Tel.: þ1-418-529-9141x6661; Fax: þ1-418-529-3548 DOI: 10.1016/B978-0-444-53825-3.00013-9
119
120
Introduction Breathing, walking, and chewing are three rhythmic movements essential for survival. While not intuitively obvious, the neural control of these three apparently very distinct movements share a large amount of similarities. All three use the following elements to produce the final motor output: voluntary descending control, automatic rhythm and pattern-generating circuits, and sensory feedback. Rhythmic movement controllers have a long phylogenetical history, demonstrating their robustness. However, to be efficient they also need to be adaptable, that is, have the ability to cope with daily as well as evolutionary demands. The goal of this chapter is to review our current understanding of this adaptive capacity, and of the neural mechanisms underlying it. Considering that locomotion received by far the most attention on this topic, it will be the focus of this chapter.
Adjusting rhythmic motor output over the lifespan: A need for adaptive capacity When observing a human or animal walk, movement always seems fluid and well controlled. Growing from children to adults, this fluidity remains regardless of the transformations in body mass and length of body segments. Executing complex rhythmic movements requires extremely well-coordinated patterns of activation of several muscle groups on a timescale of milliseconds in order to produce an appropriate level of force at the functionally appropriate moment. Therefore, for movements to remain fluid despite the changes in musculoskeletal metrics and dynamics, the stereotyped muscle activation pattern producing walking (motor output) has to be constantly adjusted, a process called motor adaptation. Motor adaptation can be considered as a form of continuous motor learning, with the goal of keeping movement optimally “tuned.” Its theoretical limit is the capacity for change of the
neural control circuitry generating the rhythmic behavior, a process called adaptive capacity. Understanding the extent of adaptive capacity is of great interest in the field of motor control and rehabilitation, as it can provide a framework for developing more efficient locomotor rehabilitation protocols.
Combined feedforward and feedback control: Implications for adaptive capacity Before considering locomotor movement adaptation, it is important to understand how this movement is neurally controlled. The generation of the locomotor muscle activation pattern involves voluntary commands, central pattern generators, and sensory feedback. As several in-depth reviews have been published recently (e.g., Rossignol et al., 2006; Chapter 16), only a very brief overview will be presented here. The most basic level of control comes from a spinal pattern generator (Grillner, 1981; MacKay-Lyons, 2002) that produces a very detailed, muscle-specific spatio-temporal pattern similar to what is observed during actual walking (Grillner and Zangger, 1979). In addition to its involvement in the timing of muscle activations, this spinal network also modulates sensory afferent activity and interneuron excitability and is therefore actively involved in the sensorimotor integration that occurs during walking (Grillner, 1981; Rossignol, 1996; Rossignol et al., 2006). This automatic level of control does not act in isolation, however. The final muscle activation pattern is also influenced by sensory feedback from joint (Grillner and Rossignol, 1978), muscle (Dietz and Duysens, 2000; Duysens and Pearson, 1980; Pearson, 1995), and skin receptors (Duysens and Pearson, 1976; Rossignol et al., 1988) as well as by higher brain centers (reviewed in Armstrong, 1988), including postural control elements (Nashner, 1983). The muscle activation pattern is adjusted to the different physical situations such as walking with a heavy backpack (load
121
compensation implemented by sensory feedback; reviewed in Dietz and Duysens, 2000) or negotiating obstacles (voluntary gait modification implemented by descending commands; reviewed in Armstrong, 1988; Patla, 1996). The control of locomotion is a very efficient system, as the autonomous management of muscle activations allows the movement to be executed without requiring full attention from the subject, thereby allowing him to perform other activities, such as planning his walking path. From a general point of view, locomotion can therefore be considered to involve two types of neural control: feedforward and feedback. Feedforward control is what the central nervous system (CNS) components (voluntary drive and pattern generators) provide to the muscle activation pattern. This includes the “predictive” or planned aspect of the movement. Feedback control is what sensory feedback components provide to the muscle activation pattern. In the case of locomotion, both positive and negative feedback loops contribute to muscle activation pattern shaping. Negative feedback uses afferent discharges from peripheral receptors to reduce the error on the ongoing movement. A stretch reflex is a good example of negative feedback control: if a limb is deviated from its planned trajectory during an ongoing movement, given muscles will be stretched, muscle spindles will be activated, and afferent feedback will be used to increase motoneuron firing, thereby shortening the muscle and returning the limb toward its planned trajectory. As far as movement control is concerned, negative feedback requires a movement error to be activated. Positive feedback is quite different. A good example of positive feedback control is a loading reflex: in cats, it has been demonstrated that during walking, when antigravity hindlimb muscles such as soleus/gastrocnemius (ankle extensors) contract, the load increase created in the muscle tendon activates Golgi tendon organs (GTOs) and afferent feedback from GTOs is used to increase motoneuron firing (Gossard
et al., 1994). In this case, no movement error is present: positive feedback assists force generation during normal movement execution. Furthermore, this assistance varies automatically depending on limb loading, that is, will be greater when the animal carries a load and lesser if it is unloaded. While separating feedforward from positive feedback control is important to further our understanding of normal movement control, it becomes essential when one wants to study the neural mechanisms underlying motor adaptation/ motor learning, as learning requires modifications in feedforward (i.e., predictive) control. Positive feedback is also a powerful mode of control. Its reactive nature will make it come “on” and “off” as needed. However, being reactive to sensory feedback limits its versatility: for example, it cannot be used to preload a muscle in the expectation of an upcoming dynamic event such as landing from a fall/jump. Furthermore, motor output adjustments generated by positive feedback will only transfer to other situations associated with similar sensory feedback, and as such is not very useful to train new motor patterns for rehabilitation purposes. On the contrary, feedforward adaptation requiring modifications in central predictive elements will have a better chance of transferring to other tasks, due to a reduced need for continuous sensory feedback in this mode of control.
Central pattern generators: Implications for adaptive capacity While changes in feedforward control to update the locomotor muscle activation pattern may seem to be a relatively simple task for the CNS, the fact that CPGs are involved in movement control could potentially complicate the situation and reduce adaptive capacity. Indeed, CPGs are designed to produce autonomously a stereotyped muscle activation pattern. Such interconnected interneuronal networks create specific restricted synergies at different moments of the step cycle (Grillner et al.,
122
1998; McCrea and Rybak, 2007). While this coupling is not absolute (e.g., burst deletions can occur; McCrea and Rybak, 2007), it is not random either (Carlson-Kuhta et al., 1998; Smith et al., 1998; Trank et al., 1996). For the sake of this chapter, it is important to keep in mind that CPGs follow rules of activation (Prochazka et al., 2002; Yakovenko et al., 2002), and therefore changing the drive to one motoneuron pool for adaptive reasons may have consequences on other, maybe more remote motoneuron pools due to network properties. As a consequence, adaptation in central drive must take this factor into consideration.
protocol (Gorassini et al., 1994). In freely walking cats, when motor output from ankle extensors measured during normal gait is compared to a situation where the support surface is rapidly removed just before the paw is expected to contact the ground, the initial part of ankle extensor activity normally involved in weight acceptance is identical in both cases. However, 50 ms after expected ground contact, ankle extensor activity rapidly goes down when the support surface is absent. The initial part of the MG burst is therefore considered to be centrally generated (feedforward control), while the rest of the burst is largely due to positive feedback from load receptors (feedback control) (Fig. 1).
Assessing the relative contribution of positive feedback and feedforward control mechanisms Research questions The presence of positive feedback complicates the assessment of the relative contributions of feedforward and feedback control to motor output during normal movement. Sensory feedback has to be unexpectedly removed in order to isolate the feedforward contribution to motor output. A good example is the “foot in the hole” (a)
Considering the distributed neural control of rhythmical movements, including the presence of a CPG and of positive feedback control, how much adaptive capacity can we expect in motor output? Furthermore, what would be the relative contributions of feedforward and feedback
(b) 0.5
Normal
130 ms
LG EMG (normalized)
−10 ms
Foot in hole
n = 29 –100
0 Time (ms)
100
200
Fig. 1. Foot in the hole protocol. (a) Line drawings illustrating the procedure. A force sensor built into the trap door detects forelimb contacts and trigger door opening at the end of forelimb stance. (b) Average rectified EMG activity of lateral gastrocnemius when the cat steps into the hole (thick line; n ¼ 29) superimposed on EMG obtained during normal stepping (thin line; n ¼ 29). The initial part of the burst is the same in both situations. Modified from Gorassini et al. (1994).
123
mechanisms to motor adaptation? Finally, regarding changes in feedforward control, what would be the CNS structures involved/necessary for this process to occur? To address these important questions, two experimental approaches are presented below: partial denervations in animal models and motor adaptations in humans.
Partial denervation model One approach to the study of adaptive capacity of the control of rhythmic movements that has received considerable attention over the last century is the model of partial denervation. In this model, one (or more) peripheral nerve is (are) transected and prevented from growing back, in an otherwise intact animal. The rhythmic movement is documented before and repeatedly after the denervation. Surgically sectioning a mixed nerve interrupts sensory afferents (reducing sensory feedback) and motor axons (producing partial paralysis), effectively changing the way the affected body part can be controlled. A motor deficit (movement error) is observed when the animal tries to move after the denervation, as the usual motor command is no longer appropriate to execute the movement. For recovery of function to be observed, modifications in feedforward control must be implemented by the nervous system, a process referred to as adaptive capacity/plasticity. Several denervation models have been used over the years (see Bouyer and Rossignol, 2001; Rossignol et al., 2004, 2008 for reviews) including cutaneous and motor nerve transections and nerve transpositions.
Extent and limits of adaptive capacity Denervation experiments have shown that quite extensive recovery can be seen after experimentally induced modifications in sensorimotor processing (Bouyer and Rossignol, 2003a; Carrier
et al., 1997; Pearson et al., 1999; Whelan and Pearson, 1997). Time course of locomotor recovery occurs over a few days, that is, faster than the time needed for muscle tissue growth (see Bouyer et al., 2001), thereby demonstrating that recovery resulted from neural modifications. Adaptive capacity has limits, however. Muscle/ nerve crossing experiments have shown that antagonistic muscles cannot have their locomotor activation pattern fully reversed. For example, even 3 years after MG/TA nerve crossing, the output of MG during walking remains that of an extensor, discharging during stance, while its mechanical action is now into swing (Forssberg and Svartengren, 1983). In an extensive review of the older literature, Sperry (1945) demonstrated that this was also true for other movements and that the reported recovery at the time may in fact have been due to compensation by nondenervated agonists rather than by a change in activation patterns in the experimentally transposed muscles. Examples of this limit have also been reported very early in the human literature (Close and Todd, 1959; Sutherland et al., 1960).
Underlying mechanisms In addition to being easy to implement and control, the denervation model is very well suited to study some of the neural mechanisms underlying adaptive capacity/plasticity. Using successive peripheral and central lesions, animal models have also shown that adaptation involves structures distributed across the nervous system both at supraspinal and spinal sites (Carrier et al., 1997; Whelan and Pearson, 1997). Furthermore, reduced preparations such as spinal cats demonstrate that substantial adaptive capacity remains even in the “disconnected” spinal cord (Bouyer and Rossignol, 2003b; Bouyer et al., 2001). An extreme example is shown in Fig. 2 where a chronic spinal cat trained to walk on a treadmill undergoes a lateral gastrocnemius and soleus (LGS) nerve section and fully recovers
124 (a)
(b)
Foot lift (8 days)
Foot contact
Pre neurectomy (2 days)
160
hip
(control)
Hip
ankle
Foot contact
Foot lift
Joint angle (⬚)
knee 110 170
Knee 110 160
Ankle
2 days post neurectomy
90
Yield control 2 days 8 days
cat3 0.4 m/s
0
0.5
1.0
Phase of step cycle (c) 180
Cat 1 Cat 2 Cat 3
170
Ankle angle (⬚)
160
8 days post neurectomy
150 140 130 120
5 cm
110
5 cm
100 90 0
2
4
6
8 29 35
Days post neurectomy
Fig. 2. LGS denervation in the spinal cat. (a) Stick figure reconstructions of the left hindlimb before (top), 2 days (middle) and 8 days (bottom) after the neurectomy. (b) Details of the angular excursions during the step cycle obtained before (n ¼ 23; pale gray traces) and 2 days (n ¼ 11; black traces) and 8 days (n ¼ 9; dark gray traces) after the LGS neurectomy for cat 3. (c) Time course of ankle angle (S.D.) at a phase of 0.3 (representing maximum yield) superimposed for all cats. *P < 0.05. Modified from Bouyer et al. (2001).
walking within a week (Bouyer et al., 2001). This recovery time course is similar to that reported in nonspinal cats (Whelan and Pearson, 1997) and is implemented in the complete absence of supraspinal structures. Interestingly, the central modifications associated with locomotor recovery seem to be different between intact and spinal cats (Bouyer et al., 2001), a phenomenon that has since been observed in other situations (Frigon and Rossignol, 2009).
Limits of the denervation model While the denervation model is very useful to study some of the mechanisms underlying the adaptive capacity of central command to remodel during rhythmic behaviors, it also has limits. For example, after a muscle nerve transection, the fact that both the sensory and the motor components of the nerve have been interrupted imposes an additional constraint. In this case,
125
limits in locomotor recovery could be due to at least three nonmutually exclusive factors: (1) limits in actual central adaptive capacity; (2) limits in motor output due to muscle paralysis subsequent to the transection of motor axons; and (3) limits in error detection, as part of the richness of sensory feedback is removed. Now having a basic understanding of the time course and limits of adaptive locomotor capacity after denervation, is there a way to study adaptive capacity in an intact individual? The answer is yes, as detailed in the next section.
Modifying the environment to study adaptive capacity The concept of force field adaptation Over the past 15 years, a new approach involving experimental manipulation of the force environment in which movement is made (force fields) has been used to study the adaptive capacity of movement control without a need to alter the integrity of the nervous system of the individual under study. This approach was largely inspired by pioneer work originating in nonrhythmic movement control (Lackner and DiZio, 1994; Shadmehr and Mussa-Ivaldi, 1994).
Original studies in the upper limb When naïve subjects make an initial reaching movement in a force field, their arm is deviated from the planned movement trajectory, thereby creating a movement error (difference between perturbed and planned trajectories). Classical theories of motor control suggest that this error is interpreted by the nervous system as a signal to recalibrate the movement (Ghez et al., 2000; Kawato, 1999; Lackner and DiZio, 2000; Loeb et al., 2000; Wolpert et al., 2001). Several movement executions are required before the movement becomes completely recalibrated.
Interestingly, modifications in the motor pattern persist temporarily after the force field is removed, demonstrating that feedforward control is modified as part of the motor adaptation (Lackner and DiZio, 1994; Shadmehr and Mussa-Ivaldi, 1994). Movement can be adapted to a large variety of force field directions (Ghez et al., 2000; Shadmehr and Moussavi, 2000; Shadmehr and Mussa-Ivaldi, 1994), and repeated exposure to the same force field leads to a more rapid compensation by the subject, that is, the motor system can “learn” new motor patterns using force fields (Joiner and Smith, 2008; Shadmehr and Holcomb, 1997; Thoroughman and Shadmehr, 2000). When subjects adapt to a force field in one region of the workspace, there is carryover of this adaptation to movements performed outside the training region (Shadmehr and Mussa-Ivaldi, 1994) when the two regions of the workspace require similar changes in muscle activity (Shadmehr and Moussavi, 2000; Thoroughman and Shadmehr, 1999).
Studies during rhythmic movement Recent studies show that when healthy, stroke, and spinal cord-injured participants walk on a treadmill while a force field is applied to the leg, motor adaptations are observed as previously shown for reaching. While this chapter focuses on force field adaptation, it must be noted that the latter represents only one of the available means of manipulating the environment in which the locomotor movement is executed. Other protocols exist, such as walking on split-belt treadmills and on rotating platforms, and produce similar results. In all cases, a movement error is initially present and gradually compensated for over the next several strides (e.g., Fig. 3). Once the force is removed, aftereffects are present and followed by a gradual return to baseline values (Blanchette and Bouyer, 2009; Bouyer et al., 2003; Choi and Bastian, 2007; Dietz et al., 1994; Emken and Reinkensmeyer,
126 (a)
(c)
6.0
FORCE FIELD
th
c sti
at
st
re
Velocity (m/s)
5.0
3.0 2.0
g len
1.0
a
El
4.0
0% 25%
0 0
250
500 0
50% 75%
(d)
100% Fmax
100 90 80 70 60 50 40 30 20 10 0
200
*
180
Velocity (% of baseline)
Force (N)
(b)
250
500 0
250
500
Stride numder
*
S1 S2 S3 S4 S5 S6 S7 S8 S9 S10
160 140 120 100 80 60
0
20 40 60 80 100 120 140 160 180 200
Baseline
Initial effect Adapted state
Elastic stretching (%)
Initial aftereffect
Baseline post
Epoch Fig. 3. Elastic force field adaptation protocol. (a) Experimental setup to produce the force field: elastic tubing was attached between the right foot and the frame of the treadmill. This arrangement generated a force that pulled the foot forward and up as shown schematically by the arrows. Force intensity was related to foot anteroposterior position, being largest around toe off. In addition, elastic stretching during stance did not require active work, but resulted from the backward movement of the loaded lower limb with the treadmill belt. (b) Force-stretching relationship for a 40-cm elastic tubing, that is, the average elastic length used in the study. (c) Peak foot velocity during swing for each stride of the three walking periods in a representative subject (S1). (d) Normalized (% of baseline) peak right foot velocity during swing for the five epochs analyzed. *P < 0.05. Modified from Blanchette and Bouyer (2009).
2005; Fortin et al., 2009; Gordon and Ferris, 2007; Gordon et al., 1995; Jensen et al., 1998; Kao and Ferris, 2009; Lam et al., 2006, 2008; Noble and Prentice, 2006; Noel et al., 2009; Prokop et al., 1995; Reisman et al., 2007, 2009; Sawicki et al., 2006; Weber et al., 1998). These findings are indicative that the adaptive capacity of the neural control of rhythmic movement can be studied in humans using noninvasive protocols that
simply manipulate the walking environment in a controlled manner. By effectively altering limb dynamics, force field adaptation protocols are in many ways similar to what animals and humans have to go through when growing up (changes in body weight, in limb segment length, etc.) and also when aging (reduction in muscle force, etc.), and it is therefore believed to tap into similar underlying adaptive mechanisms.
127
Using modern robotics to create tailored force fields Within the past few years, several robotic devices and exoskeletons (Andersen and Sinkjaer, 1995; Blaya and Herr, 2004; Colombo et al., 2000; Emken and Reinkensmeyer, 2005; Gordon and Ferris, 2007; Noel et al., 2008) have been developed to study human locomotor adaptation/ learning (reviewed in Ferris, 2009). With advances in robotics, there is now little limit to the types and number of altered dynamics (i.e., force fields) that can be created to test this rhythmic movement. While most of the force field adaptation protocols have so far only been tested in humans, walking robots also exist for animal models such as rodents. De Leon et al. (2002) have shown motor adaptation in rats, opening a new field for future neurophysiological studies.
Using catch trials to identify modifications in feedforward control While force field exposure can be used to induce motor adaptations in rhythmic movements, the relative contribution of feedforward and feedback mechanisms to this process cannot be quantified by simply looking at muscle activation patterns during movement execution in the force field. Indeed, the mere presence of the force field will modify muscle loading, and hence also positive feedback contribution to motor output. While complete separation between feedforward and feedback control cannot be achieved using noninvasive methods, the modification in feedforward control can be estimated relatively easily by unexpectedly removing the force field for individual movement cycles (catch trials). In this situation, motor output used during the catch trial will involve the modified central command, but the altered muscle loading will not be available for positive feedback to influence motor output. Therefore, using a similar reasoning as was presented earlier for the “foot in the hole” protocol,
by comparing the catch trial motor output to control movement, the differences will reveal modifications in feedforward control (Lam et al., 2006; Noel et al., 2009).
Phase-dependent feedforward adaptive capacity Using a velocity-dependent resistance applied against hip and knee movement during the swing phase, Lam et al. (2006) compared the EMG activity of several muscles of the lower limb during catch trials (unexpected force removal during one stride) to baseline walking. They concluded that adapted muscle activation patterns in rectus femoris (knee extensor/hip flexor) and tibialis anterior (ankle dorsiflexor) during swing were controlled by feedback mechanisms (catch EMG activity was not different from baseline), while modifications in biceps femoris and medial hamstring (knee flexors/hip extensors) activity during preswing involved modifications in feedforward control (catch EMG was different from baseline). The authors suggested that the mechanisms underlying force field adaptation during locomotion were therefore different from one muscle group to another. However, there is an alternative hypothesis that could explain their results: considering that the resistance applied to knee flexion and knee extension did not occur at the same moment in the gait cycle, the difference in feedforward adaptation between the muscle groups studied may have depended on the timing in the gait cycle where adaptation had to occur. Indeed, neurophysiological studies have shown that central and peripheral drive to lower limb motoneurons is modulated as a function of the timing in the gait cycle (Bo Nielsen, 2002; Capaday and Stein, 1987; Duysens et al., 1992; Sinkjaer et al., 1996; Yang and Stein, 1990; Zehr et al., 1997). Therefore, it is very likely that feedforward and feedback contributions to the adapted motor output may not be constant throughout a stride. To test this hypothesis, Blanchette and Bouyer (2009) used a
128
position-dependent force field that modified the force environment over a wide portion of the gait cycle (preswing and swing phases). Their specific objectives were to describe the modifications in the walking pattern during and after force field exposure and to relate the muscle activation aftereffects to the modifications that occurred in the presence of the force field. In the adapted state, hamstring EMG activity started earlier and remained elevated throughout swing. After force field exposure, aftereffects in hamstring EMGs consisted of increased activity around toe off, but contrary to the adapted state, this increase was not maintained during the rest of swing (Fig. 4). These results suggest that adapted STANCE
350
SWING
EMG activity (% of baseline)
LH
250 150
** * *** *** *
50 0
*
180
*
* **
MH
120
*
* ***
60
******* **
*
0 0
20
40
60
80
100
% of stride Fig. 4. Phase-specific carryover of force field adaptation aftereffects. Group average (n ¼ 7) Lateral Hamstring (LH) and Medial Hamstring (MH) rectified EMG activity. Comparisons between adapted motor output (thick black line) and initial aftereffect (thick gray line) show that carryover after force field removal varied depending on the phase of the gait cycle. Error bars correspond to S.E. Data were time normalized to 60% for stance and 40% for swing. Data were also amplitude normalized to the peak locomotor EMG measured during baseline. *P < 0.05. From Blanchette and Bouyer (2009).
hamstring EMG activity may rely more on feedforward mechanisms around toe off and more on feedback mechanisms during the rest of swing thereby supporting the hypothesis of phasedependent adaptive capacity. In another recent study, Noel et al. (2009) used a robotized ankle exoskeleton (Noel et al., 2008) to apply short duration force fields during the stance phase of human walking and applied catch trials at pseudorandomly selected gait cycles within a 5-min adaptation period (Fig. 5). When exposed to a mid-stance force field, subjects initially showed a large movement error (increased ankle dorsiflexion velocity). Catch trials applied early into the force field exposure period presented a kinematic pattern identical to baseline, suggesting that feedforward control had not been modified yet. Subjects gradually adapted by returning ankle velocity to baseline over 50 strides. Catch trials applied thereafter showed decreased ankle velocity when compared to baseline, indicating the presence of feedforward adaptation. These results are compatible with the idea of a gradual updating of feedforward control based on recent experience. Noel et al. (2009) also exposed the same subjects to a similar force field, but this time presented during push-off (end of stance). Plantar flexion velocity was initially reduced at the moment of force field application, but no adaptation occurred over the 5-min exposure period. Catch trials kinematics remained similar to baseline at all times, suggesting that the same subjects that could easily adapt to a mid-stance force field had no adaptive capacity during the push-off phase of gait, at least over these short duration exposures. Taken together, the above studies suggest that while the neural control of rhythmic movement can adapt to force field exposure, the mechanisms underlying this adaptation may vary according to the timing in the gait cycle. In addition, these results show that robotized exoskeletons such as the Lokomat (Lam et al., 2006) and the EHO (Noel et al., 2009) are useful tools to study phase-dependent adaptive control of movement.
129
Applied torque (Nm)
WA
12 8
MS 1/2 FF
PO
(e)
Swing Baseline ± 2 S.T.D FF early FF late Last catch
4 0 –4
250
CONTROL
FORCE FIELD
POST
200 Ankle velocity (% baseline)
(a)
150 100 50
(b) Knee angular position (⬚)
80 60 40
0
0
0 100 200 Stride number
0
100
200
Ext
20 10
DF
0 –20
(f)
100
PF
Mean velocity Difference (Δ%)
–10
(d) Ankle velocity (⬚/s)
100
20
(c) Ankle angular position (⬚)
0 Flex
200 100 0 –100 –200
FORCE FIELD
POST
60 20 –20
*
–60 0 10 20 30 40 50 60 70 80 90 100 HS % Stride
CATCH
*
early
late
*
last 1st Epoch
early
late
Fig. 5. Example of force field adaptation with pseudorandomly inserted catch trials. (a) Torque applied on a subject's ankle by an electrohydraulic ankle foot orthosis (Noel et al., 2008) during mid-stance. Baseline (gray band) and force field late (thick black line). Outside of the force field application zone, the robotized orthosis applied a null field to minimize its influence on the subject's walking pattern. (b) Knee angular displacements superimposed for baseline (gray band), force field early (thin black line), force field late (thick black line), and last catch (dashed line). (c) Ankle angular displacements superimposed for baseline (gray band), force field early (thin black line), force field late (thick black line), and last catch (dashed line). (d) Ankle angular velocity for the same traces as in (c). Gray box: zone used for velocity measurement. Gray bands represent mean value 2 S.T.D. For all conditions, data were synchronized on heel strike. Abbreviations: WA, weight acceptance; MS, mid-stance; PO, push-off; DF, dorsiflexion; PF, plantar flexion; HS, heel strike. (E) Time course of ankle velocity across walking conditions. Each gray symbol represents a stride. Black symbols represent 11 points moving average. Open symbols represent catch strides; arrows focus on early and late catch trials velocity. (F) Group results (n ¼ 11) expressed as % difference from control for the two epochs in each walking condition. Error bars represent 95% confidence intervals. *: Epochs statistically different from baseline (P < 0.05; repeated measure ANOVA with Bonferonni correction). Modified from Noel et al. (2009).
130
Effects of force field training on aftereffects, next day performance, and retention of feedforward control parameters A review of the adaptive capacity of feedforward control of rhythmic movement could not be complete without touching upon two additional questions: (1) How long should one be exposed to force fields to obtain motor adaptation? (2) How much retention can be expected on subsequent exposures to the same force field? The first of these questions was recently addressed by Fortin et al. (2009). Subjects (n ¼ 17) were randomly assigned to one of several predetermined force field exposure durations (range 1–30 min, i.e., 49–1629 strides); the effects of exposure duration were measured on aftereffect duration and amplitude and on next day performance in the same force field. This study demonstrated that while aftereffect duration was correlated to exposure duration (Fig. 6), all subjects performed equally better on the 24-h retest. These results suggest that during walking, even short daily exposures to a force field ( 49 strides) lead to significant retention of the adapted feedforward parameters. They also suggest that neural consolidation mechanisms
General conclusion This chapter summarizes our current state of knowledge on the adaptive capacity of the neural control of rhythmic movements. Results show that despite the additional complexity brought about by automatic pattern-generating circuits, locomotion nevertheless shows a large potential for adaptive modification of its motor output. Regardless if the origin for the need for modification is internal (such as after nerve lesions) or external (such as due to changes in environment
40
50 40 30 20 10 0 80 60 Fo rce fiel 40 d (N) intens iit
0 y
2000 1500 on i 1000 urat d ) 500 re s su ide po (str x E
ity error Absolute veloc (Δ% CL)
Velocity aftereffect duration (strides)
60
triggered by force field exposure continue to act after the end of exposure. An example of the extent of memory of the adapted feedforward parameters and of our ability to rapidly recall the adapted muscle activation pattern is demonstrated in Fig. 7, where a nonnaïve subject is tested using a catch trial protocol similar to that of Fig. 5. This time, a change in catch trial performance was measured as early as 2–3 strides into the force field walking, a period way too short for a novel adaptation to occur, showing that feedforward control can be rapidly updated based on previous experience.
30 20 10 0 80 Fo rce 60 fiel 40 d (N) intens ity
0
2000 1500 on i 1000 urat d ) 500 e ur es os strid p ( Ex
Fig. 6. Effects of elastic force field adaptation duration and force field intensity on aftereffects duration. Each point represents one subject. It can be seen that aftereffect duration is proportional to force field exposure duration (left panel) and that aftereffect magnitude is proportional to force field intensity (right panel). From Fortin et al. (2009).
131 CONTROL
FORCE FIELD
POST
30
20
Ankle velocity error (deg/s)
10
0 -10 -20 -30 -40
0
100
0
200 100 Stride number
0
100
200
Fig. 7. Force field adaptation with catch trials (red circles) in a nonnaïve subject. When compared to Fig. 4e, it can be seen that in this case, ankle velocity error on the first catch trial (left arrow) was already as large as after 5 min of force field adaptation (right arrow). The nonnaïve subject anticipated well how to perform during force field walking. This is an indication of storage capacity of the previously adapted feedforward control parameters, that is, a sign of motor learning (Bouyer et al., unpublished observations).
dynamics), the system copes with the challenge rapidly and efficiently. Neural structures involved are distributed, but even reduced preparations such as low spinal cats still show extensive adaptive capacity. The adaptive capacity is not absolute, however. Functional flexors cannot be turned into extensors, and vice versa. In addition, recent evidence suggest that adaptive capacity may be dependent on the moment in the movement (phase specificity) where the adaptation is required, with some phases being more amendable to change than others. Clearly, while lots of progress have been achieved using models of denervation and motor adaptation, many questions remain to be answered, especially regarding the mechanisms underlying adaptation and retention, as well as regarding how sensory inputs are used to trigger adaptation. Recent advances in robotics together with the design of simple, yet clever protocols such as catch trials are very promising new tools to provide more
answers. Modern robotics open a new field of study with essentially unlimited types of force fields and the ability to introduce catch trials. Finally, on a more global perspective, the general finding that rhythmic movements are capable of substantial adaptation in the adult, and with relatively short training regimens is very positive for rehabilitation. Indeed, these findings provide key experimental evidence to support rhythmic movement rehabilitation training in patients. References Andersen, J. B., & Sinkjaer, T. (1995). An actuator system for investigating electrophysiological and biomechanical features around the human ankle joint during gait. IEEE Transactions on Rehabilitation Engineering, 3, 299–306. Armstrong, D. M. (1988). The supraspinal control of mammalian locomotion. Journal of Physiology, 405, 1–37. Blanchette, A., & Bouyer, L. J. (2009). Timing-specific transfer of adapted muscle activity after walking in an elastic force field. Journal of Neurophysiology, 102, 568–577.
132 Blaya, J. A., & Herr, H. (2004). Adaptive control of a variableimpedance ankle-foot orthosis to assist drop-foot gait. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 12, 24–31. Bo Nielsen, J. (2002). Motoneuronal drive during human walking. Brain Research. Brain Research Reviews, 40, 192–201. Bouyer, L., & Rossignol, S. (2001). Spinal cord plasticity associated with locomotor compensation to peripheral nerve lesions in the cat. In M. M. Patterson & J. W. Grau (Eds.), Spinal cord plasticity: Alterations in reflex function (pp. 207–224). Boston, MA: Kluwer Academic Publishers. Bouyer, L. J., & Rossignol, S. (2003a). Contribution of cutaneous inputs from the hindpaw to the control of locomotion. I. Intact cats. Journal of Neurophysiology, 90, 3625–3639. Bouyer, L. J., & Rossignol, S. (2003b). Contribution of cutaneous inputs from the hindpaw to the control of locomotion. II. Spinal cats. Journal of Neurophysiology, 90, 3640–3653. Bouyer, L. J., Whelan, P. J., Pearson, K. G., & Rossignol, S. (2001). Adaptive locomotor plasticity in chronic spinal cats after ankle extensors neurectomy. The Journal of Neuroscience, 21, 3531–3541. Bouyer, L. J., DiZio, P., & Lackner, J. R. (2003). Adaptive modifications of human locomotion by Coriolis force. In Proceedings of the 33rd annual meeting of the Society for Neuroscience. New Orleans, LA. Capaday, C., & Stein, R. B. (1987). Difference in the amplitude of the human soleus H reflex during walking and running. Journal of Physiology, 392, 513–522. Carlson-Kuhta, P., Trank, T. V., & Smith, J. L. (1998). Forms of forward quadrupedal locomotion. II. A comparison of posture, hindlimb kinematics, and motor patterns for upslope and level walking. Journal of Neurophysiology, 79, 1687–1701. Carrier, L., Brustein, E., & Rossignol, S. (1997). Locomotion of the hindlimbs after neurectomy of ankle flexors in intact and spinal cats: Model for the study of locomotor plasticity. Journal of Neurophysiology, 77, 1979–1993. Choi, J. T., & Bastian, A. (2007). Adaptation reveals independent control networks for human walking. Nature Neuroscience, 10, 1055–1062. Close, J. R., & Todd, F. N. (1959). The phasic activity of the muscles of the lower extremity and the effect of tendon transfer. The Journal of Bone and Joint Surgery, 41, 189–235. Colombo, G., Joerg, M., Schreier, R., & Dietz, V. (2000). Treadmill training of paraplegic patients using a robotic orthosis. Journal of Rehabilitation Research and Development, 37, 693–700. De Leon, R. D., Kubasak, M. D., Phelps, P. E., Timoszyk, W. K., Reinkensmeyer, D. J., Roy, R. R., et al. (2002). Using robotics to teach the spinal cord to walk. Brain Research. Brain Research Reviews, 40, 267–273.
Dietz, V., & Duysens, J. (2000). Significance of load receptor input during locomotion: A review. Gait & Posture, 11, 102–110. Dietz, V., Zijlstra, W., & Duysens, J. (1994). Human neuronal interlimb coordination during split-belt locomotion. Experimental Brain Research, 101, 513–520. Duysens, J., & Pearson, K. G. (1976). The role of cutaneous afferents from the distal hindlimb in the regulation of the step cycle of thalamic cats. Experimental Brain Research, 24, 245–255. Duysens, J., & Pearson, K. G. (1980). Inhibition of flexor burst generation by loading ankle extensor muscles in walking cats. Brain Research, 187, 321–332. Duysens, J., Tax, A. A., Trippel, M., & Dietz, V. (1992). Phase-dependent reversal of reflexly induced movements during human gait. Experimental Brain Research, 90, 404–414. Emken, J., & Reinkensmeyer, D. (2005). Robot-enhanced motor learning: Accelerating internal model formation during locomotion by transient dynamic amplification. IEEE Transactions on Neural Systems and Rehabilitation Engineering: A Publication of the IEEE Engineering in Medicine and Biology Society, 13, 33–39. Ferris, D. P. (2009). The exoskeletons are here. Journal of Neuroengineering and Rehabilitation, 6, 17. Forssberg, H., & Svartengren, G. (1983). Hardwired locomotor network in cat revealed by a retained motor pattern to gastrocnemius after muscle transposition. Neuroscience Letters, 41, 283–288. Fortin, K., Blanchette, A., McFadyen, B. J., & Bouyer, L. J. (2009). Effects of walking in a force field for varying durations on aftereffects and on next day performance. Experimental Brain Research, 199, 145–155. Frigon, A., & Rossignol, S. (2009). Partial denervation of ankle extensors prior to spinalization in cats impacts the expression of locomotion and the phasic modulation of reflexes. Neuroscience, 158, 1675–1690. Ghez, C., Krakauer, J. W., Sainburg, R. L., & Ghilardi, M. F. (2000). Spatial representations and internal models of limb dynamics in motor learning. In M. S. Gazzaniga (Ed.), The new cognitive neurosciences (pp. 501–514). Cambridge, MA: MIT Press. ISBN-10: 0-262-07195-9 ISBN-13: 978-0262-07195-6. Gorassini, M. A., Prochazka, A., Hiebert, G. W., & Gauthier, M. J. (1994). Corrective responses to loss of ground support during walking. I. Intact cats. Journal of Neurophysiology, 71, 603–610. Gordon, K. E., & Ferris, D. P. (2007). Learning to walk with a robotic ankle exoskeleton. Journal of Biomechanics, 40, 2636–2644. Gordon, C. R., Fletcher, W. A., Melvill, J. G., & Block, E. W. (1995). Adaptive plasticity in the control of locomotor trajectory. Experimental Brain Research, 102, 540–545.
133 Gossard, J. P., Brownstone, R. M., Barajon, I., & Hultborn, H. (1994). Transmission in a locomotor-related group Ib pathway from hindlimb extensor muscles in the cat. Experimental Brain Research, 98, 213–228. Grillner, S. (1981). Control of locomotion in bipeds, tetrapods, and fish. In J. M. Brookhart & V. B. Mountcastle (Eds.), Handbook of physiology. The nervous system II (pp. 1179–1236). Bethesda, MD: American Physiological Society. Grillner, S., & Rossignol, S. (1978). On the initiation of the swing phase of locomotion in chronic spinal cats. Brain Research, 146, 269–277. Grillner, S., & Zangger, P. (1979). On the central generation of locomotion in the low spinal cat. Experimental Brain Research, 34, 241–261. Grillner, S., Ekeberg, Ö., El Manira, A., Lansner, A., Parker, D., Tegner, J., & Wallen, P. (1998). Intrinsic function of a neuronal network—A vertebrate central pattern generator. Brain Research Brain Research Reviews, 26, 184–197. Jensen, L., Prokop, T., & Dietz, V. (1998). Adaptational effects during human split-belt walking: Influence of afferent input. Experimental Brain Research, 118, 126–130. Joiner, W. M., & Smith, M. A. (2008). Long-term retention explained by a model of short-term learning in the adaptive control of reaching. Journal of Neurophysiology, 100, 2948–2955. Kao, P. C., & Ferris, D. P. (2009). Motor adaptation during dorsiflexion-assisted walking with a powered orthosis. Gait & Posture, 29, 230–236. Kawato, M. (1999). Internal models for motor control and trajectory planning. Current Opinion in Neurobiology, 9, 718–727. Lackner, J. R., & DiZio, P. (1994). Rapid adaptation to Coriolis force perturbations of arm trajectory. Journal of Neurophysiology, 72, 299–313. Lackner, J. R., & DiZio, P. A. (2000). Aspects of body self-calibration. Trends in Cognitive Sciences, 4, 279–288. Lam, T., Anderschitz, M., & Dietz, V. (2006). Contribution of feedback and feedforward strategies to locomotor adaptations. Journal of Neurophysiology, 95, 766–773. Lam, T., Wirz, M., Lunenburger, L., & Dietz, V. (2008). Swing phase resistance enhances flexor muscle activity during treadmill locomotion in incomplete spinal cord injury. Neurorehabilitation and Neural Repair, 22, 438–446. Loeb, E. P., Giszter, S. F., Saltiel, P., Bizzi, E., & MussaIvaldi, F. A. (2000). Output units of motor behavior: An experimental and modeling study. Journal of Cognitive Neuroscience, 12, 78–97. MacKay-Lyons, M. (2002). Central pattern generation of locomotion: A review of the evidence. Physical Therapy, 82, 69–83. McCrea, D. A., & Rybak, I. A. (2007). Modeling the mammalian locomotor CPG: Insights from mistakes and perturbations. Progress in Brain Research, 165, 235–253.
Nashner, L. M. (1983). Analysis of movement control in man using the movable platform. Advances in Neurology, 39, 607–619. Noble, J., & Prentice, S. (2006). Adaptation to unilateral change in lower limb mechanical properties during human walking. Experimental Brain Research, 169, 482–495. Noel, M., Cantin, B., Lambert, S., Gosselin, C. M., & Bouyer, L. J. (2008). An electrohydraulic actuated ankle foot orthosis to generate force fields and to test proprioceptive reflexes during human walking. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 16, 390–399. Noel, M., Fortin, K., & Bouyer, L. J. (2009). Using an electrohydraulic ankle foot orthosis to study modifications in feedforward control during locomotor adaptation to force fields applied in stance. Journal of Neuroengineering and Rehabilitation, 6, 16. Patla, A. E. (1996). “Neurobiomechanical bases for the control of human locomotion,” in Clinical Aspects of Balance and Gait Disorders, A. Bronstein, T. Brandt, and M. Woollacott, Eds. London: Arnold, pp. 19–40A. Pearson, K. G. (1995). Proprioceptive regulation of locomotion. Current Opinion in Neurobiology, 5, 786–791. Pearson, K. G., Fouad, K., & Misiaszek, J. E. (1999). Adaptive changes in motor activity associated with functional recovery following muscle denervation in walking cats. Journal of Neurophysiology, 82, 370–381. Prochazka, A., Mushahwar, V., & Yakovenko, S. (2002). Activation and coordination of spinal motoneuron pools after spinal cord injury. Progress in Brain Research, 137, 109–124. Prokop, T., Berger, W., Zijlstra, W., & Dietz, V. (1995). Adaptational and learning processes during human split-belt locomotion: Interaction between central mechanisms and afferent input. Experimental Brain Research, 106, 449–456. Reisman, D. S., Wityk, R., Silver, K., & Bastian, A. (2007). Locomotor adaptation on a split-belt treadmill can improve walking symmetry post-stroke. Brain, 130, 1861–1872. Reisman, D. S., Wityk, R., Silver, K., & Bastian, A. J. (2009). Split-belt treadmill adaptation transfers to overground walking in persons poststroke. Neurorehabilitation and Neural Repair, 23, 735–744. Rossignol, S. (1996). Neural control of stereotypic limb movements. In L. B. Rowell & J. T. Sheperd (Eds.), Handbook of physiology, section 12. Exercise: Regulation and integration of multiple systems (pp. 173–216). New York: Oxford University Press. Rossignol, S., Lund, J. P., & Drew, T. (1988). The role of sensory inputs in regulating patterns of rhythmical movements in higher vertebrates. A comparison between locomotion, respiration, and mastication. In A. Cohen, S. Rossignol & S. Grillner (Eds.), Neural control of rhythmic movements in vertebrates (pp. 201–283). New York: John Wiley and Sons, Inc. ISBN 0471819689 (ISBN13: 9780471819684).
134 Rossignol, S., Brustein, E., Bouyer, L., Barthelemy, D., Langlet, C., & Leblond, H. (2004). Adaptive changes of locomotion after central and peripheral lesions. Canadian Journal of Physiology and Pharmacology, 82, 617–627. Rossignol, S., Dubuc, R., & Gossard, J. P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86, 89–154. Rossignol, S., Barriere, G., Frigon, A., Barthelemy, D., Bouyer, L., Provencher, J., et al. (2008). Plasticity of locomotor sensorimotor interactions after peripheral and/ or spinal lesions. Brain Research Reviews, 57, 228–240. Sawicki, G. S., Domingo, A., & Ferris, D. P. (2006). The effects of powered ankle-foot orthoses on joint kinematics and muscle activation during walking in individuals with incomplete spinal cord injury. Journal of Neuroengineering and Rehabilitation, 3, 3. Shadmehr, R., & Holcomb, H. H. (1997). Neural correlates of motor memory consolidation. Science, 277, 821–825. Shadmehr, R., & Moussavi, Z. M. (2000). Spatial generalization from learning dynamics of reaching movements. The Journal of Neuroscience, 20, 7807–7815. Shadmehr, R., & Mussa-Ivaldi, F. A. (1994). Adaptive representation of dynamics during learning of a motor task. The Journal of Neuroscience, 14, 3208–3224. Sinkjaer, T., Andersen, J. B., & Larsen, B. (1996). Soleus stretch reflex modulation during gait in humans. Journal of Neurophysiology, 76, 1112–1120. Smith, J. L., Carlson-Kuhta, P., & Trank, T. V. (1998). Forms of forward quadrupedal locomotion. III. A comparison of posture, hindlimb kinematics, and motor patterns for downslope and level walking. Journal of Neurophysiology, 79, 1702–1716. Sperry, R. W. (1945). The problem of central nervous reorganization after nerve regeneration and muscle transposition. The Quarterly Review of Biology, 20, 311–369.
Sutherland, D. H., Bost, F. C., & Schottstaedt, E. R. (1960). Electromyographic study of transplanted muscles about the knee in poliomyelitic patients. The Journal of Bone and Joint Surgery, 42, 919–939. Thoroughman, K. A., & Shadmehr, R. (1999). Electromyographic correlates of learning an internal model of reaching movements. The Journal of Neuroscience, 19, 8573–8588. Thoroughman, K. A., & Shadmehr, R. (2000). Learning of action through adaptive combination of motor primitives. Nature, 407, 742–747. Trank, T. V., Chen, C., & Smith, J. L. (1996). Forms of forward quadrupedal locomotion. I. A comparison of posture, hindlimb kinematics, and motor patterns for normal and crouched walking. Journal of Neurophysiology, 76, 2316–2326. Weber, K. D., Fletcher, W. A., Gordon, C. R., Melvill, J. G., & Block, E. W. (1998). Motor learning in the “podokinetic” system and its role in spatial orientation during locomotion. Experimental Brain Research, 120, 377–385. Whelan, P. J., & Pearson, K. G. (1997). Plasticity in reflex pathways controlling stepping in the cat. Journal of Neurophysiology, 78, 1643–1650. Wolpert, D. M., Ghahramani, Z., & Flanagan, J. R. (2001). Perspectives and problems in motor learning. Trends in Cognitive Sciences, 5, 487–494. Yakovenko, S., Mushahwar, V., VanderHorst, V., Holstege, G., & Prochazka, A. (2002). Spatiotemporal activation of lumbosacral motoneurons in the locomotor step cycle. Journal of Neurophysiology, 87, 1542–1553. Yang, J. F., & Stein, R. B. (1990). Phase-dependent reflex reversal in human leg muscles during walking. Journal of Neurophysiology, 63, 1109–1117. Zehr, E. P., Komiyama, T., & Stein, R. B. (1997). Cutaneous reflexes during human gait: Electromyographic and kinematic responses to electrical stimulation. Journal of Neurophysiology, 77, 3311–3325.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 9
Face sensorimotor cortex neuroplasticity associated with intraoral alterations Limor Avivi-Arber{,*, Jye-Chang Lee{ and Barry J. Sessle{ { {
Department of Prosthodontics, Faculty of Dentistry, University of Toronto, Ontario, Canada Department of Oral Physiology, Faculty of Dentistry, University of Toronto, Ontario, Canada
Abstract: Loss of teeth or dental attrition is a common clinical occurrence associated with altered somatosensation and impaired oral motor behavior (e.g., mastication, deglutition, phonation). Oral rehabilitation aims at restoring these sensorimotor functions to improve patients' quality of life. Recent studies have implicated neuroplastic changes within the primary motor cortex (M1) in the control of limb motor behaviors following manipulations of sensory inputs to or motor outputs from the central nervous system as well as in learning and adaptation processes. However, limited data are available of the neuroplastic capabilities of face-M1 in relation to orofacial motor functions. The overall objective of our series of studies was to use intracortical microstimulation (ICMS) and recordings of evoked muscle electromyographic activity to test if neuroplastic changes occur in the ICMS-defined motor representations of the tongue-protrusive (genioglossus, GG) and jaw-opening (anterior digastric, AD) muscles within the rat face-M1 and adjacent face primary somatosensory cortex (face-S1) following several different types of intraoral manipulations. We found that a change in diet consistency was not associated with statistically significant changes in AD and GG motor representations. However, incisor extraction resulted, one week later, in a significantly increased AD representation within the contralateral face-M1 and face-S1, and incisor trimming produced timedependent changes in the AD motor representation. These novel findings underscore the neuroplastic capabilities of the face sensorimotor cortex and point to its possible role in adaptation to an altered peripheral state or altered sensorimotor behavior. Further insights into the neuroplastic capabilities of the face sensorimotor cortex promise to improve therapeutic strategies aimed at the restoration of oral functions, particularly in patients suffering from orofacial sensorimotor deficits or pain. Keywords: dental; intracortical microstimulation; orofacial; muscles.
*Corresponding author. Tel.: þ1-416-979-4900x4618; Fax: þ1-416-979-4900x4936 E-mail:
[email protected] DOI: 10.1016/B978-0-444-53825-3.00014-0
135
136
Introduction In addition to the well-documented role of subcortical mechanisms in the genesis and control of chewing (as well as breathing and walking), recent studies suggest that the primary motor cortex (M1) also plays an important role in chewing and other semiautomatic orofacial motor functions (e.g., swallowing and rhythmic whisking of the vibrissae) as well as in the control of elemental orofacial movements such as jaw opening and tongue protrusion (for review, see Chapter 5). Analogous studies have revealed that the somatosensory system, including the face primary somatosensory cortex (face-S1), is not just involved in perceptional processes but may also assist in the generation and modulation of orofacial movements (Ebner, 2005; Murray et al., 2001; Chapter 5). These studies include those using the intracortical microstimulation (ICMS) technique to map the functional organization of cortical motor outputs, as well as single neuron recordings to define the functional organization of orofacial somatosensory inputs and mechanoreceptive fields and the movement-related activities of neurons within the faceM1 and face-S1. In subprimates and primates, short-train ICMS can evoke elemental jaw, tongue, and facial movements when applied to face-M1 and also when applied to face-S1, and long-train ICMS can evoke rhythmic jaw movements, swallowing, or vibrissal whisking when applied to face-M1 and face-S1 (e.g., Burish et al., 2008; Gioanni and Lamarche, 1985; Hiraba et al., 1997; Huang et al., 1989a; Lund et al., 1984; Martin et al., 1999; Neafsey et al., 1986; Sapienza et al., 1981). Single unit recordings reveal that face-M1 as well as face-S1 neurons receive somatosensory inputs from the orofacial region (e.g., Farkas et al., 1999; Henry and Catania, 2006; Iyengar et al., 2007; Kaas et al., 2006; Murray et al., 2001; Remple et al., 2003; Sapienza et al., 1981). These studies have also demonstrated that the orofacial region has extensive somatosensory and motor representations that often overlap within the face sensorimotor cortex.
One striking finding of the many studies using ICMS or single neuron recordings of the limb and vibrissal motor or somatosensory representations within M1 and S1, respectively, is that these representations are dynamic and can be substantially remodeled during development and in adulthood by altered somatosensory inputs or motor outputs (Ebner, 2005; Frostig, 2006; Kaas et al., 2008; Navarro et al., 2007; Sessle, 2009; Toldi, 2008). Reorganization of the representations has been associated with increased muscle use or disuse (i.e., use-dependent neuroplasticity) and may also occur in association with the training of animals and humans in a novel motor behavior and thereby contribute to mechanisms underlying the acquisition and retention of learned motor skills (Ebner, 2005; Guic et al., 2008; Kleim and Jones, 2008; Kleim et al., 2004; Nudo, 2003; Remple et al., 2001). Reorganization may also occur following injuries peripherally (e.g., amputation, deefferentation, deafferentation) (Huntley, 1997; Lotze et al., 1999; Merzenich et al., 1984; Tandon et al., 2009; Toldi et al., 1996; Veronesi et al., 2006) or centrally (e.g., stroke involving the cortex) (Samuels and Feske, 2003). It may be manifested as increased representations of the M1 or S1 regions neighboring the deprived or damaged cortical areas and may play a crucial role in the subsequent recovery of these sensorimotor functions (Hallett, 2001; Kleim and Jones, 2008; Lotze et al., 1999; Nudo, 2003; Rossignol, 2006; Teasell et al., 2006). However, limited published data are available of the neuroplastic capabilities of face sensorimotor cortex following intraoral manipulations, as noted below. No study has addressed whether neuroplastic changes occur in the ICMS-defined motor representations within face-M1 following intraoral manipulations such as altering the dental occlusion by tooth extraction or trimming, or a change in diet consistency. Furthermore, no study has tested whether peripheral manipulations result in neuroplastic changes in the orofacial motor representations within face-S1 despite
137
existing evidence of neuroplastic changes within face-S1 somatosensory representations following incisor extraction in mole rats (Henry et al., 2005) and despite the evidence of the role that face-S1 plays in the modulation of orofacial motor functions (see Chapter 5; Murray et al., 2001). Such information is of clinical significance, since loss of teeth or dental attrition is a common occurrence, as are other conditions such as pain, nerve injury, and stroke affecting orofacial functions; such conditions are often associated with altered somatosensation or impaired mastication, deglutition, and phonation which may impact the quality of life of the affected individuals (Brennan et al., 2008; Haas and Lennon, 1995; Martin, 2009). Moreover, besides their effect on nutritional intake and general health, loss of teeth and reduced masticatory function, particularly in elderly patients, have been associated with impaired cognitive functions (e.g., learning and memory) and implicated in neurological disorders such as Alzheimer's disease (Bergdahl et al., 2007; Sakamoto et al., 2009). Since novel therapies targeting limb motor dysfunction as a result of peripheral injuries, spinal cord injuries, or stroke have taken advantage of the documented neuroplastic capacity of limb sensorimotor cortex to promote recovery of limb motor function in animal models as well as in humans (Kleim and Jones, 2008; Lotze et al., 1999; Nudo, 2003; Rossignol, 2006), it is possible that oral rehabilitation of patients with altered sensorimotor function produces its beneficial effects, at least in part, by utilizing the neuroplastic capabilities of the face sensorimotor cortex. Thus, the purpose of the present studies was to determine whether neuroplastic changes occur in the motor outputs from face-M1 and the adjacent face-S1 following various intraoral manipulations.
Objectives and methods To use ICMS and recordings of evoked muscle electromyographic (EMG) activity to test if
neuroplastic changes occur in the ICMS-defined motor representations of the tongue-protrusive (genioglossus, GG) and jaw-opening (anterior digastric, AD) muscles within the rat face-M1 or adjacent face-S1 following: 1. a change in diet consistency 2. mandibular incisor trimming 3. mandibular incisor extraction The next section summarizes the common methodology applied to the three different studies that is extensively detailed in Adachi et al. (2007) and Avivi-Arber et al. (2010b); therefore, only a brief outline follows. Rats were housed in individual cages containing a plastic tube used as a shelter and gnawing device. Water and food were available ad libitum. In each study (diet consistency, incisor trimming, incisor extraction), rats were divided into experimental, sham, and naïve groups. The manipulations provided to each study group are summarized in Table 1. Rats are usually kept on a hard diet from their weaning day; however, since extraction may impose some discomfort while chewing on the hard diet, the rats in the extraction and sham extraction groups (see Table 1) were kept on soft diet (mashed chow) from their arrival day at the vivarium. Since diet consistency may affect masticatory functions (Inoue et al., 2004; Okayasu et al., 2003; Proschel and Hofmann, 1988), we also tested whether a change in diet consistency has any effect on the jaw and tongue motor representations, and therefore, one group was kept on the hard (chow) diet while the other group was given the soft diet. Within each study, animals were monitored regularly to assess body weight, food consumption, and any postoperative complications to ensure continuous and similar growth and normal behavior. All dental procedures were carried out under general anesthesia, and incisor extraction was supplemented by local anesthesia. Rat incisors normally erupt continuously at a rate of 1–2 mm/
138 Table 1. The study groups, the number of rats within each study group, the intraoral manipulation, and the time of the ICMS mapping experiment. Data from Avivi-Arber et al., 2010a,b Experiment Diet
Incisor trimming
Incisor extraction
Group a
Soft diet (n ¼ 6) Hard diet (n ¼ 6) Naive (n ¼ 7) Trim (n ¼ 7) Sham trim (n ¼ 7) Trim recovered (n ¼ 7) Extraction (n ¼ 8) Sham extraction (n ¼ 6)
Diet
Manipulation
ICMS mapping
Soft
Neither general anesthesia nor any dental treatment
Following 2–3 weeks on soft diet
Hard
Neither general anesthesia nor any dental treatment
Hard
Neither general anesthesia nor incisor trimming
Hard
General anesthesia, bilateral trimming of the incisal edge of the lower incisors every 2 days 4 times General anesthesia but no trimming every 2 days
Following 2–3 weeks on hard diet Following 1–2 weeks after arrival 1 day following last trimming day 1 day following last sham trimming day 7 days following single trimming day and regaining of normal occlusal contacts 7 days following extraction
Hard Hard
General anesthesia, bilateral trimming of the incisal edges of the lower incisors
Soft
General and local anesthesia and surgical extraction of right lower incisor General and local anesthesia, surgical procedure but no actual tooth extraction
Soft
7 days following sham extraction
a
The soft diet group served as a naïve control group in the extraction experiments.
day (Burnmurdoch, 1995; Sessle, 1966) and their diminutive pulpal innervation terminates 2 mm away from the incisal edge (Paxinos, 2004; Zhang and Deschenes, 1998). Therefore, to keep the mandibular incisors in a state of reduced occlusal contacts with no pulp exposure, trimming of the incisal edge of both mandibular incisors was limited to 1–2 mm/day (see Lee et al., 2010; Table 1). Extraction of the right mandibular incisor was carried out under aseptic conditions and included full thickness mucoperiosteal flap elevation and labial bone trimming (see Avivi-Arber et al., 2010b; Table 1). ICMS mapping was carried out in rats to define the general organizational features of the ICMSdefined jaw and tongue motor representations within the face-M1 and adjacent face-S1. The mapping time for each study group is summarized in Table 1. Rats were maintained throughout the ICMS experiments under a stable level of general anesthesia with ketamine HCL. EMG electrodes were used to record activity from the left or right
AD (LAD, RAD), masseter, GG, vibrissal, and neck muscles. Systematic cortical mapping extended from 2.5–4.0 mm rostral to Bregma (i.e., anteroposterior (AP) planes 2.5, 3.0, 3.5, and 4.0) and 1.5–5.0 mm lateral to Bregma within the left and right face-M1 and with a horizontal spatial resolution of 0.5 mm (Fig. 1). At each penetration site, ICMS was applied every 0.2 mm of microelectrode penetration depth. Five ICMS trains (at 333 Hz, 12 pulses of 0.2 ms) were delivered at 1 Hz with maximum ICMS intensity of 60 mA. Electrolytic lesions placed at the bottom of each positive ICMS penetration were used for subsequent histological verification of the ICMS sites within the cytoarchitectonically defined gray matter of the granular (S1) or agranular (M1) cortex (for details, see Adachi et al., 2007; Avivi-Arber et al., 2010b). A customized software written in Spike2 script (CED, Cambridge Electronic Design, Cambridge, UK) and LabView (National Instruments, Austin, TX, USA) was used to analyze data files off-line.
139 (b) Centre of Gravity within face-Ml 2600 Extraction
Extraction Left
Sham-Extraction Right
No response GG LAD RAD
Left
Right
Depth (mm)
Motor Maps within face-M1 and face-S1
(a)
Sham-extraction
Naive
2800 3000 AP 2.5 3200 3400
2 3
5
3600 -6
-4
-2
0
2
4
6
2
4
6
2
4
6
-4 -2 0 2 4 Mediolateral distance (mm)
6
6
2600 2800
Depth (mm)
AP 4.0
3000 AP 3.0
3200 3400 3600 -6
AP 3.5
-4
-2
0
2600
Depth (mm)
2800
AP 3.0
3000 AP 3.5
3200 3400 3600 -6
-4
-2
0
2600
Depth (mm)
2800
AP 2.5
3000 3200
AP 4.0
3400
1 mm
3600 -6
Fig. 1. (a) Representative motor maps of LAD, RAD, and GG in a rat from the extraction group as compared with a rat from the sham extraction group. Any site where ICMS could evoke LAD, RAD, or GG EMG activity was plotted on templates from Swanson's maps of the rat brain (Brain Maps: Structure of the Rat Brain, third revised edition, 2004) corresponding to the coronal cross sections at AP planes 2.5–4.0 mm anterior to Bregma: LAD in gray circles, RAD in black circles, and GG in white circles. Black dots represent sites where ICMS could not evoke EMG activity in LAD, RAD, or GG muscles. Note the extensive, bilateral representation of LAD, RAD, and GG within S1 as well as M1. (b) Center of gravity by mapping plane. At AP 3.0 and AP 3.5, the center of gravity was located significantly more lateral in the extraction group than in the sham extraction and naïve groups. In all study groups, the center of gravity was positioned between AP3.0 and AP3.5 with no significant differences across the study groups or between left and right face-M1.
For each muscle, an ICMS site was defined and counted as a “positive ICMS site” if at least three out of the five ICMS trains could evoke an EMG response with an onset latency 40 ms and a peak activity exceeding the mean value of the initial 10 ms of the EMG response plus two standard
deviations. The number of positive ICMS sites outlined the extent of that muscle motor representation within the face sensorimotor cortex. At some sites, ICMS evoked EMG activity simultaneously in more than one muscle, and these sites were defined as overlapping representation
140
sites. The onset latency of the ICMS-evoked EMG response was also noted for each muscle at each positive ICMS site. The locations of the positive ICMS sites were marked on rat atlas templates corresponding to the coronal cross sections at AP planes 2.5–4.0 mm anterior to Bregma (Swanson, 2004) (e.g., see Fig. 1). The center of gravity, which defines the mean three-dimensional center position of the motor representations, was calculated for each of the LAD, RAD, and GG muscles by taking into account the mean number of positive ICMS sites obtained at each ML, AP, and depth coordinates, thereby providing the position of the motor maps weighted relative to the extent of the motor representation (Ridding et al., 2000). Statistical analyses used t-tests when only two groups were being compared and ANOVAs followed by post hoc tests when three groups were being compared, p < 0.05 indicating statistical significance.
Features of rat face-M1 and face-S1 motor representations The ICMS studies documented the general organizational features of the jaw-opening (AD) and tongue protrusion (GG) motor representations within the rat faceM1 and the adjacent face-S1. While ICMS within the mapped area evoked AD and GG EMG activity, masseter (jaw-closing) EMG activity was rarely observed, consistent with studies demonstrating a marked paucity of jaw-closing motor representation in monkeys and subprimates (e.g., Huang et al., 1988; Neafsey et al., 1986). Within the mapped area, which was limited to AD and GG motor representations, there were very few positive ICMS sites for the vibrissal and neck muscles, consistent with a more medial and caudal representations for these latter muscles (Brecht et al., 2004; Tandon et al., 2008). In each of the three studies summarized in this chapter, there were no significant differences between the sham and naïve groups in any of the study measures. AD and GG had large motor representations within the left and right face-M1
that spanned the entire depth of the cortical layers V–VI. AD and GG motor representations were characterized by multiple neighboring and intermingled representations that often overlapped, and LAD and RAD had a significant contralateral predominance (p < 0.01) (Table 2, Fig. 1). In addition, LAD and RAD had a significantly shorter onset latency within the contralateral faceM1 (left: LAD: 22.6 2.1(mean S.E.M), RAD: 14.2 1.0; right: LAD: 14.2 0.6, RAD: 21.2 1.6; p < 0.0001). These characteristics of jaw and tongue motor representations are consistent with previous studies in rats (Adachi et al., 2007; Donoghue and Wise, 1982; Gioanni and Lamarche, 1985; Neafsey et al., 1986; Sapienza et al., 1981; Tandon et al., 2008), monkeys (Burish et al., 2008; Huang et al., 1988, 1989b), and humans (Martin et al., 2004; Nordstrom, 2007). Overlapping of AD/GG motor representations was observed in a large proportion (42%) of the positive ICMS sites within face-M1. Overlapping motor representations within the limb-M1 along with a rich network of intracortical connectivity have been thought to reflect a “shared neural substrate,” whereby neurons within different representation areas are interconnected, and have been considered important for the spatiotemporal coordination of several muscles during movements (Aroniadou and Keller, 1993; Nudo et al., 1996; Sanes et al., 1995). Such overlapping may explain observations whereby long-train ICMS delivered to different sites within face-M1 can evoke different types of rhythmic masticatorylike jaw movements (see Chapter 5) and longtrain ICMS, matching the time course of the motor function being studied, can evoke coordinated and complex movements in space such as movements of the mouth, lips, and tongue toward a specific orofacial posture (Graziano and Aflalo, 2007). Noteworthy are our findings of LAD, RAD, and GG motor representations within face-S1. These EMG findings are in accord with earlier results showing that jaw and/or tongue movements can be evoked by short-train ICMS of face-S1 in rats and marmosets (e.g., Burish
141 Table 2. Number of positive ICMS sites within face-M1 and face-S1
M1
Cortical side
Muscle
Soft diet group
Hard diet group
Sham extraction group
Extraction group
Left
LAD RAD GG AD þ GG LAD RAD GG AD þ GG LAD RAD GG AD þ GG LAD RAD GG AD þ GG
18.3 4.4 23.3 5.2{ 21.3 6.2 13.8 4.2 25.5 3.7{ 9.8 3.2 18.0 3.9 13.0 2.6 3.0 1.7 4.8 1.3{ 10.0 4.7 2.5 1.7 9.3 2.9{ 3.0 1.6 9.3 3.1 3.7 1.6
19.2 1.9 24.7 1.6{ 13.5 4.4 10.5 2.7 25.3 1.8{ 8.7 2.2 10.5 3.2 8.2 2.1 5.8 2.1 8.8 4.1{ 8.2 4.3 3.7 1.6 13.7 4.6{ 4.3 1.8 7.3 2.7 4.7 1.6
13.5 3.5 23.2 4.2{ 26.3 12.6 14.7 5.2 23.8 2.5{ 23.8 2.5 20.3 7.4 12.5 3.0 2.3 0.8 6.7 1.7{ 5.2 2.0 3.0 1.5 9.7 5.5{ 3.7 1.6 7.5 3.1 3.8 2.0
21.5 3.2 51.3 4.6{,** 30.6 4.4 25.8 2.9 38.9 5.5{ 38.9 5.5 24.9 4.9 20.0 4.0 3.6 1.3 21.3 3.0{ 13.9 4.7 7.9 2.4 20.5 4.0{,* 3.6 1.0 13.8 5.8 6.1 2.7
Right
S1
Left
Right
The number of positive ICMS sites (mean SEM) obtained for each muscle or group of muscles within the left and right face-M1 or face-S1 of each of the study groups at 60 mA ICMS intensity. Within each group, LAD had significantly more sites within the right face-M1 or face-S1, and RAD had significantly more sites within the left face-M1 or face-S1 ({p < 0.05). There were no significant differences between the soft and hard diet groups (p > 0.05). Within the left face-M1 and face-S1, in comparison with the sham and naïve groups, the extraction group had significantly more RAD positive ICMS sites (*p < 0.05; **p < 0.01) (LAD, left anterior digastric; RAD, right anterior digastric; GG, genioglossus).
et al., 2008; Gioanni and Lamarche, 1985; Neafsey et al., 1986; Sapienza et al., 1981). It could be argued that the observed ICMSdefined motor representations within face-S1 in our studies were the result of activation of distant face-M1 neurons through current spread or transynaptic interactions (Cheney, 2002). However, several lines of evidence indicate that this is unlikely and that face-S1 does possess motor outputs (see also Avivi-Arber et al., 2010b). Some of the positive ICMS sites within face-S1 were positioned away from face-M1 (Fig. 1) and had short onset latencies comparable to those of LAD, RAD, and GG responses evoked from face-M1. In addition, under similar stimulation parameters, different intraoral manipulations had different yet specific and significant effects on the AD motor representations within the contralateral face-S1 comparable to the findings within face-M1 (see below). Finally, these findings of motor outputs from face-S1 are in agreement with anatomical studies showing efferent projections from S1 to
brainstem regions involved in orofacial motor control (Yoshida et al., 2009; Zhang and Sasamoto, 1990).
Effects of intraoral manipulation on face-M1 and face-S1 motor representations We found that a change in diet consistency for a period of 2–3 weeks had no significant (p > 0.05) effects on the extent or topographical organization of AD and GG motor representations within face-M1 and the adjacent face-S1 (Avivi-Arber et al., 2010a), as shown in Table 2. However, keeping the two mandibular incisors out of occlusal contacts for 1 week (by trimming every 2 days) significantly (p < 0.05) decreased the AD motor representation as reflected in the number of AD sites within the left and right face-M1 (Fig. 2); this decrease in AD sites was also associated with a shift of the AD and GG centers of gravity to a more superficial position (p < 0.05). Nevertheless,
142 Number of positive ICMS sites within face-sensorimotor cortex
Number of positive ICMS sites
160
GG AD
140 120 100
extraction and naive groups (p ¼ 0.033). Within face-S1, RAD onset latency was significantly shorter in the extraction group than in the sham extraction and naive groups (p ¼ 0.0016 and 0.029, respectively) and GG onset latency was significantly shorter in the extraction group than in the sham extraction group (p ¼ 0.033).
*
80
Implications of findings to sensorimotor behavior
60 40 20 0 Ctrl
M-Trim
R-Trim
Fig. 2. Effects of bilateral trimming of the mandibular incisors on the number of positive ICMS (60 mA) sites for GG and AD within the left and right face sensorimotor cortex (i.e., left and right face-M1 and S1). Note the significant (*p < 0.05) reduction in the number of AD sites in the trim group (M-Trim) as compared with the control group and trimrecovered (R-Trim) group. Each bar is the mean standard error for the group.
these changes were not evident after a single trim and 1 week of no trimming and regaining of normal occlusal contacts, as shown in Fig. 2 (trim-recovered group; p > 0.05) (Lee et al., 2010; Sessle et al., 2007). While bilateral trimming of the lower incisors resulted in a decreased bilateral representation of the AD, unilateral extraction of a lower incisor, as compared with sham extraction and naive groups, resulted, 1 week later, in a significant (p < 0.0001) increase in the RAD motor representation within the contralateral face-M1, along with a substantial (p < 0.05) lateral shift of the LAD and RAD centers of gravity and a significantly (p < 0.01) increased RAD motor representation within the contralateral face-S1 (Fig. 1, Table 2). While trimming, extraction, or diet consistency had no significant effect on GG motor representations within face-M1 or face-S1, extraction had a significant effect on GG onset latency within face-M1 that was significantly shorter in the extraction group as compared with the sham
The novel findings of the present studies demonstrate that several different intraoral manipulations can be associated with time-dependent neuroplastic changes within the face-M1 manifested as a directionally selective increase or decrease in the number of sites representing AD or GG. The studies also document for the first time that face-S1 motor outputs also have the capacity to undergo neuroplastic changes following peripheral manipulations. The importance of peripheral somatosensory feedback to the perception and control of jaw and tongue positions and movements (see Haas and Lennon, 1995; Murray et al., 2001) is reflected in the high innervation density of the oral tissues including the teeth and the projection of somatosensory afferents to the face-M1 as well as face-S1 (Iyengar et al., 2007; Miles et al., 2004; Paxinos, 2004; Sessle, 2009). Alterations in orofacial somatosensory inputs induced by sustained innocuous stimulation or transient or permanent peripheral deafferentation produced by peripheral nerve block or nerve transection, respectively, may result in neuroplastic changes within face-M1, as manifested in changes in cortical excitability and/ or altered motor representations (Adachi et al., 2007; Ebner, 2005; Franchi, 2001; Halkjaer et al., 2006; Hamdy et al., 1998; Yildiz et al., 2004). Therefore, it is possible that in the present studies, the reduced occlusal contacts induced by either dental extraction or trimming as well as extractioninduced deafferentation of tooth pulp and periodontal tissues may have altered the somatosensory inputs from the teeth to face-S1 and
143
face-M1 and contributed to the observed changes within face-M1, analogous to the occurrence of neuroplastic changes within limb-M1 following limb deafferentation (Ebner, 2005; Navarro et al., 2007; Sessle, 2009; Toldi, 2008). Limb deafferentation may also induce neuroplastic changes in the somatosensory representations within limb-S1 (Kaas et al., 2008; Toldi, 2008). Manipulations (e.g., trimming, lesioning, stimulation) of rodent vibrissae may induce reorganization of the vibrissal and neighboring receptive fields within face-S1 (e.g., Guic et al., 2008; Katz et al., 1999; for review, see Ebner, 2005; Frostig, 2006; Toldi, 2008) and extraction of the mandibular incisor in young mole rats results in expansion of neighboring orofacial somatosensory representation areas into the deprived incisor representation area (Henry et al., 2005). We have documented for the first time that incisor extraction may also induce neuroplastic changes in the ICMS-defined motor representations within face-S1. It has been well documented that experimental acute pain may result in decreased limb-M1 excitability, and chronic pain conditions including phantom limb pain have been associated with reorganization of sensory or motor representations within limb-S1 or limb-M1, respectively (e.g., Farina et al., 2001; Krause et al., 2006; Lotze et al., 1999; Vartiainen et al., 2009). Treatment with myoelectric prostheses results in re-reorganization of limb and lip representations and less phantom limb pain (Lotze et al., 1999). Experimental noxious stimulation induced by injection of the algesic glutamate into the tongue in rats or application of capsaicin to the tongue in healthy humans is also associated with decreased face-M1 excitability (Adachi et al., 2008; Boudreau et al., 2007). Oral surgery including dental extraction is associated with acute postoperative pain, and dental trimming may expose dentin tubules and tooth pulp and thereby render the tooth more sensitive to mechanical, thermal, and even noxious stimuli. Therefore, the possible confounding effect of postoperative pain in the trim and extraction groups cannot be excluded as a factor in the neuroplastic
changes observed in the present studies. However, this possibility is unlikely because we did not observe any pulp exposure in the incisor trimming experiments, and trimming was limited to 1–2 mm of the incisal edge that is believed to be free of pulpal axonal terminals (Paxinos, 2004; Zhang and Deschenes, 1998). In addition, neither in the trim nor in the extraction groups were there any obvious behavioral changes or changes in body weight to indicate that the rats were in pain (e.g., see Avivi-Arber et al., 2010b). Furthermore, the sham extraction group which also underwent intraoral surgery did not show any changes in the ICMS features as compared with the naïve rats. In addition, while experimentally induced intraoral noxious stimuli can induce decreased face-M1 excitability (see above) suggestive of a decreased motor representation (Ridding and Rothwell, 1997), incisor extraction resulted in increased motor representations. There is, however, another possible factor to consider. Rodents use their incisors for biting, fighting, and other oral motor functions, but gnawing and the associated dental attrition is a unique oral motor behavior that compensates for the continuous eruption of the incisors (Burnmurdoch, 1995; Sessle, 1966). Modification to the dental occlusion in rodents and humans induced by dental trimming or extraction can result in altered patterns of jaw and tongue movements during mastication (for review, see Klineberg and Jagger, 2004). Therefore, it is possible that dental trimming or extraction of the incisors also modified the rat's gnawing or other oral motor behaviors. It has been shown in monkeys that alterations in oral motor functions can affect the somatosensory inputs from the orofacial tissues involved in the motor functions (for review, see Murray et al., 2001). Such changes in somatosensory inputs may have contributed to changes in motor representations within the face-M1 (see above). In addition, changes in oral motor behavior may also contribute directly to the changes within face-M1, consistent with the concept of use-dependent
144
neuroplasticity whereby motor representations are altered by increased or decreased muscle use and by motor experience (Ebner, 2005; Kleim and Jones, 2008; Nudo, 2003; Remple et al., 2001). However, we did not record EMG or movement parameters during the animal's oral sensorimotor functions, so further studies are needed to determine if incisal trimming or extraction is associated with alterations in oral sensorimotor behavior. Changes in diet consistency may also be associated with altered motor behavior such as altered biting and chewing forces and different patterns of mastication (Inoue et al., 2004; Okayasu et al., 2003; Proschel and Hofmann, 1988). In contrast to the significant changes in AD or GG motor representations within faceM1 produced by dental extraction and trimming, a change in diet consistency had no significant effects. One possible explanation is that different forms of neuroplastic changes (e.g., changes in a specific muscle's motor representation) in different directions (i.e., increase, decrease, or no change) may be linked to different modes of peripheral manipulations. For example, in rodents, bilateral trimming of the vibrissae for 5 days has been reported to result in decreased vibrissal and increased limb motor representations within the vibrissal-M1 (Keller et al., 1996), whereas sensory denervation of the infraorbital nerve supplying sensory innervation to the vibrissae has been reported to result in a decreased excitability of the vibrissal-M1 but no apparent changes in motor representations (Franchi, 2001). In humans, it has been observed that lingual nerve block is associated with decreased excitability of the tongue motor representation within face-M1 (Halkjaer et al., 2006), whereas local anesthesia of the facial skin is associated with increased face-M1 excitability (Yildiz et al., 2004). Another possible factor contributing to the differential effects across our study groups could be related to time-dependent neuroplastic changes. Different but specific forms of neuroplastic
changes may occur at different time points following a peripheral manipulation. For example, unilateral transection of the lingual nerve results, 1–2 weeks later, in a significantly decreased GG representation within face-M1, and 3–4 weeks later, in a significantly increased GG representation (Adachi et al., 2007). Similar to our dental trimming data (see above), bilateral trimming of the vibrissae for 5 days has resulted in decreased vibrissal (and increased limb) motor representations that reversed and reestablished normal motor representations once the vibrissae are allowed to regrow for at least 3 days (Keller et al., 1996). Therefore, although the present ICMS study could not detect reorganization of orofacial motor representations following a change in diet consistency, the possibility cannot be ruled out that the underlying mechanisms required more time in order to be manifested as reorganization of motor representations. Also noteworthy is that one of the most consistent findings involving limb motor skill training studies is an increased overlapping representation within limb-M1 of the movements involved in the acquisition of the limb motor skill (Nudo et al., 1996). In addition, only limb motor skill training, but not nonskilled training, induces cortical reorganization within limb-M1 (Kleim et al., 2002; Remple et al., 2001). It is interesting to note that dental extraction resulted in a significantly increased RAD representation and RAD/GG overlapping representations within face-M1. Hence, it is possible that while dental extraction (or trimming) might have resulted in significant changes in oral motor behavior necessitating adoption of novel oral motor function (such as unilateral biting) and novel skilled coordination of jaw and tongue movements to compensate for the dental alterations, a change in diet consistency likely involved changes in nonskilled motor functions such as changes in the magnitude of forces applied during biting. Consequently, this could explain why dental extraction and trimming and not a change in diet consistency had a significant effect on face-M1 motor representations.
145
This explanation is consistent with the occurrence of face-M1 neuroplasticity that has been documented in humans and animals trained to acquire a novel tongue motor skill (e.g., Boudreau et al., 2007; Guggenmos et al., 2009; Sessle et al., 2007; Svensson et al., 2006).
Possible role of neuroplastic changes in other brain areas Reorganization of somatosensory inputs and motor outputs following peripheral manipulations is not limited to the sensorimotor cortex. The sensorimotor cortex receives afferent inputs from other central regions and face-M1 receives somatosensory inputs from the orofacial region indirectly through face-S1 as well as directly through the thalamus that relays the ascending orofacial inputs to face-M1 and face-S1 predominantly via the trigeminal brainstem sensory nuclei (Hatanaka et al., 2005; Henry and Catania, 2006; Iyengar et al., 2007; Miyashita et al., 1994; Sessle, 2009). Face-M1 and Face-S1 also project to thalamic and brainstem regions involved in orofacial somatosensory and motor functions (for review, see Hatanaka et al., 2005; Sessle, 2009). The brainstem sensory nuclei have been shown to undergo neuroplastic changes following tooth extraction or pulp extirpation (Hu et al., 1999; Kwan et al., 1993; Linden and Scott, 1989). Also, transection of the facial nerve supplying motor innervation to the vibrissae can induce motor reorganization not only just within the vibrissal-M1 but also within the brainstem motor nuclei (Kis et al., 2004). Therefore, although the present studies have demonstrated neuroplastic changes within face-M1 and face-S1 following dental extraction or trimming, we cannot rule out the possibility that changes may have also occurred in other cortical areas or in subcortical regions involved in orofacial sensorimotor function and contributed, at least in part, to the changes observed by applying ICMS at the cortical level. In addition, despite the lack of observed neuroplastic changes within face-M1 following a change
in diet consistency, changes may have occurred in other cortical or subcortical areas. A change in diet consistency may be associated with an altered masticatory function (see above) and it has been well documented that the cortical masticatory area (for review, see Sessle, 2009, chapter 5) and brainstem central pattern generator for chewing (for review, see Lund and Kolta, 2006; Sessle, 2009) play important roles in the generation and control of masticatory movements. Therefore, it is possible that these areas underwent changes that contributed to any altered oral motor behavior that may have occurred as a result of the change in diet consistency.
Clinical significance of the findings Intraoral manipulations can induce neuroplastic changes in the jaw or tongue motor representations not only within the face-M1 but also within the adjacent face-S1. Such cortical changes may play a role in the functional adaptation (or maladaptation) of the masticatory system to an altered oral state and be related to the ability of patients undergoing oral rehabilitation to restore (or not) their lost orofacial sensorimotor functions. This information is important since injuries to the oral tissues and modifications to the dental occlusion induced by dental extraction, attrition, or trimming are common clinical occurrences in humans as are many neurological disorders (e.g., brain injury, stroke, Parkinson disease) involving the orofacial regions. Such conditions may be accompanied by impaired sensorimotor functions that sometimes make the most vital functions of eating, swallowing, and speaking difficult and thereby negatively impact the patient's quality of life (Brennan et al., 2008; Haas and Lennon, 1995; Martin, 2009; Samuels and Feske, 2003). Treatment with implantsupported prostheses can significantly improve tactile capabilities and motor functions (e.g., Haraldson and Zarb, 1988) and a recent study has documented increased fMRI activity within
146
face-S1 of patients treated with implantsupported prostheses as compared with edentulous patients treated with conventional dentures (Yan et al., 2008). Regrettably, the cortical and subcortical neuroplastic mechanisms underlying orofacial sensorimotor functions, dysfunction, and rehabilitation have received very limited investigation in animals or humans. Further understanding of these neuroplastic mechanisms promises to lead to new or improved treatment strategies to facilitate recovery of patients suffering from pain and sensorimotor deficits and improve their quality of life. Indeed, in recent years, principles of M1 neuroplasticity have been translated into novel treatment strategies that induce cortical neuroplasticity or reverse redundant or undesirable neuroplastic changes in order to enhance the effectiveness of rehabilitation of patients suffering from sensorimotor disorders such as impaired deglutition and phonation, locomotion, and even phantom limb pain (Kleim and Jones, 2008; Lotze et al., 1999; Martin, 2009; Nudo, 2003; Robbins et al., 2008; Rossignol, 2006). For example, recent studies in rats (Adkins et al., 2008), monkeys (Frost et al., 2003), and humans (Brown et al., 2006) have shown that pairing rehabilitative training with cortical electrical stimulation induces better behavioral improvement than training alone. Our animal model has proven to be appropriate for further studies of face-M1 and face-S1 neuroplastic capabilities in rats, and complementary studies are being designed for monkeys and humans.
Acknowledgments This research was supported by grant MOP-4918 to B. J. S. from the Canadian Institutes of Health Research. The support of the Canadian Foundation for Innovation and Ontario Innovation Trust and Ministry of Research and Innovation is also gratefully acknowledged. B. J. S. is the holder of a Canada Research Chair.
Abbreviations AD AP EMG fMRI GG ICMS LAD LTD M1 ML RAD S1 SEM
anterior digastric anteroposterior electromyographic functional magnetic resonance imaging genioglossus intracortical microstimulation left anterior digastric long-term depression primary motor cortex mediolateral right anterior digastric primary somatosensory cortex standard error of the means
References Adachi, K., Lee, J. C., Hu, J. W., Yao, D., & Sessle, B. J. (2007). Motor cortex neuroplasticity associated with lingual nerve injury in rats. Somatosensory & Motor Research, 24, 97–109. Adachi, K., Murray, G. M., Lee, J. C., & Sessle, B. J. (2008). Noxious lingual stimulation influences the excitability of the face primary motor cerebral cortex (face MI) in the rat. Journal of Neurophysiology, 100, 1234–1244. Adkins, D. L., Hsu, J. E., & Jones, T. A. (2008). Motor cortical stimulation promotes synaptic plasticity and behavioral improvements following sensorimotor cortex lesions. Experimental Neurology, 212, 14–28. Aroniadou, V. A., & Keller, A. (1993). The patterns and synaptic properties of horizontal intracortical connections in the rat motor cortex. Journal of Neurophysiology, 70, 1553–1569. Avivi-Arber, L., Lee, J. C., & Sessle, B. J. (2010a). Cortical orofacial motor representation: Effect of diet consistency. Journal of Dental Research, 89(10), 1142–1147. Avivi-Arber, L., Lee, J. C., & Sessle, B. J. (2010b). Effects of incisor extraction on jaw and tongue motor representations within face sensorimotor cortex of adult rats. The Journal of Comparative Neurology, 518(7), 1030–1045. Bergdahl, M., Habib, R., Bergdahl, J., Nyberg, L., & Nilsson, L. G. (2007). Natural teeth and cognitive function in humans. Scandinavian Journal of Psychology, 48, 557–565. Boudreau, S., Romaniello, A., Wang, K., Svensson, P., Sessle, B. J., & Arendt-Nielsen, L. (2007). The effects of intra-oral pain on motor cortex neuroplasticity associated
147 with short-term novel tongue-protrusion training in humans. Pain, 132, 169–178. Brecht, M., Krauss, A., Muhammad, S., Sinai-Esfahani, L., Bellanca, S., & Margrie, T. W. (2004). Organization of rat vibrissa motor cortex and adjacent areas according to cytoarchitectonics, microstimulation, and intracellular stimulation of identified cells. The Journal of Comparative Neurology, 479, 360–373. Brennan, D. S., Spencer, A. J., & Roberts-Thomson, K. F. (2008). Tooth loss, chewing ability and quality of life. Quality of Life Research, 17, 227–235. Brown, J. A., Lutsep, H. L., Weinand, M., & Cramer, S. C. (2006). Motor cortex stimulation for the enhancement of recovery from stroke: A prospective, multicenter safety study. Neurosurgery, 58, 464–473. Burish, M. J., Stepniewska, I., & Kaas, J. H. (2008). Microstimulation and architectonics of frontoparietal cortex in common marmosets (Callithrix jacchus). The Journal of Comparative Neurology, 507, 1151–1168. Burnmurdoch, R. A. (1995). The effect of shortening incisor teeth on the eruption rates and lengths of the other incisors in the rat. Archives of Oral Biology, 40, 467–471. Cheney, P. D. (2002). Electrophysiological methods for mapping brain motor circuits. In A. W. Toga & J. C. Mazziotta (Eds.), Brain mapping: The methods. (2nd ed.). New York, NY: Academic Press. Donoghue, J. P., & Wise, S. P. (1982). The motor cortex of the rat—Cytoarchitecture and microstimulation mapping. The Journal of Comparative Neurology, 212, 76–88. Ebner, F. F. (2005). Neural plasticity in adult somatic sensorymotor systems. Florida: CRS Press. Farina, S., Valeriani, M., Rosso, T., Aglioti, S., Tamburin, S., Fiaschi, A., et al. (2001). Transient inhibition of the human motor cortex by capsaicin-induced pain. A study with transcranial magnetic stimulation. Neuroscience Letters, 314, 97–101. Farkas, T., Kis, Z., Toldi, J., & Wolff, J. R. (1999). Activation of the primary motor cortex by somatosensory stimulation in adult rats is mediated mainly by associational connections from the somatosensory cortex. Neuroscience, 90, 353–361. Franchi, G. (2001). Persistence of vibrissal motor representation following vibrissal pad deafferentation in adult rats. Experimental Brain Research, 137, 180–189. Frost, S. B., Barbay, S., Friel, K. M., Plautz, E. J., & Nudo, R. J. (2003). Reorganization of remote cortical regions after ischemic brain injury: A potential substrate for stroke recovery. Journal of Neurophysiology, 89, 3205–3214. Frostig, R. D. (2006). Functional organization and plasticity in the adult rat barrel cortex: Moving out-of-the-box. Current Opinion in Neurobiology, 16, 445–450.
Gioanni, Y., & Lamarche, M. (1985). A reappraisal of rat motor cortex organization by intracortical microstimulation. Brain Research, 344, 49–61. Graziano, M. S., & Aflalo, T. N. (2007). Mapping behavioral repertoire onto the cortex. Neuron, 56, 239–251. Guggenmos, D. J., Barbay, S., Bethel-Brown, C., Nudo, R. J., & Stanford, J. A. (2009). Effects of tongue force training on orolingual motor cortical representation. Behavioural Brain Research, 201, 229–232. Guic, E., Carrasco, X., Rodriguez, E., Robles, I., & Merzenich, M. M. (2008). Plasticity in primary somatosensory cortex resulting from environmentally enriched stimulation and sensory discrimination training. Biological Research, 41, 425–437. Haas, D. A., & Lennon, D. (1995). A 21 year retrospective study of reports of paresthesia following local anesthetic administration. Journal (Canadian Dental Association), 61, 319–320 323–326, 329–330. Halkjaer, L., Melsen, B., Mcmillan, A. S., & Svensson, P. (2006). Influence of sensory deprivation and perturbation of trigeminal afferent fibers on corticomotor control of human tongue musculature. Experimental Brain Research, 170, 199–205. Hallett, M. (2001). Plasticity of the human motor cortex and recovery from stroke. Switzerland: Warth. Hamdy, S., Rothwell, J. C., Aziz, Q., Singh, K. D., & Thompson, D. G. (1998). Long-term reorganization of human motor cortex driven by short-term sensory stimulation. Nature Neuroscience, 1, 64–68. Haraldson, T., & Zarb, G. (1988). A 10-year follow-up study of the masticatory system after treatment with osseointegrated implant bridges. Scandinavian Journal of Dental Research, 96, 243–252. Hatanaka, N., Tokuno, H., Nambu, A., Inoue, T., & Takada, M. (2005). Input–output organization of jaw movement-related areas in monkey frontal cortex. The Journal of Comparative Neurology, 492, 401–425. Henry, E. C., & Catania, K. C. (2006). Cortical, callosal, and thalamic connections from primary somatosensory cortex in the naked mole-rat (Heterocephalus glaber), with special emphasis on the connectivity of the incisor representation. The Anatomical Record. Part A, Discoveries in Molecular, Cellular, and Evolutionary Biology, 288, 626–645. Henry, E. C., Marasco, P. D., & Catania, K. C. (2005). Plasticity of the cortical dentition representation after tooth extraction in naked mole-rats. The Journal of Comparative Neurology, 485, 64–74. Hiraba, H., Yamaguchi, Y., & Iwamura, Y. (1997). Mastication-related neurons in the orofacial first somatosensory cortex of awake cats. Somatosensory & Motor Research, 14, 126–137. Hu, J. W., Woda, A., & Sessle, B. J. (1999). Effects of pre-emptive local anaesthesia on tooth pulp deafferentation-induced
148 neuroplastic changes in cat trigeminal brainstem neurones. Archives of Oral Biology, 44, 287–293. Huang, C. S., Sirisko, M. A., Hiraba, H., Murray, G. M., & Sessle, B. J. (1988). Organization of the primate face motor cortex as revealed by intracortical microstimulation and electrophysiological identification of afferent inputs and corticobulbar projections. Journal of Neurophysiology, 59, 796–818. Huang, C. S., Hiraba, H., Murray, G. M., & Sessle, B. J. (1989a). Topographical distribution and functional properties of cortically induced rhythmical jaw movements in the monkey (Macaca fascicularis). Journal of Neurophysiology, 61, 635–650. Huang, C. S., Hiraba, H., & Sessle, B. J. (1989b). Input–output relationships of the primary face motor cortex in the monkey (Macaca fascicularis). Journal of Neurophysiology, 61, 350–362. Huntley, G. W. (1997). Differential effects of abnormal tactile experience on shaping representation patterns in developing and adult motor cortex. The Journal of Neuroscience, 17, 9220–9232. Inoue, M., Harasawa, Y., Yamamura, K., Ariyasinghe, S., & Yamada, Y. (2004). Effects of food consistency on the pattern of extrinsic tongue muscle activities during mastication in freely moving rabbits. Neuroscience Letters, 368, 192–196. Iyengar, S., Qi, H. X., Jain, N., & Kaas, J. H. (2007). Cortical and thalamic connections of the representations of the teeth and tongue in somatosensory cortex of new world monkeys. The Journal of Comparative Neurology, 501, 95–120. Kaas, J. H., Qi, H. X., & Iyengar, S. (2006). Cortical network for representing the teeth and tongue in primates. The Anatomical Record. Part A Discoveries in Molecular, Cellular, and Evolutionary Biology, 288, 182–190. Kaas, J. H., Qi, H. X., Burish, M. J., Gharbawie, O. A., Onifer, S. M., & Massey, J. M. (2008). Cortical and subcortical plasticity in the brains of humans, primates, and rats after damage to sensory afferents in the dorsal columns of the spinal cord. Experimental Neurology, 209, 407–416. Katz, D. B., Simon, S. A., Moody, A., & Nicolelis, M. A. (1999). Simultaneous reorganization in thalamocortical ensembles evolves over several hours after perioral capsaicin injections. Journal of Neurophysiology, 82, 963–977. Keller, A., Weintraub, N. D., & Miyashita, E. (1996). Tactile experience determines the organization of movement representations in rat motor cortex. NeuroReport, 7, 2373–2378. Kis, Z., Rakos, G., Farkas, T., Horvath, S., & Toldi, J. (2004). Facial nerve injury induces facilitation of responses in both trigeminal and facial nuclei of rat. Neuroscience Letters, 358, 223–225. Kleim, J. A., & Jones, T. A. (2008). Principles of experiencedependent neural plasticity: Implications for rehabilitation
after brain damage. Journal of Speech, Language, and Hearing Research, 51, S225–S239. Kleim, J. A., Cooper, N. R., & Vandenberg, P. M. (2002). Exercise induces angiogenesis but does not alter movement representations within rat motor cortex. Brain Research, 934, 1–6. Kleim, J. A., Hogg, T. M., Vandenberg, P. M., Cooper, N. R., Bruneau, R., & Remple, M. (2004). Cortical synaptogenesis and motor map reorganization occur during late, but not early, phase of motor skill learning. The Journal of Neuroscience, 24, 628–633. Klineberg, I., & Jagger, R. (2004). Occlusion and Clinical Practice—An Evidence Based Approach. London: Elsevier. Krause, P., Forderreuther, S., & Straube, A. (2006). TMS motor cortical brain mapping in patients with complex regional pain syndrome type I. Clinical Neurophysiology, 117, 169–176. Kwan, C. L., Hu, J. W., & Sessle, B. J. (1993). Effects of toothpulp deafferentation on brain-stem neurons of the rat trigeminal subnucleus oralis. Somatosensory & Motor Research, 10, 115–131. Lee J -C., Avivi-Arber L., Adachi K., Yao D., & Sessle B. J. (2005). Motor cortex (MI) neuroplasticity associated with single or multiple trimmings of the rat incisors. Society for Neuroscience Abstract 174, 1. Linden, R. W., & Scott, B. J. (1989). The effect of tooth extraction on periodontal ligament mechanoreceptors represented in the mesencephalic nucleus of the cat. Archives of Oral Biology, 34, 937–941. Lotze, M., Grodd, W., Birbaumer, N., Erb, M., Huse, E., & Flor, H. (1999). Does use of a myoelectric prosthesis prevent cortical reorganization and phantom limb pain? Nature Neuroscience, 2, 501–502. Lund, J. P., & Kolta, A. (2006). Generation of the central masticatory pattern and its modification by sensory feedback. Dysphagia, 21, 167–174. Lund, J. P., Sasamoto, K., Murakami, T., & Olsson, K. A. (1984). Analysis of rhythmical jaw movements produced by electrical stimulation of motor-sensory cortex of rabbits. Journal of Neurophysiology, 52, 1014–1029. Martin, R. E. (2009). Neuroplasticity and swallowing. Dysphagia, 24, 218–229. Martin, R. E., Kemppainen, P., Masuda, Y., Yao, D., Murray, G. M., & Sessle, B. J. (1999). Features of cortically evoked swallowing in the awake primate (Macaca fascicularis). Journal of Neurophysiology, 82, 1529–1541. Martin, R. E., Macintosh, B. J., Smith, R. C., Barr, A. M., Stevens, T. K., Gati, J. S., et al. (2004). Cerebral areas processing swallowing and tongue movement are overlapping but distinct: A functional magnetic resonance imaging study. Journal of Neurophysiology, 92, 2428–2443. Merzenich, M. M., Nelson, R. J., Stryker, M. P., Cynader, M. S., Schoppmann, A., & Zook, J. M. (1984).
149 Somatosensory cortical map changes following digit amputation in adult monkeys. The Journal of Comparative Neurology, 224, 591–605. Miles, T. S., Nauntofte, B., & Svensson, P. (2004). Clinical oral physiology. Copenhagen: Quintessence. Miyashita, E., Keller, A., & Asanuma, H. (1994). Input–output organization of the rat vibrissal motor cortex. Experimental Brain Research, 99, 223–232. Murray, G. M., Lin, L. D., Yao, D., & Sessle, B. (2001). Sensory and motor functions of face primary somatosensory cortex in the primate. In M. Rowe & Y. Iwamura (Eds.), Somatosensory processing: From single neuron to brain imaging. Amsterdam, The Netherlands: Harwood Academic. Navarro, X., Vivo, M., & Valero-Cabre, A. (2007). Neural plasticity after peripheral nerve injury and regeneration. Progress in Neurobiology, 82, 163–201. Neafsey, E. J., Bold, E. L., Haas, G., Hurley-Gius, K. M., Quirk, G., Sievert, C. F., et al. (1986). The organization of the rat motor cortex: A microstimulation mapping study. Brain Research, 396, 77–96. Nordstrom, M. A. (2007). Insights into the bilateral cortical control of human masticatory muscles revealed by transcranial magnetic stimulation. Archives of Oral Biology, 52, 338–342. Nudo, R. J. (2003). Adaptive plasticity in motor cortex: Implications for rehabilitation after brain injury. Journal of Rehabilitation Medicine, Suppl. 41, 7–10. Nudo, R. J., Milliken, G. W., Jenkins, W. M., & Merzenich, M. M. (1996). Use-dependent alterations of movement representations in primary motor cortex of adult squirrel monkeys. The Journal of Neuroscience, 16, 785–807. Okayasu, I., Yamada, Y., Kohno, S., & Yoshida, N. (2003). New animal model for studying mastication in oral motor disorders. Journal of Dental Research, 82, 318–321. Paxinos, G. (2004). Rat nervous system. San Diego: Elsevier Academic Press. Proschel, P., & Hofmann, M. (1988). Frontal chewing patterns of the incisor point and their dependence on resistance of food and type of occlusion. The Journal of Prosthetic Dentistry, 59, 617–624. Remple, M. S., Bruneau, R. M., Vandenberg, P. M., Goertzen, C., & Kleim, J. A. (2001). Sensitivity of cortical movement representations to motor experience: Evidence that skill learning but not strength training induces cortical reorganization. Behavioural Brain Research, 123, 133–141. Remple, M. S., Henry, E. C., & Catania, K. C. (2003). Organization of somatosensory cortex in the laboratory rat (Rattus norvegicus): Evidence for two lateral areas joined at the representation of the teeth. The Journal of Comparative Neurology, 467, 105–118. Ridding, M. C., & Rothwell, J. C. (1997). Stimulus/ response curves as a method of measuring motor cortical excitability in man. Electroencephalography and Clinical
Neurophysiology/ Electromyography and Motor Control, 105(5), 340–344. Ridding, M. C., Brouwer, B., Miles, T. S., Pitcher, J. B., & Thompson, P. D. (2000). Changes in muscle responses to stimulation of the motor cortex induced by peripheral nerve stimulation in human subjects. Experimental Brain Research, 131, 135–143. Robbins, J., Butler, S. G., Daniels, S. K., Diez, G. R., Langmore, S., Lazarus, C. L., et al. (2008). Swallowing and dysphagia rehabilitation: Translating principles of neural plasticity into clinically oriented evidence. Journal of Speech, Language, and Hearing Research, 51, S276–S300. Rossignol, S. (2006). Plasticity of connections underlying locomotor recovery after central and/or peripheral lesions in the adult mammals. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 361, 1647–1671. Sakamoto, K., Nakata, H., & Kakigi, R. (2009). The effect of mastication on human cognitive processing: A study using event-related potentials. Clinical Neurophysiology, 120, 41–50. Samuels, M. A., & Feske, S. K. (2003). Office practice of neurology. New York: Churchill Livingston. Sanes, J. N., Donoghue, J. P., Thangaraj, V., Edelman, R. R., & Warach, S. (1995). Shared neural substrates controlling hand movements in human motor cortex. Science, 268, 1775–1777. Sapienza, S., Talbi, B., Jacquemin, J., & Albe-Fessard, D. (1981). Relationship between input and output of cells in motor and somatosensory cortices of the chronic awake rat. A study using glass micropipettes. Experimental Brain Research, 43, 47–56. Sessle, B. J. (1966). Attrition and eruption rates of the rat lower incisor. Journal of Dental Research, 45, 1571. Sessle, B. J. (2009). Orofacial motor control. In: L. Squire (Ed.), Encyclopedia of neuroscience. Vol. 7. Oxford: Academic Press. Sessle, B. J., Adachi, K., Avivi-Arber, L., Lee, J., Nishiura, H., Yao, D., et al. (2007). Neuroplasticity of face primary motor cortex control of orofacial movements. Archives of Oral Biology, 52, 334–337. Svensson, P., Romaniello, A., Wang, K., Arendt-Nielsen, L., & Sessle, B. J. (2006). One hour of tongue-task training is associated with plasticity in corticomotor control of the human tongue musculature. Experimental Brain Research, 173, 165–173. Swanson, L. W. (2004). Brain maps: Structure of the rat brain. Amsterdam: Elsevier Inc. Tandon, S., Kambi, N., & Jain, N. (2008). Overlapping representations of the neck and whiskers in the rat motor cortex revealed by mapping at different anaesthetic depths. The European Journal of Neuroscience, 27, 228–237.
150 Tandon, S., Kambi, N., Lazar, L., Mohammed, H., & Jain, N. (2009). Large-scale expansion of the face representation in somatosensory areas of the lateral sulcus after spinal cord injuries in monkeys. The Journal of Neuroscience, 29, 12009–12019. Teasell, R., Bayona, N., Salter, K., Hellings, C., & Bitensky, J. (2006). Progress in clinical neurosciences: Stroke recovery and rehabilitation. The Canadian Journal of Neurological Sciences, 33, 357–364. Toldi, J. (2008). Representational plasticity in the mammalian brain cortex. (Review article). Acta Physiologica Hungarica, 95, 229–245. Toldi, J., Laskawi, R., Landgrebe, M., & Wolff, J. R. (1996). Biphasic reorganization of somatotopy in the primary motor cortex follows facial nerve lesions in adult rats. Neuroscience Letters, 203, 179–182. Vartiainen, N., Kirveskari, E., Kallio-Laine, K., Kalso, E., & Forss, N. (2009). Cortical reorganization in primary somatosensory cortex in patients with unilateral chronic pain. The Journal of Pain, 10, 854–859. Veronesi, C., Maggiolini, E., & Franchi, G. (2006). Postnatal development of vibrissae motor output following neonatal
infraorbital nerve manipulation. Experimental Neurology, 200, 332–342. Yan, C., Ye, L., Zhen, J., Ke, L., & Gang, L. (2008). Neuroplasticity of edentulous patients with implantsupported full dentures. European Journal of Oral Sciences, 116, 387–393. Yildiz, N., Yildiz, S., Ertekin, C., Aydogdu, I., & Uludag, B. (2004). Changes in the perioral muscle responses to cortical TMS induced by decrease of sensory input and electrical stimulation to lower facial region. Clinical Neurophysiology, 115, 2343–2349. Yoshida, A., Taki, I., Chang, Z., Iida, C., Haque, T., Tomita, A., et al. (2009). Corticofugal projections to trigeminal motoneurons innervating antagonistic jaw muscles in rats as demonstrated by anterograde and retrograde tract tracing. The Journal of Comparative Neurology, 514, 368–386. Zhang, Z. W., & Deschenes, M. (1998). Projections to layer VI of the posteromedial barrel field in the rat: A reappraisal of the role of corticothalamic pathways. Cerebral Cortex, 8, 428–436. Zhang, G. X., & Sasamoto, K. (1990). Projections of two separate cortical areas for rhythmical jaw movements in the rat. Brain Research Bulletin, 24, 221–230.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 10
A hierarchical perspective on rhythm generation for locomotor control Sergiy Yakovenko* Département de Physiologie, Université de Montréal, Pavillon Paul-G. Desmarais, Montreal, Quebec, Canada
Abstract: The control of locomotion is a complex dynamic task solved with apparent ease by our body. How this is accomplished still remains an intriguing mystery. This chapter first describes classical and recent findings relevant to understanding the complexity of the question on the verge of several fields of neurophysiology, biomechanics, and computational neuroscience. Then, control of locomotion is analyzed with numerical simulations to reveal some basic characteristics responsible for modulation of the locomotor rhythm and high-level control of steering in the whole animal. In this study, the concept of a central pattern generator (CPG) for controlling locomotor rhythm first proposed by Brown was implemented in a “simple” model with bilateral half-center oscillators consisting of reciprocally organized integrators. The parameters of the CPG were determined by the process of optimization of its phase-duration characteristic that satisfies biomechanical requirements of the overground locomotion. The general finding of this study is that the modality of the control signal that drives CPGs for each limb corresponds to the desired speed of forward progression. This supports the idea that the descending and sensory feedback inputs to the spinal CPG are combined to produce a highlevel control signal that sets forward velocity. The same mechanism may be responsible for the control of steering by generating a differential input of speed commands to different limbs. Keywords: motor control; locomotion; rhythm and pattern generator; CPG; model.
Historical perspective and overview
prey or escaping from predators. It is likely that relatively crude neural mechanisms of early animals were already sufficient to accommodate the complexity of the dynamic control required for this behavior. Any additional neural structures involved in the control of movement have been acquired through evolutionary selection to add precision, steering, posture, and to
Any form of locomotion, the ability to move within environment, offers evolutionary rewards, for example, in the form of pursuing and catching *Corresponding author. Tel.: þ1-514-343-6111x3333; Fax: þ1-514-343-6113 DOI: 10.1016/B978-0-444-53825-3.00015-2
151
152
enable different modes of locomotion and their control. Even reaching, a focal nonstereotypic movement, was proposed to have evolved from locomotion (Georgopoulos and Grillner, 1989). In vertebrates, the mechanisms responsible for overground locomotion are expressed in a complex hierarchical system that contains multiple levels of control (Fig. 1). The evolution of this hierarchy has occurred through sequential modifications of both mechanical and neural structures. The phylogenetic conservation of these structures and control principles is evident in the comparative examination of the organization of locomotor control across animals at different levels on the evolutionary ladder (Prochazka, 1993). However, the functional contribution and integration of additional mechanisms responsible for execution of novel
Supraspinal
Neural system
CPG
Rhythm & pattern generator
Mechanical system
Muscle
Sensory feedback
Motoneurons
Environment
Fig. 1. The diagram of hypothesized levels of locomotor control in vertebrates. Multiple descending pathways converge on the CPG, which is coupled to the mechanical system by sensory feedback.
behaviors are progressively harder to describe due to the growing complexity and associated emergent properties of the system, for example, self-stabilizing properties of multiple negative feedback loops of muscle length control (Blickhan et al., 2007; Prochazka, 1999; Yakovenko et al., 2004) and neuromechanical tuning of locomotor phase control in vertebrates (Prochazka and Yakovenko, 2007a; Taga, 1995) and insects (Ausborn et al., 2007). This chapter describes the essential demands of locomotion and the simplest mechanisms that are sufficient to satisfy these demands. A large set of animal behaviors is produced by repetitive stereotypic actions, for example, chewing, breathing, and different forms of locomotion in water, air, and/or over ground. The ability of animals to generate rhythmic coordinated movements shortly after spinalization has been known for centuries (de La Mettrie, 1745; Chapter 16). The mechanism responsible for this behavior was initially attributed to the interplay between consecutive antagonistic actions generating locomotor phases and the proprioceptive reflexes they evoke (Freusberg, 1874; Sherrington, 1910). The alternative explanation attributed sensory signals only regulatory and not the essential role for locomotion based on the observations of stepping in spinal deafferented animals (Brown, 1911). In this scheme, the stepping can be generated by “the intrinsic factor” alone, without either sensory feedback or descending inputs largely responsible for adapting the individual movement components to the demands of the environment. For over half of the century, the existence of this mechanism, now termed a central pattern generator (CPG), could not overcome a favored viewpoint of the reflexive stepping (Stuart and Hultborn, 2008). In part, this was due to the popularity of the reductionist methods applied to the study of spinal reflexes (Prochazka et al., 2002). The revival of the intrinsic spinal rhythmogenesis concept in 1960s coincided with the renewed holistic interest in the complexity of locomotion and other rhythmic movements across different animals.
153
Today, there is a growing consensus that both intrinsic and sensory feedback signals play a crucial role in controlling the act of locomotion (Prochazka and Yakovenko, 2007b; Rossignol et al., 2006). However, our understanding of the locomotor control is not complete without the theoretical description of the rhythmogenic mechanism of the CPG and its regulation by sensory and descending pathways that converge on it. The challenge is to describe the hierarchical system with emergent properties arising from pluralistic interactions between multiple integrated elements of this system. Since such properties are not accounted by function of individual components, which can be readily studied with typical reductionist techniques, the contribution of holistic computational approaches is essential. The goal of this computational study is to probe how the locomotor phase timing is regulated by the dynamic rhythmogenic spinal mechanism. Stabilizing properties of the locomotor system Stable limbed locomotion can be produced without muscle contractions, provided appropriate initial conditions and sufficient potential or kinetic energy. Locomotion is an oscillatory behavior resulting from interactions of the animal with its environment. It has been noted over 50 years ago that simple mechanical models, even those consisting of a mass bouncing on weightless limbs, can describe general features of walking and running (Alexander, 1976; Saunders et al., 1953). These models were based on observations that periodic interactions of limbs with the ground produce arc trajectories of the hip and, consequently, the center of mass during overground walking of animals, quadrupeds and bipeds. The dynamics in this system can be described by oscillations of potential and kinetic energy, and it is closely related to that of inverted pendulum (Cavagna et al., 1976, 1977). Consider the following sequence of events that occurs during locomotion of an inverted pendulum model. At ground
contact or onset of stance phase of locomotion, some of the energy in the system is lost due to the interaction with the ground. However, most of it is converted from kinetic to potential energy and then recovered when the center of mass starts falling after mid-stance. To ensure stable gait in the anteroposterior body plane, the amount of energy has to be restored and maintained, for example, by using the potential energy of walking down slope. This sequence of energy oscillations repeats when the other leg collides with the ground. When running, the center of mass of an animal is bouncing with a minimum during mid-stance. As the center of mass moves down after the ground contact, muscles and ligaments of extensors stretch and store the contact energy as the elastic (and potential) energy, which is then recovered during the following propulsive phase of stance. The mechanism that can describe this behavior is analogous to a mass-spring system, a familiar pendulum with a spring element between the point of ground contact and the center of mass. It is remarkable that the dynamics of this mechanism can not only describe the overall properties of interactions with the environment and energy conservation during walking, running, hopping, and trotting (Cavagna et al., 1977; McMahon and Cheng, 1990) but also predict transitions between these different modes of gait (Srinivasan and Ruina, 2006). The robustness of this simple mechanism has led to the suggestion that the mass-spring inverted pendulum is a template of locomotion, a simplest model that exhibits a desired behavior and can be used for simplifying the control of walking robots or animals (Full and Koditschek, 1999; Full et al., 2002). While the studies of simple mechanical models have provided a conceptual framework for the description of locomotion, their predictions of locomotor frequency (Bertram and Ruina, 2001) and energy expenditure in step-to-step transitions for overground walking do not predict accurately the experimental results (Kuo et al., 2005).
154
Some of the discrepancies in the cost of locomotion are due to the higher complexity of physical interactions between body appendages. For example, the inertial contribution of multiple leg segments and the metabolic energy that is used during stance for body support are excluded from calculations. Other errors could originate from inadequate model assumptions. For example, the models that rely on optimal efficiency of gait may omit various exogenous conditions that modulate behavior like uneven terrain with the possibility of a ground slip or different task context (consider a skipping gait). Yet other inconsistencies could be due to inadequate description of the neuromuscular organization, for example, arrangement of mechanical muscle actions, contribution from autogenic and heterogenic sensory feedback pathways. For example, the large yaw displacements of the ankle joint can be counteracted with large extrasaggital levering of extensor muscles, which increases torques that stabilize quadrupeds and bipeds during locomotion (Lawrence and Nichols, 1999; Lawrence et al., 1993). Thus, to describe the details of the motor control mechanisms, simple models of passive walkers need to be augmented with models of muscles and sensory feedback. In essence, muscles are tunable springs (Hogan, 1985) with stabilizing intrinsic properties that manifest themselves as immediate negative length feedback and velocity damping. However, the muscle modeling is a treacherous field where the level of necessary details is not readily evident. For example, the most commonly used description of the length–force relationship is based on the sarcomere muscle model with a characteristic isometric “bell-shaped” curve, which predicts a nonelastic decrease in muscle force as the muscle is stretched beyond its optimal length (Zajac, 1989). The molecular mechanism responsible for the descending limb may be explained to a large degree by the sliding-filament theory (Huxley, 1974). The force production decreases with the decrease in overlap of actin and myosin filaments as sarcomeres lengthen (Gordon et al., 1966). However, the models that use the classical bell-shaped
profile alone to describe the length–force relationship and do not take into account other nonlinear history-dependent muscle properties, for example, the short-range stiffness (Joyce et al., 1969), are worse than a simple linear model at describing experimentally derived dynamics of muscle behavior at longer lengths (Gillard et al., 2000). For that reason, choosing a good approximation of muscle dynamics depends heavily on the studied details of the behavior. The short-latency sensory feedback mechanisms add another stabilizing closed-loop feedback on top of the intrinsic zero-delay feedback of muscles. In part, the complexity of this organization that arises from interactions within the closed-loop system is responsible for the long standing debate about the relative contribution of the stretch reflexes during locomotion with the estimates ranging from 20% to 60% of the overall muscle activity (Bennett et al., 1996; Pearson and Misiaszek, 2000; Sinkjaer et al., 2000; Stein et al., 2000; Yang et al., 1991). If hips of a decerebrate cat walking on a treadmill are fixed to provide lateral stability at the expense of preventing the normal bouncing motion of the center of mass then the extensor muscles may be forced to stretch more than during the unconstrained gait for the same level of activity. This experimental condition recreated in a biomechanical model shows that average stretch reflex contribution of 30% could be overestimated to contribute about 80% of force and more than 50% of EMG of ankle extensors (Yakovenko et al., 2004). This modeling result is in agreement with experiments where the gain of the closed-loop sensory feedback included the contribution of the muscle force–length characteristic (Bennett et al., 1996; Donelan et al., 2009). In general, short-latency autogenic sensory pathways (arising from a muscle's own spindles) further enhance the stabilizing correction of the intrinsic negative feedback of muscle length (Burkholder and Nicols, 2000; Wilmink and Nichols, 2003; Yakovenko et al., 2004), when the heterogenic sensory pathways (arising from other muscles) compensate for inertial nonuniformities of the limb
155
and may contribute to inter-joint coordination (Wilmink and Nichols, 2003). Organization and function of the CPG The neural mechanism responsible for generating both rhythm and pattern of locomotion is thought to be composed of the tripartite control system consisting of multiple supraspinal and sensory feedback pathways interacting with spinal oscillatory network. While the importance of sensory feedback for pattern generation was recognized early (Philippson, 1905; Sherrington, 1910), Thomas Brown discovered “the intrinsic factor” or the CPG and outlined its basic principles (Brown, 1911). It is remarkable that the first concept of the CPG described it as an oscillatory mechanism consisting of pairs of dynamic reciprocally coupled units under regulatory influence of descending pathways and sensory feedback, the view that did not change dramatically for 100 years. Although most of the advanced analytical and computational tools required for the theoretical description of such a complex bistable system, that is, with stable reciprocal states, were not available at the time, the first attempts to simulate the dynamics of the half-center oscillator proposed by Thomas Brown were done just a decade later. Fritz Vérzar, a Hungarian physiologist, pioneered dynamic simulations in neurobiology using mechanical models or robots (Verzar, 1923), and this technique is now part of an active research field at the junction of robotics, artificial intelligence, and neuroscience (Datteri and Tamburrini, 2007). The CPG in Vérzar's model was made of two water-operated vacuum pumps connected with tubes and instrumented with pressure gages that tracked the circulation of liquid with fine pebbles to simulate activity-dependent fatigue. These pebbles could terminate the leak through one of the half-centers and force the “activity” of the contralateral side. The main findings of this pioneering study supported Brown's half-center oscillator hypothesis by
showing that the rhythm generation could be produced by two antagonistic integrators and, in addition, the phase switching can be regulated by an intrinsic fatigue-related mechanism. Presently, our understanding of the CPG can be greatly expanded by applying advanced analytical and computational tools to derive a theoretical description of the mechanisms of such a complex bistable system (Daun et al., 2009). Yet, despite the significant research effort, the anatomical details of the localization and connectivity within interneuronal network constituting mammalian CPG still remain unknown (see section on “Historical Perspective and Overview”), whereas this level of details is available in some invertebrates (for review, see Grillner and Wallén, 1985; McCrea and Rybak, 2008). Much of what we do know about the motor control strategies for vertebrate locomotion comes from studying motor activity during artificial interventions in reduced preparations. It has been shown that the vertebrate CPG is distributed across several spinal segments (Arshavsky et al., 1997; Deliagina et al., 1983; Kiehn and Kjaerulff, 1998; Kremer and Lev-Tov, 1997). The excitability of the spinal segments that contain the CPG circuitry appears to be biased toward rostral segments, which may “lead” rhythmogenesis in caudal segments (Arshavsky et al., 1997; Deliagina et al., 1983; Kiehn and Kjaerulff, 1998; Marcoux and Rossignol, 2000). Multiple inputs to spinal CPG accommodate its locomotor output to different mechanical and task demands. The locomotor activity of the spinal cord can be evoked and modulated by stimulation of either descending (Bretzner and Drew, 2005; Orlovsky, 1972; Perreault et al., 1994; Rho et al., 1999; Shik and Orlovsky, 1976; Shik et al., 1966), peripheral sensory (Andersson and Grillner, 1983; Duysens and Pearson, 1980), or propriospinal (Ribotta et al., 2000; Yakovenko et al., 2007) pathways of the spinal cord. By definition, the CPG can be intrinsically rhythmogenic even in the absence of other inputs. The pattern of bursting with many features of regular locomotion emerges with
156
cutaneous stimulation or spontaneously in motor nerves of decerebrate preparation (Côté and Gossard, 2003; Pearson and Rossignol, 1991), or it can be evoked by increasing the excitability of the appropriate spinal segments using pharmacological agents such as L-dihydroxy-phenylalanine (L-DOPA) or N-methyl-D-aspartate (NMDA) (Grillner, 1969). Several molecular mechanisms can be responsible for the appropriate bursting regime of the CPG (Daun et al., 2009), even in a network with highly variable sets of parameters (Goaillard et al., 2009; Grashow et al., 2009). For example, Grashow et al. (2009) have used the dynamic clamp to simulate different combinations of synaptic and intrinsic parameters in isolated circuits of crab gastric mill neurons. They found that while individual circuits could produce anomalous responses to two neuromodulators, serotonin and oxotremorine (muscarinic agonist), the overall mean population response was statistically reproducible across all parameter regimes. This supports the counterintuitive and reassuring idea that biological neural networks consisting of oscillatory circuits with complex dynamics, diverse structure, and intrinsic properties can produce “simple” stable network behavior. In other words, the complex dynamic structure of the CPG may not preclude our understanding of its main functions. The neural circuits that form the CPG have evolved to control dynamics of locomotion and to simplify its hierarchical control. Even though the CPG can be endogenously active, sensory feedback signals are the primary control inputs to the rhythm-generating circuitry, which might not have been an accident but inevitability of evolution (Brown, 1914). It turns out that a combination of feedforward and feedback signals is a general optimal solution for the control of pendular dynamics, which describes movement of “passive walkers” (Kuo, 2002). Even though this view might appear counterintuitive, the control system under influence of motion-related sensory feedback can be thought of being itself controlled by the oscillatory dynamics of the musculoskeletal
system (Taga et al., 1991). Along the same lines of reasoning, the mechanism of CPG is suggested to be an internal model of the controlled mechanical oscillations during locomotion (Brooke and Zehr, 2006; Full and Koditschek, 1999; Kuo, 2002; Pearson et al., 1999; Taga, 1998), that is, the CPG contains the input–output relationship of the motor behavior it controls. Such internal models embedded at the lowest level of a hierarchical sensorimotor system would reduce the complexity of the high-level control signals by offloading some control tasks to local autonomic controllers (Pfeifer et al., 2007). This process of embedding complex dynamic transformations into the dedicated neural networks have been also proposed for acquisition of novel motor skills through learning and evolution (Fry et al., 2009; Jordan, 1996; Kawato, 1999; Mussa-Ivaldi, 1999; Prentice et al., 2001). The basic question is then what is the control signal that drives the spinal CPGs? One approach is to use the Hodgkin–Huxley formalism or its variants to describe neural complexity leading to the generation of locomotor patterns (Bashor et al., 1998; Rybak et al., 2006a,b). This largescale modeling is a promising development in understanding function at multiple scales, that is, from single channels and neurons to networks and the whole transformation in the sensorimotor pathways. The main pitfall of this approach is the accumulating errors that creep in the analysis with every choice of a parameter that cannot be empirically measured. Such complex systems with many more degrees of freedom or parameters in the controller than in the task it controls are also difficult to constrain to physiological solutions, because they can have multiple different ways to solve the same relatively “simple” behavior. Admittedly, the subset of all these possible solutions could be physiological. The alternative is to simplify the model to known characteristics of large subpopulations of neurons. For example, Brown (1911) has identified two features of the control mechanism based on the observed behavior: (i) the endogenously active networks contain
157
reciprocal inhibitory connections between spinal networks forming antagonistic groups of half-center oscillators and (ii) these networks integrate multiple feedforward and feedback control signals to adjust the pattern of locomotion. A simple mathematical model that implements these two features with integrate-to-threshold units organized with mutually inhibitory connections was sufficient to describe complex properties of behavior, such as different regimes of the extensor- and flexor-dominant locomotion observed during electrical stimulation of the mesencephalic locomotor region (MLR), MLR-induced fictive locomotion (Prochazka and Yakovenko, 2007a; Yakovenko et al., 2005). Simple rhythm generation model
proposed and tested by Yakovenko and colleagues (Prochazka and Yakovenko, 2007a; Yakovenko et al., 2005) that generates bilateral rhythm based on the interactions within and between the half-center elements. The testing task for this model is to adjust appropriately locomotor phases required for overground locomotion at different speeds (Goslow et al., 1973; Halbertsma, 1983). The simple model of the process controlling phase durations is based on the ability of the network to integrate inputs until a critical threshold is reached, which causes phase resetting. Then, the model can be expressed as the system of differential equations consisting of two parts (1) the transformation of the input signals and (2) the intrinsic interactions within the system. BL x_ ¼ x0 þ Gu u þ GUL x x þ Gx ð1 xÞjx>0
The locomotor rhythm-generating model is based on the classical principle of intrinsic rhythmogenesis of spinal networks described in Brown's half-center oscillator hypothesis (Brown, 1911). The implementation shown in Fig. 2 is the extension of the single oscillator model for the description of phase dominance previously
where Gu matrix represents gains of input signals u, Gx matrices represent the strength of unilateral and bilateral connections between elements of the pattern generator with the internal states x that have the constant bias x0. The internal states are limited to positive values with the threshold of switching set to 1. The unilateral and bilateral Gx matrices have the following form: GUL x ¼ Irleak
CONTRA
IPSI I
III
2
r13
GBL x
r14
II
r23
IV
r24
Fig. 2. The schematic of Brown's locomotor pattern generator. The bilateral model consists of four half-center integrators with intrinsic connections that provide reciprocal excitation and inhibition of the ipsilateral and contralateral neighboring integrators (excitation and inhibition are shown as a line and a filled circle, respectively).
ð1Þ
0 6 0 ¼6 4 r24 r14
0 0 r23 r13
r13 r23 0 0
ð2Þ 3 r14 r24 7 7 0 5 0
ð3Þ
where I is the identity matrix, rleak is the constant that determines intrinsic state-dependent feedback, rij is the coupling term that represents the effect between i and j elements in the model as shown in Fig. 2. The phase adjustments of flexor and extensor phases that are characteristic of the level overground locomotion were simulated by solving Eq. (1) with Runge–Kutta (fourth order) method. To determine the appropriate parameters of the
158
CPG model, the space of parameters was searched using Nelder–Mead multidimensional unconstrained nonlinear error minimization (state tolerance 5.0 10 3 and function tolerance 5.0 10 3) algorithm. The minimized cost function was expressed as the sum of two errors: (1) the phase-duration characteristic error in the relationship between experimental and the simulated phase and cycle durations and (2) the error in the range of simulated cycle periods (see Fig. 6 in Halbertsma, 1983). In these simulations, out-ofphase initial conditions were set for the reciprocal elements of the half-center integrators. With the iteratively increasing input signal, the algorithm converged on the solution of parameters x0, Gx, and Gu of Eq. (1), shown in Table 1. The CPG model driven by a ramping input signal (Fig. 3a) produces oscillations of variable phase duration arising from interactions between bilateral half-center integrators. The integration to threshold process is reciprocal for each halfcenter and out-of-phase with the contralateral elements shown in Fig. 3b. Shown in Fig. 4, the phase-duration characteristic of phases and the corresponding cycles of the bilateral rhythm generator with the optimized parameters from Table 1 accurately reproduce experimental results (Rswing2 and Rstance2 are 0.915 and 0.999, respectively). Note that the simulated phase-duration characteristics have a slight nonlinear deviation, which is also observed in the original experimental data of Halbertsma (1983). Note that the phase-duration characteristic for both half-centers is symmetrical for symmetrical input signal. One of the central questions in motor control is to identify the modality of the CPG input signals from multiple descending and sensory feedback
pathways. These input signals could represent position, velocity, force, effort or comprise of low- and high-order variables appropriate for multiple task-specific goals of locomotion. For example, the modality of the primary afferents of muscle spindles is mostly velocity of muscle stretch, although muscle length and its activity related to acceleration are also represented (Prochazka, 1999). However, the representation of the low-level control parameters at high levels of the neural hierarchy would indicate that the complexity of the musculoskeletal organization and its interactions with the environment are not resolved by the low-levels in the hierarchy. This statement is unlikely in absolute terms; therefore, the modality of inputs converging on the rhythmgenerating networks has global context, for example, the speed of forward progression (Prochazka & Sorensen, 2009; Markin et al., 2010). To test this hypothesis, the output of the CPG model can be solved as a function of its inputs. First, the relationship between the CPG input in the model and the simulated locomotor phases can be derived (Fig. 4a). This relationship can be further transformed to test the association between the CPG input and the velocity based on the previously described power function between cycle duration and velocity, Tc ¼ 0.5445 V 0.5925 (Goslow et al., 1973). Then, the relationship between the CPG input and the velocity of gait can be calculated from the values of the cycle duration in Fig. 4a. It turns out that the result shown in Fig. 4b is a highly linear relationship between the CPG input and the forward velocity (R2 ¼ 0.9978). Thus, the CPG input expressed as the desired speed of movement can automatically adjust locomotor phases so that they are
Table 1. The CPG model parameters of optimization x01 0.0007
x02 2.4256
g1 0.6203
g2 0.4882
rleak 0.0094
r13 0.1339
r14 0.0485
r23 0.0823
r24 0.0981
The parameters of Eqs. (1)–(3), where the vector of intrinsic bias, which determines the phase dominance, is x0 ¼ [x01, x02, x01, x02] and the input gain vector is Gu ¼ [g1, g2, g1, g2].
159 (a)
4.5
input command (nu)
4 3.5 3 2.5 2 1.5 1 0.5 0
(b)
1
L
0
0.5
1
2
3 4 5 time (s)
6
7
8
0
R
1 0.5 0
2 tim
phase duration (s)
e
(c)
0 1 2 3 4 5 6 7 8 9 10 time (sec)
stance 1 swing
0
0
1 cycle duration (s)
2
Fig. 3. The bilateral pattern of locomotor phase modulation with optimized CPG model parameters. (a) The ramping input driving the bilateral half-center oscillators shown for the beginning of 80 s simulation. (b) The corresponding integration processes within the model shown for each oscillator, where black and white lines represent the dominant and nondominant half-centers, respectively. (c) The relationship between phase and cycle duration for each integrator is plotted together with the desired Halbertsma (1983) best-fit linear model.
appropriate for that speed. This finding supports the idea that the control signals that drive the rhythm-generating network are expressed in the modality of locomotor speed. This modeling result is supported by classical observations of stimulating MLR in cats (Shik et al., 1969). Shik and colleagues (1969) have reported that a ramping increase of the stimulation strength produces continuous transition from slow walking to trot and gallop with appropriate modulation of locomotor phases. Another computational study of Collins and Richmond (1994) has shown that all these locomotor behaviors can be generated by a single model of coupled oscillators with hard-wired parameters across all behaviors. Thus, the same neural mechanism could integrate multiple inputs to induce nonlinear transformation resulting in the change of velocity. Furthermore, the speed of locomotion can be increased by mechanical stimulation of primary afferents in humans (Ivanenko et al., 2000) and by increasing the gain of just the homonymous stretch reflex feedback in neuromechanical models (Prochazka et al., 2002). This points to the built-in mechanical intelligence of the neuromuscular organization. Moreover, the visuomotor transformation of optic flow also has a modulatory effect on the velocity of locomotion in animals (Davis and Ayers, 1972; Fry et al., 2009), including humans (Prokop et al., 1997). Fry et al. (2009) have found that the flight speed control in fruit flies depends mostly on the retinal slip speed and not on the other components of the multimodal control signal, which are commonly used in engineering applications. Thus, the visuomotor transformation of optic flow speed into the motor command signal may directly specify locomotor speed and acts as an online feedback to control locomotor velocity (see Chapter 6). The hypothesis of the velocity control in the CPG model is further tested for asymmetric gait control. Halbertsma (1983) has studied the splitbelt locomotion of cats, where speed of left and right sides of the treadmill could be separately controlled. Figure 5a shows that when the speeds
160 (a)
(b)
2
1
y = 0.2357x – 0.1272
velocity (m/s)
time (s)
R2 = 0.9978
cy
cle
1 sta
nc
e
0.58
0.5
0.16
0
swing 0
1
2
3
4
5
CPG input
0
0
1
2 3 CPG input
4
5
Fig. 4. Modality of the CPG command signal. (a) The power function relationships between the input to the CPG model and the locomotor phases. (b) The relationship between the input signal and the forward velocity corresponding to the simulated cycle durations in A is a linear function (R2 ¼ 0.998).
were set to 0.16 and 0.57 m/s, the animal produced double-stepping on the fast side. Here, the position of two contralateral hind toes is shown relative to a stationary camera during forward gait on a split-belt treadmill. The brief upward phases indicate the change of toe position during swing. To simulate this result in the model, the input strength to the half-center oscillator of each limb was found using the linear relationship between the CPG input and the speed of locomotion (Fig. 4). Not surprisingly, the simulated relationship between limbs was exactly out-of-phase when the same desired speed of 0.57 m/s was used as input for both limbs (Fig. 5b, simulation start). However, when one of the limbs (gray) was set to walk with the desired speed of about 3.6 times slower than the initial speed, the duration of the simulated extensor phase was prolonged (Fig. 5b, solid gray at simulation end) and closely matched the duration of the experimental stance phase in Fig. 5a. Furthermore, the phase adjustments followed the experimentally observed phase modulation with different locomotion velocities (Fig. 3c). This result supports the idea that the velocity control signal of the CPG may be composed from separate speed signals for each limb. When walking on the ground and not the split-belt treadmill, this simple mechanism would provide smooth
steering control of the whole animal. The analogy of this mechanism is the control of turning in a tracked vehicle rather than steering of a car. The idea of steering control by the differential of speed commands to the CPG of each limb is supported by the observations of turning in lampreys by muscle contraction on one side of the body in response to the increased activity of the ipsilateral reticulospinal neurons (Deliagina et al., 2000), or in response to asymmetrical electrical stimulation of the cutaneous afferents (Fagerstedt and Ullén, 2001) or MLR (Sirota et al., 2000). The steering with the asymmetric command to the CPGs has also been shown in simulations of cat locomotion (Kozlov et al., 2009). In human walking along a curved path, the adjustments of stance duration of the inner and outer legs on the path are also consistent with the idea that the speed of the outer leg is increased relative to the inner leg (Courtine and Schieppati, 2003). Similarly, the steering can be produced either by asymmetric vibration during gait or asymmetric torsional loading of human subjects prior to gait initiation (Ivanenko et al., 2000, 2006). The results of studies of human split-belt locomotion show that phases of “fast” and “slow” limbs are appropriately adjusted according to the phase-duration characteristic (Dietz et al., 1994; Reisman et al., 2005), similar
161 (a) hind toe
0.16 vs 0.58 m/s
0 (b)
1
2 time (s)
0.58 vs 0.58 m/s
3
4
0.16 vs 0.58 m/s
1 0.8 0.6 0.4 0.2 0
0.5
0
(c)
1
phase duration (s)
2
1.5
L
1.5 1 0.58 m/s 0.5 0
0
0.5 1 1.5 cycle duration (s)
2 time (s) (d) phase duration (s)
half-center state (nu)
L R
2
2.5
3
3.5
4
R
2 /s
1.5
16
m
0.
1 /s
8 .5
0.5 0
m
0
0
0.5 1 1.5 cycle duration (s)
2
Fig. 5. Simulation of double-stepping using velocity as the input signal in the model. (a) The positions of left and right hind toes (black and gray) are shown during locomotion on a split-belt treadmill set to speeds 0.16 and 0.58 m/s (modified from Fig. 21 in Halbertsma, 1983). (b) The states in the model simulated using velocity as input signals. The states of “left” and “right” halfcenters are shown in black and gray in time. The termination of the shaded period (2 s) indicates the time when the input velocity signal based on the relationship in Fig. 4b was switched from symmetric (0.58 vs. 0.58 m/s) to asymmetric (0.58 vs. 0.16 m/s) pattern. (c , d) The phase and cycle duration relationships are shown for the left and right sides of the CPG model. Note that the change in the right half-center oscillator follows the expected modulation of the phases for the input speeds.
to the result in Fig. 5b. Moreover, studies of motor adaptation in people with motor dysfunctions during locomotion on the split-belt treadmill provide further support for the idea that separate signals for each limb from multiple descending
pathways are involved in locomotor phase adjustments in humans. In people with cerebral and cerebellar damage, the adaptation to the asymmetric gait imposes similar characteristic stance duration asymmetry observed in normal
162
subjects and predicted by the model (Morton and Bastian, 2006; Reisman et al., 2007, 2009). This plasticity may involve different contribution of the spinal networks and spared descending pathways to the resulting coordination, which suggests that the redundancy of the locomotor phase control can be used for rehabilitation of locomotion. Summary Despite the high complexity of dynamic interactions between body and its environment, the CPG mechanism simplifies the task of the high-level control signals necessary for the regulation of phases during different modes of overground locomotion. In this study, a simple oscillator model based on reciprocal integration proposed by Thomas Graham Brown was sufficient to describe with high precision the necessary locomotor phase adjustments. Moreover, the model predicts that the input signals that converge on the CPG may be largely determined by the high-level velocity control strategy, where spinal rhythm generators for each limb are driven by the desired speed signals. This mechanism can also describe steering as the differential between the control signal to each limb that determines speeds. This simplification of control paradigm may have evolved from the ability of biological systems to embed complex nonlinear behaviors within the hierarchical control structure. In conclusion, the proposed model of the locomotor phase coordination between multiple descending and sensory feedback pathways realistically reproduces and mechanistically explains experimentally observed intra- and interlimb coordination. Acknowledgments This work was supported in part by Wissenschaftskolleg zu Berlin (WIKO) and the Canadian Institutes of Health Research (CIHR). The author wishes to thank Arthur Prochazka
for inspiring discussions that led to this work and Valeriya Gritsenko for helpful comments on this chapter. Abbreviations CPG MLR EMG
central pattern generator mesencephalic locomotor region electromyography
References Alexander, R. (1976). Mechanics of bipedal locomotion. In P. S. Davies (Ed.), Perspectives in experimental biology (pp. 493–504). Oxford, UK: Pergamon Press. Andersson, O., & Grillner, S. (1983). Peripheral control of the cat's step cycle. II. Entrainment of the central pattern generators for locomotion by sinusoidal hip movements during “fictive locomotion” Acta Physiologica Scandinavica, 118(3), 229–239. Arshavsky, Y. I., Deliagina, T. G., & Orlovsky, G. N. (1997). Pattern generation. Current Opinion in Neurology, 7(6), 781–789. Ausborn, J., Stein, W., & Wolf, H. (2007). Frequency control of motor patterning by negative sensory feedback. The Journal of Neuroscience, 27(35), 9319–9328. Bashor, D. P., Dai, Y., Kriellaars, D. J., & Jordan, L. M. (1998). Pattern generators for muscles crossing more than one joint. Annals of the New York Academy of Sciences, 860, 444–447. Bennett, D. J., De Serres, S. J., & Stein, R. B. (1996). Gain of the triceps surae stretch reflex in decerebrate and spinal cats during postural and locomotor activities. Journal de Physiologie, 496(3), 837–850. Bertram, J. E. A., & Ruina, A. (2001). Multiple walking speedfrequency relations are predicted by constrained optimization. Journal of Theoretical Biology, 209(4), 445–453. Blickhan, R., Seyfarth, A., Geyer, H., Grimmer, S., Wagner, H., & Günther, M. (2007). Intelligence by mechanics. Philosophical Transactions of the Royal Society of London. Series A. Mathematical and Physical Sciences, 365 (1850), 199–220. Bretzner, F., & Drew, T. (2005). Contribution of the motor cortex to the structure and the timing of hindlimb locomotion in the cat: A microstimulation study. Journal of Neurophysiology, 94(1), 657–672. Brooke, J. D., & Zehr, E. P. (2006). Limits to fast-conducting somatosensory feedback in movement control. Exercise and Sport Sciences Reviews, 34(1), 22–28.
163 Brown, T. G. (1911). The intrinsic factors in the act of progression in the mammal. Proceedings of the Royal Society of London. Series B, Biological Science, 84(572), 308–319. Brown, T. G. (1914). On the nature of the fundamental activity of the nervous centres; together with an analysis of the conditioning of rhythmic activity in progression, and a theory of the evolution of function in the nervous system. Journal of Physiology, 48(1), 18–46. Burkholder, T. J., & Nicols, T. R. (2000). The mechanical action of proprioceptive length feedback in a model of cat hindlimb. Motor Control, 4(2), 201–220. Cavagna, G. A., Heglund, N. C., & Taylor, C. R. (1977). Mechanical work in terrestrial locomotion: Two basic mechanisms for minimizing energy expenditure. The American Journal of Physiology, 233(5), R243–R261. Cavagna, G. A., Thys, H., & Zamboni, A. (1976). The sources of external work in level walking and running. Journal de Physiologie, 262(3), 639–657. Côté, M.-P., & Gossard, J.-P. (2003). Task-dependent presynaptic inhibition. The Journal of Neuroscience, 23(5), 1886–1893. Collins, J. J., & Richmond, S. A. (1994). Hard-wired central pattern generators for quadrupedal locomotion. Biological Cybernetics, 71(5), 375–385. Courtine, G., & Schieppati, M. (2003). Human walking along a curved path. II. Gait features and EMG patterns. The European Journal of Neuroscience, 18(1), 191–205. Datteri, E., & Tamburrini, G. (2007). Biorobotic experiments for the discovery of biological mechanisms. Philosophy of Science, 74, 409–430. Daun, S., Rubin, J. E., & Rybak, I. A. (2009). Control of oscillation periods and phase durations in half-center central pattern generators: A comparative mechanistic analysis. Journal of Computational Neuroscience, 27(1), 3–36. Davis, W. J., & Ayers, J. L. (1972). Locomotion: Control by positive-feedback optokinetic responses. Science, 177(44), 183–185. de La Mettrie, J. (1745). Histoire naturelle de l'âme (pp. 1–398). Paris. Deliagina, T. G., Orlovsky, G. N., & Pavlova, G. A. (1983). The capacity for generation of rhythmic oscillations is distributed in the lumbosacral spinal cord of the cat. Experimental Brain Research, 53(1), 81–90. Deliagina, T. G., Zelenin, P. V., Fagerstedt, P., Grillner, S., & Orlovsky, G. N. (2000). Activity of reticulospinal neurons during locomotion in the freely behaving lamprey. Journal of Neurophysiology, 83(2), 853–863. Dietz, V., Zijlstra, W., & Duysens, J. (1994). Human neuronal interlimb coordination during split-belt locomotion. Experimental Brain Research, 101(3), 513–520. Donelan, J. M., McVea, D. A., & Pearson, K. G. (2009). Force-regulation of ankle extensor muscle activity in freely walking cats. Journal of Neurophysiology, 101, 360–371.
Duysens, J., & Pearson, K. G. (1980). Inhibition of flexor burst generation by loading ankle extensor muscles in walking cats. Brain Research, 187(2), 321–332. Fagerstedt, P., & Ullén, F. (2001). Lateral turns in the Lamprey. I. Patterns of motoneuron activity. Journal of Neurophysiology, 86(5), 2246–2256. Freusberg, A. (1874). Reflexbewegungen beim hunde. Pflügers Archiv: European Journal of Physiology, 9(1), 358–391. Fry, S. N., Rohrseitz, N., Straw, A. D., & Dickinson, M. H. (2009). Visual control of flight speed in Drosophila melanogaster. The Journal of Experimental Biology, 212(Pt. 8), 1120–1130. Full, R. J., & Koditschek, D. (1999). Templates and anchors: Neuromechanical hypotheses of legged locomotion on land. The Journal of Experimental Biology, 202(Pt. 23), 3325–3332. Full, R. J., Kubow, T., Schmitt, J., Holmes, P., & Koditschek, D. (2002). Quantifying dynamic stability and maneuverability in legged locomotion. Integrative and Comparative Biology, 42(1), 149–157. Georgopoulos, A. P., & Grillner, S. (1989). Visuomotor coordination in reaching and locomotion. Science, 245(4923), 1209–1210. Gillard, D., Yakovenko, S., Cameron, T., & Prochazka, A. (2000). Isometric muscle length–tension curves do not predict angle–torque curves of human wrist in continuous active movements. Journal of Biomechanics, 33(11), 1341–1348. Goaillard, J.-M., Taylor, A. L., Schulz, D. J., & Marder, E. (2009). Functional consequences of animal-to-animal variation in circuit parameters. Nature Neuroscience, 12(11), 1424–1430. Gordon, A. M., Huxley, A. F., & Julian, F. J. (1966). The variation in isometric tension with sarcomere length in vertebrate muscle fibres. Journal de Physiologie, 184(1), 170–192. Goslow, G. E., Reinking, R. M., & Stuart, D. G. (1973). The cat step cycle: Hind limb joint angles and muscle lengths during unrestrained locomotion. Journal of Morphology, 141(1), 1–41. Grashow, R., Brookings, T., & Marder, E. (2009). Reliable neuromodulation from circuits with variable underlying structure. Proceedings of the National Academy of Sciences of the United States of America, 106(28), 11742–11746. Grillner, S. (1969). The influence of DOPA on the static and the dynamic fusimotor activity to the triceps surae of the spinal cat. Acta Physiologica Scandinavica, 77(4), 490–509. Grillner, S., & Wallén, P. (1985). Central pattern generators for locomotion, with special reference to vertebrates. Annual Review of Neuroscience, 8, 233–261. Halbertsma, J. M. (1983). The stride cycle of the cat: The modelling of locomotion by computerized analysis of automatic recordings. Acta Physiologica Scandinavica, 521, 1–75.
164 Hogan, N. (1985). The mechanics of multi-joint posture and movement control. Biological Cybernetics, 52(5), 315–331. Huxley, A. F. (1974). Muscular contraction. The Journal of Physiology, 243(1), 1–43. Ivanenko, Y. P., Grasso, R., & Lacquaniti, F. (2000). Influence of leg muscle vibration on human walking. Journal of Neurophysiology, 84(4), 1737–1747. Ivanenko, Y. P., Wright, W. G., Gurfinkel, V. S., Horak, F., & Cordo, P. (2006). Interaction of involuntary post-contraction activity with locomotor movements. Experimental Brain Research, 169(2), 255–260. Jordan, M. (1996). Computational aspects of motor control and motor learning. In H. Heuer & S. W. Keele (Eds.), Handbook of perception and action: Motor skills. Vol. 2. (pp. 71–120). London: Academic Press. Joyce, G. C., Rack, P. M., & Westbury, D. R. (1969). The mechanical properties of cat soleus muscle during controlled lengthening and shortening movements. Journal de Physiologie, 204(2), 461–474. Kawato, M. (1999). Internal models for motor control and trajectory planning. Current Opinion in Neurology, 9(6), 718–727. Kiehn, O., & Kjaerulff, O. (1998). Distribution of central pattern generators for rhythmic motor outputs in the spinal cord of limbed vertebrates. Annals of the New York Academy of Sciences, 860, 110–129. Kozlov, A., Huss, M., Lansner, A., Kotaleski, J. H., & Grillner, S. (2009). Simple cellular and network control principles govern complex patterns of motor behavior. Proceedings of the National Academy of Sciences of the United States of America, 106(47), 20027–20032. Kremer, E., & Lev-Tov, A. (1997). Localization of the spinal network associated with generation of hindlimb locomotion in the neonatal rat and organization of its transverse coupling system. Journal of Neurophysiology, 77(3), 1155–1170. Kuo, A. D. (2002). The relative roles of feedforward and feedback in the control of rhythmic movements. Motor Control, 6(2), 129–145. Kuo, A. D., Donelan, J. M., & Ruina, A. (2005). Energetic consequences of walking like an inverted pendulum: Stepto-step transitions. Exercise and Sport Sciences Reviews, 33 (2), 88–97. Lawrence, J. H., & Nichols, T. R. (1999). A three-dimensional biomechanical analysis of the cat ankle joint complex: I. Active and passive postural mechanics. Journal of Applied Biomechanics, 15, 95–105. Lawrence, J. H., Nichols, T. R., & English, A. W. (1993). Cat hindlimb muscles exert substantial torques outside the sagittal plane. Journal of Neurophysiology, 69(1), 282–285. Marcoux, J., & Rossignol, S. (2000). Initiating or blocking locomotion in spinal cats by applying noradrenergic drugs
to restricted lumbar spinal segments. The Journal of Neuroscience, 20(22), 8577–8585. Markin, S. N., Klishko, A. N., Shevtsova, N. A., Lemay, M. A., Prilutsky, B. I., & Rybak, I. A. (2010). Afferent control of locomotor CPG: insights from a simple neuromechanical model. Annals of the New York Academy of Sciences, 1198, 21–34. McCrea, D. A., & Rybak, I. A. (2008). Organization of mammalian locomotor rhythm and pattern generation. Brain Research Reviews, 57(1), 134–146. McMahon, T. A., & Cheng, G. C. (1990). The mechanics of running: How does stiffness couple with speed? Journal of Biomechanics, 23(Suppl. 1), 65–78. Morton, S. M., & Bastian, A. J. (2006). Cerebellar contributions to locomotor adaptations during splitbelt treadmill walking. The Journal of Neuroscience, 26(36), 9107–9116. Mussa-Ivaldi, F. A. (1999). Modular features of motor control and learning. Current Opinion in Neurology, 9(6), 713–717. Orlovsky, G. N. (1972). The effect of different descending systems on flexor and extensor activity during locomotion. Brain Research, 40(2), 359–371. Pearson, K. G., Fouad, K., & Misiaszek, J. E. (1999). Adaptive changes in motor activity associated with functional recovery following muscle denervation in walking cats. Journal of Neurophysiology, 82(1), 370–381. Pearson, K. G., & Misiaszek, J. E. (2000). Use-dependent gain change in the reflex contribution to extensor activity in walking cats. Brain Research, 883(1), 131–134. Pearson, K. G., & Rossignol, S. (1991). Fictive motor patterns in chronic spinal cats. Journal of Neurophysiology, 66(6), 1874–1887. Perreault, M. C., Rossignol, S., & Drew, T. (1994). Microstimulation of the medullary reticular formation during fictive locomotion. Journal of Neurophysiology, 71(1), 229–245. Pfeifer, R., Lungarella, M., & Iida, F. (2007). Self-organization, embodiment, and biologically inspired robotics. Science, 318(5853), 1088–1093. Philippson, M. (1905). L'autonomie et la centralisation dans le systeme nerveux des animaux. Travaux du Laboratoire de Physiologie, Instituts Solvay (Bruxelles), 7(2), 1–208. Prentice, S. D., Patla, A. E., & Stacey, D. A. (2001). Artificial neural network model for the generation of muscle activation patterns for human locomotion. Journal of Electromyography Kinesiology, 11(1), 19–30. Prochazka, A. (1993). Comparison of natural and artificial control of movement. Rehabilitation Engineering, 1(1), 7–17. Prochazka, A. (1999). Quantifying proprioception. Progress in Brain Research, 123, 133–142. Prochazka, A., Gritsenko, V., & Yakovenko, S. (2002). Sensory control of locomotion: Reflexes versus higher-level control. Advances in Experimental Medicine and Biology, 508, 357–367.
165 Prochazka, A., & Yakovenko, S. (2007a). The neuromechanical tuning hypothesis. Progress in Brain Research, 165, 255–265. Prochazka, A., & Yakovenko, S. (2007b). Predictive and reactive tuning of the locomotor CPG. Integrative and Comparative Biology, 47(4), 474–481. Prokop, T., Schubert, M., & Berger, W. (1997). Visual influence on human locomotion. Modulation to changes in optic flow. Experimental Brain Research, 114(1), 63–70. Prochazka, A., & Sorensen, C. (2009). Biomechanical imperatives in the neural control of locomotion. Comparative Biochemistry and Physiology - Part A. Molecular & Integrative Physiology, 153(2). Supplement 1, S135–S136. Reisman, D. S., Block, H. J., & Bastian, A. J. (2005). Interlimb coordination during locomotion: What can be adapted and stored? Journal of Neurophysiology, 94(4), 2403–2415. Reisman, D. S., Wityk, R., Silver, K., & Bastian, A. J. (2007). Locomotor adaptation on a split-belt treadmill can improve walking symmetry post-stroke. Brain, 130(Pt. 7), 1861–1872. Reisman, D. S., Wityk, R., Silver, K., & Bastian, A. J. (2009). Split-belt treadmill adaptation transfers to overground walking in persons poststroke. Neurorehabilitation and Neural Repair, 23(7), 735–744. Rho, M. J., Lavoie, S., & Drew, T. (1999). Effects of red nucleus microstimulation on the locomotor pattern and timing in the intact cat: A comparison with the motor cortex. Journal of Neurophysiology, 81(5), 2297–2315. Ribotta, M., Provencher, J., Feraboli-Lohnherr, D., Rossignol, S., Privat, A., & Orsal, D. (2000). Activation of locomotion in adult chronic spinal rats is achieved by transplantation of embryonic raphe cells reinnervating a precise lumbar level. The Journal of Neuroscience, 20(13), 5144–5152. Rossignol, S., Dubuc, R., & Gossard, J.-P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86(1), 89–154. Rybak, I. A., Shevtsova, N. A., Lafreniere-Roula, M., & McCrea, D. A. (2006). Modelling spinal circuitry involved in locomotor pattern generation: Insights from deletions during fictive locomotion. Journal de Physiologie, 577 (Pt. 2), 617–639. Rybak, I. A., Stecina, K., Shevtsova, N. A., & McCrea, D. A. (2006). Modelling spinal circuitry involved in locomotor pattern generation: Insights from the effects of afferent stimulation. Journal de Physiologie, 577(Pt. 2), 641–658. Saunders, J. B., Inman, V. T., & Eberhart, H. D. (1953). The major determinants in normal and pathological gait. The Journal of Bone and Joint Surgery. American Volume, 35-A(3), 543–558. Sherrington, C. S. (1910). Flexion-reflex of the limb, crossed extension-reflex, and reflex stepping and standing. Journal de Physiologie, 40(1–2), 28–121.
Shik, M. L., & Orlovsky, G. N. (1976). Neurophysiology of locomotor automatism. Physiological Reviews, 56(3), 465–501. Shik, M. L., Severin, F. V., & Orlovskiĭ, G. N. (1966). Control of walking and running by means of electric stimulation of the midbrain. Biofizika, 11(4), 659–666. Shik, M. L., Severin, F. V., & Orlovsky, G. N. (1969). Control of walking and running by means of electrical stimulation of the mesencephalon. Electroencephalography & Clinical Neurophysiology, 26(5), 549. Sinkjaer, T., Andersen, J. B., Ladouceur, M., Christensen, L. O., & Nielsen, J. B. (2000). Major role for sensory feedback in soleus EMG activity in the stance phase of walking in man. Journal de Physiologie, 523(Pt. 3), 817–827. Sirota, M. G., Di Prisco, G. V., & Dubuc, R. (2000). Stimulation of the mesencephalic locomotor region elicits controlled swimming in semi-intact lampreys. The European Journal of Neuroscience, 12(11), 4081–4092. Srinivasan, M., & Ruina, A. (2006). Computer optimization of a minimal biped model discovers walking and running. Nature, 439(7072), 72–75. Stein, R. B., Misiaszek, J. E., & Pearson, K. G. (2000). Functional role of muscle reflexes for force generation in the decerebrate walking cat. Journal de Physiologie, 525(Pt. 3), 781–791. Stuart, D. G., & Hultborn, H. (2008). Thomas Graham Brown (1882–1965), Anders Lundberg (1920-), and the neural control of stepping. Brain Research Reviews, 59(1), 74–95. Taga, G. (1995). A model of the neuro-musculo-skeletal system for human locomotion. I. Emergence of basic gait. Biological Cybernetics, 73(2), 97–111. Taga, G. (1998). A model of the neuro-musculo-skeletal system for anticipatory adjustment of human locomotion during obstacle avoidance. Biological Cybernetics, 78(1), 9–17. Taga, G., Yamaguchi, Y., & Shimizu, H. (1991). Self-organized control of bipedal locomotion by neural oscillators in unpredictable environment. Biological Cybernetics, 65(3), 147–159. Verzar, F. (1923). Reflexumkehr (paradoxe Reflexe) durch zentrale Ermudung beim Warmbluter. Pflügers Archiv: European Journal of Physiology, 199, 109–124. Wilmink, R. J. H., & Nichols, T. R. (2003). Distribution of heterogenic reflexes among the quadriceps and triceps surae muscles of the cat hind limb. Journal of Neurophysiology, 90(4), 2310–2324. Yakovenko, S., Gritsenko, V., & Prochazka, A. (2004). Contribution of stretch reflexes to locomotor control: A modeling study. Biological Cybernetics, 90(2), 146–155. Yakovenko, S., Kowalczewski, J., & Prochazka, A. (2007). Intraspinal stimulation caudal to spinal cord transections in rats. Testing the propriospinal hypothesis. Journal of Neurophysiology, 97(3), 2570–2574.
166 Yakovenko, S., McCrea, D. A., Stecina, K., & Prochazka, A. (2005). Control of locomotor cycle durations. Journal of Neurophysiology, 94(2), 1057–1065. Yang, J. F., Stein, R. B., & James, K. B. (1991). Contribution of peripheral afferents to the activation of the soleus muscle
during walking in humans. Experimental Brain Research, 87(3), 679–687. Zajac, F. E. (1989). Muscle and tendon: Properties, models, scaling, and application to biomechanics and motor control. Critical Reviews in Biomedical Engineering, 17(4), 359–411.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 11
Novel mechanism for hyperreflexia and spasticity C. Yates{,{, K. Garrison{,{, N. B. Reese{, A. Charlesworth{ and E. Garcia-Rill{,* {
Center for Translational Neuroscience, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA { Department of Physical Therapy, University of Central Arkansas, Conway, Arkansas, USA
Abstract: We established that hyperreflexia is delayed after spinal transection in the adult rat and that passive exercise could normalize low frequency-dependent depression of the H-reflex. We were also able to show that such passive exercise will normalize hyperreflexia in patients with spinal cord injury (SCI). Recent results demonstrate that spinal transection results in changes in the neuronal gap junction protein connexin 36 below the level of the lesion. Moreover, a drug known to increase electrical coupling was found to normalize hyperreflexia in the absence of passive exercise, suggesting that changes in electrical coupling may be involved in hyperreflexia. We also present results showing that a measure of spasticity, the stretch reflex, is rendered abnormal by transection and normalized by the same drug. These data suggest that electrical coupling may be dysregulated in SCI, leading to some of the symptoms observed. A novel therapy for hyperreflexia and spasticity may require modulation of electrical coupling. Keywords: electrical coupling; gap junctions; H-reflex; hyperreflexia; spasticity; spinal cord injury; stretch reflex.
hippocampus (Traub et al., 2003), olfactory bulb (Friedman and Strowbridge, 2003), amygdala (Sinfield and Collins, 2006), inferior olive (Leznik and Llinas, 2005; Long et al., 2002), and locus coeruleus (Christie et al., 1989). In addition, electrical coupling is evident in respiratory brainstem regions (Bou-Flores and Berger, 2001; Rekling and Feldman, 1997) and spinal cord (Connors and Long, 2004; Kiehn and Tresch, 2002). Gap junction blockers have been reported to depress rhythmogenesis in the respiratory pre-Botzinger complex (Elsen et al., 2008), the retrotrapezoid
Introduction Electrical coupling in breathing, walking, and chewing Oscillatory activity mediated by electrical coupling via gap junctions has been demonstrated in a number of regions including the neocortex (Blatow et al., 2003; Roopun et al., 2006), *Corresponding author. Tel.: þ1-501-686-5167; Fax: þ1-501-526-7928 DOI: 10.1016/B978-0-444-53825-3.00016-4
167
168
nucleus (Hewitt et al., 2004), and the nucleus of the tractus solitarius (Parisian et al., 2004). Although few studies on chewing and swallowing have been carried out using gap junction blockers, there is little doubt that gap junctions are present in the trigeminal nucleus (Honma et al., 2004), and hypoglossal (Cifra et al., 2009) neurons are electrically coupled. Jim Lund, in collaboration with Verdier and Kolta, was instrumental in determining that at least some electrical coupling is present in the trigeminal mesencephalic nucleus (Verdier et al., 2004). As far as locomotion is concerned, electrical coupling has been found to modulate swimming in tadpoles (Li et al., 2009) and below we discuss evidence in mammals. How do spinal cord gap junctions influence motor coordination? Electrical synapses may contribute to the generation and maintenance of synchronized neuronal bursting firing patterns (Kistler et al., 2002). Others demonstrated that motor patterns in the neonatal rat spinal cord were observed during blockade of chemical synapses, probably through the synchronization of bursting through gap junctions (Tresch and Kiehn, 2000). They suggested, “Gap junctionmediated neuronal coordination contributes to the basic function and organization of spinal motor systems.” These authors also suggested the existence of numerous independent rhythms in distinct motor pools. However, the modulation of gap junction communication in the adult spinal system and the adult system after injury are poorly understood.
Hyperreflexia and spasticity in spinal cord injury Spinal cord injury (SCI) results in numerous deficits of motor and sensory systems, including paralysis, anesthesia, hyperreflexia, and spasticity below the level of the lesion. Many of these individuals will require lifetime care due to motor deficits suffered as a result of the injury. Hyperreflexia and spasticity are evident in both humans
and animals following SCI and can contribute to significant functional limitation experienced by patients following SCI. The mechanisms responsible for hyperreflexia are not known, however, several observed physiological changes that have been postulated to contribute to hyperreflexia include alpha motoneuron hyperexcitability (Landau and Clare, 1964; Magladery et al., 1952; Milanov, 1994), changes in the intrinsic properties of alpha motoneurons (Bennett et al., 2001; Eken et al., 1989; Li and Bennett, 2003), reduced postactivation depression of transmission from Ia fibers (Hultborn, 2003; Nielsen et al., 1995), synapse growth (Little et al., 1999), alterations in morphology of alpha motoneurons (Kitzman, 2005), and decreased presynaptic inhibition of Ia terminals (Angel and Hofmann, 1963; Calancie et al., 1993; Faist et al., 1994; Nielsen et al., 1995; Pierrot-Deseilligny, 1990; Schindler-Ivens and Shields, 2000). The time course of spinal changes after injury has been proposed to include an early postsynaptic mechanism, possibly involving an increase in excitability and/or receptor upregulation, and a late change involving presynaptic mechanisms possibly involving synaptic growth in spared descending pathways and in reflex pathways (Kitzman, 2005). Spasticity is classically referred to as resistance to passive limb movement in proportion to the velocity of movement (Lance, 1980). This velocity-dependent resistance is thought to be due to increased stretch reflex responses in the lengthened muscle (Ju et al., 2000; Kuhn, 1950; Powers and Rymer, 1988), although increased stiffness in limb compliance is also a factor (Dietz et al., 1981; Sinkjaer et al., 1993). A test of excitability, in general, and of frequency-dependent depression can also be applied to the stretch reflex (Thompson et al., 2001). Therefore, the use of both H-reflex and stretch reflex would yield assessments of hyperreflexia and spasticity. However, the utility of the H-reflex test is limited by the invasive nature of the test (direct exposure of the tibial nerve to provide stimulation), typically a terminal procedure, and the inability to
169
assess the velocity-dependent nature of spasticity. The development of (a) a preparation in which reflexes are measured longitudinally and (b) a mechanical stretching device would overcome these limitations and enable studies of the emergence of hyperreflexia and spasticity post-SCI and effects of interventions. Such developments are discussed below.
The H-reflex One measure used by numerous investigators to quantify hyperreflexia is the electrical analogue of the classic tendon jerk reflex, the Hoffman or Hreflex (Angel and Hofmann, 1963; Faist et al., 1994; Little and Halar, 1985; Olsen and Diamantopoulos, 1967; Yablon and Stokic, 2004). The H-reflex is a compound electromyographic (EMG) response elicited by the synaptic activation of motoneurons by muscle afferents following stimulation of muscle nerves. Thompson et al. (1992) investigated four measures of H-reflex excitability in a contusion model of SCI in the rat. Results of their studies led these researchers to conclude that rate-sensitive depression of the Hreflex was of particular importance in the assessment of hyperreflexia following SCI. Other groups have reached similar conclusions regarding the importance of changes in H-reflex rate-sensitive depression as a measure of the effects of SCI (Chen et al., 2001). In spinally intact individuals, the Hreflex demonstrates depressed amplitude, due to marked frequency-dependent depression, once stimulus frequencies reach or exceed 1 Hz (Ishikawa et al., 1966; Kiser et al., 2005). However, frequency-dependent depression of the H-reflex is less evident in patients or animals with chronic SCI (Calancie et al., 1993; Ishikawa et al., 1966; Kiser et al., 2005; Skinner et al., 1996). Schindler-Ivens and Shields (2000) examined changes in the human H-reflex in a longitudinal study over 44 weeks and concluded that attenuation of frequency-dependent depression occurs gradually. They also reported gradual changes
from high- to low-rate sensitivity between 6 and 18 weeks in the human that correlated with the transition from flaccid to a spastic state (Schindler-Ivens and Shields, 2004). Thompson et al. (1992) examined changes in reflex excitability in the rat contusion model and determined that there was no difference in control animals compared with animals 6 days postcontusion regarding the threshold for reflex initiation. They also concluded that animals that were 28 days postcontusion were significantly different than control animals in rate-dependent depression at 1–5 Hz. Animals continued to be significantly different than controls at 60 days postcontusion. The fact that hyperreflexia does not set in immediately after transection, and that reflexes can be elicited with normal habituation at 6 days, suggests that a major component of hyperreflexia involves reorganization and perhaps changes in structure, rather than the elimination of presynaptic effects by the lesion. That is, whatever changes take place in the spinal cord after transection require > 1 week to become manifested.
The stretch reflex The noninvasive measurement of EMG and torque response to a movement perturbation has been reported in the human (Schmit et al., 1999, 2002). Thompson et al. (1996) have also developed a device to quantify the stretch reflex in rats and documented the velocity-dependent response in normal rats and in rats with a contusion injury of the spinal cord (Bose et al., 2002). Furthermore, Thompson documented the time course of the response postinjury. However, the stretch reflex response to a complete transection injury has not been reported, and evidence from human studies indicates that responses differ based on completeness of injury (Calancie et al., 2002; Nakazawa et al., 2006). From a methodological standpoint, single session quantification of the EMG response to imposed stretch is possible, but longitudinal comparison of EMG responses in rodents is
170
problematic because of the inability to normalize the EMG signal. Validity of nonnormalized comparisons can be compromised because measured differences could be due to altered electrode location or spacing rather than a true treatment effect (Soderberg and Knutson, 2000). One paradigm that overcomes this limitation is the study of the windup of repeated stretches. In a windup protocol, the EMG response to the first stretch is used to normalize subsequent responses. Windup behavior is characterized by temporal facilitation that results in increased amplitude and duration of the reflex responses. This has been demonstrated for flexor reflexes (Hornby et al., 2003), and recently in stretch reflexes in humans post-SCI (Hornby et al., 2006). It has been suggested that the mechanism for this prolonged reflex response is due to alterations in intrinsic motoneuron properties, namely persistent inward currents (PICs) (Bennett et al., 1998). In reduced preparations, PICs demonstrate the ability to amplify and prolong the response to brief inputs (Crone et al., 1988; Hounsgaard et al., 1988), and their reemergence is linked to the onset of hyperreflexia (Bennett et al., 1999, 2001). Therefore, we developed an actuator that is capable of producing controlled stretches using (a) a windup protocol to provide another measure of hyperreflexia and (b) a velocity protocol to simultaneously provide a measure of changes in spasticity due to interventions such as exercise and pharmacological agents. Coupled with the ability to carry out these measures longitudinally, the serious limitations in this type of research can be overcome, and we can now assess the effects of passive exercise and pharmacological interventions as abnormal symptoms develop and are reversed.
Passive exercise Role We developed a method of providing continuous passive alternating movement of the hind limbs,
known as motorized bicycle exercise training (MBET), to rats that had undergone T10 spinal cord complete transection. We reported the ability of long-term passive exercise therapy to produce frequency-dependent depression of the Hreflex in adult rats with complete spinal cord transection. We find that a period of 3 months of MBET, performed in sessions of 1 h/day, 5 days/ week, was capable of producing frequency-dependent depression of the H-reflex equivalent to that of intact animals (Reese et al., 2006; Skinner et al., 1996). Most of the H-reflex work in rats has involved acute testing where the animals are anesthetized, undergo H-reflex testing, and sacrificed at the end of the experiment (Meinck, 1976; Skinner et al., 1996; Thompson et al., 1992). Recently, we developed a method of testing the H-reflex longitudinally in the same rats using percutaneous electrodes (Arfaj et al., 2007). This method allows us to perform the testing in unanesthetized animals, more closely mimicking the testing conditions employed with human subjects. In addition, the same animals can be tested on multiple occasions throughout the course of the experiment. Thus, use of the percutaneous electrode method of testing requires fewer animals and allows one to track changes in the same animal over time in response to the intervention being applied. Our studies employ groups of spinally transected animals that are followed over time to determine the time course of the decrease in hyperreflexia (assessed using the H-reflex) and spasticity (assessed using the stretch reflex) seen with exercise, and if a shorter duration of exercise is capable of producing similar degrees of H-reflex frequency-dependent depression or stretch reflex habituation in adult rats following spinal cord transection. Hyperreflexia and spasticity are present following a multitude of upper motor neuron disorders. Pharmacological agents such as Baclofen, Diazepam, Tizanidine, Dantrolene, and others have been used in human subjects in an attempt to decrease spasticity following SCI. However, all of these drugs have some undesirable side effects
171
Effects on the H-reflex Using the complete transection model in the adult rat, we recently reported that rate-dependent depression in spinally transected animals, like contusion model animals, become significantly different from nontransected controls when tested 30 days posttransection (Reese et al., 2006). We also found that hyperreflexia does not set in immediately posttransection and that it may be present as early as 14 days posttransection (see Fig. 1) (Yates et al., 2008a). Determination of the exact time course for the onset of hyperreflexia provides us with the framework around which to (a) examine the molecular and cellular H-reflex
120
Control Tx + 7D Tx + 14D
100 H-reflex ave. amplitude (%)
on patients and some have no effects on hyperreflexia (Kita and Goodkin, 2000). The effects of modafinil (MOD) on hyperreflexia and spasticity have not been investigated extensively, despite the fact that we found it is effective in treating patients with neglect after stroke injury (Woods et al., 2006). Moreover, recent reports suggest MOD may be useful in the treatment of spasticity arising from cerebral palsy (Hurst and Cedrone, 2006; Hurst et al., 2006). Why would a stimulant used as an antinarcoleptic agent have salutary effects on spasticity? The mechanism of action of MOD was unknown until recently, but was credited with increasing glutamate, acetylcholine, noradrenaline, and serotonin release and decreasing GABA release (Ballon and Feifel, 2006). However, MOD was recently found to increase electrical coupling between nerve cells in the inferior olivary nucleus, cortical interneurons, and thalamic reticular neurons (Urbano et al., 2007). Following pharmacological blockade of connexin permeability, MOD restored electrotonic coupling within 30 min. The effects of MOD were counteracted by the gap junction blocker mefloquine. These authors proposed that MOD may be acting in a wide variety of cerebral areas by increasing electrotonic coupling in such a way that the high input resistance typical of GABAergic neurons is reduced. This “shunting effect” of MOD may activate the entire thalamocortical system by slightly diminishing inhibitory networks and, at the same time, increasing synchronous activation of both interneurons and noninhibitory neurons. We confirmed that MOD increased electrical coupling in cell groups in the reticular activating system, thus accounting for its stimulant effects on arousal (Garcia-Rill et al., 2007; Heister et al., 2007). Studies described below show that daily oral administration of MOD can also be used to prevent the loss of frequency-dependent depression of the H-reflex in acutely spinalized rats. These findings raise a number of intriguing questions regarding the mechanisms behind hyperreflexia and spasticity.
Tx + 30D Tx + Ex Tx + MOD
80
60
40
20
0
0
2
4 6 8 Frequency (Hz)
10
12
Fig. 1. H-reflex. H-reflex amplitude at 0.2, 1, 5, and 10 Hz for intact animals (control, open circles, n ¼ 10), in rats 7 days (Tx þ 7D, open squares, n ¼ 7), 14 days (Tx þ 14D, filled triangles, n ¼ 7), 30 days after transection (Tx þ 30D, filled squares, n ¼ 16), in rats 30 days after MBET treatment (Tx þ Ex, open triangles, n ¼ 16), and in rats 30 days after daily modafinil (Tx þ MOD, filled circle, n ¼ 16). Frequencydependent depression of the H-reflex at 0.2 Hz was designated 100%, and statistical comparisons made against the transection-only 30-day group. Note that (a) hyperreflexia does not set in until 14 days after transection, (b) both passive exercise and MOD independently prevent hyperreflexia after transection.
172
mechanisms involved and (b) determine optimal parameters for interventions following SCI. We measured group mean H-reflex amplitudes at 0.2, 1, 5, and 10 Hz for intact animals (control, open circles, n ¼ 10), in rats 7 days (Tx þ 7D), 14 days (Tx þ 14D), and 30 days posttransection (Tx þ 30D). Frequency-dependent depression of the H-reflex at 0.2 Hz was designated as 100%, and all statistical comparisons in this figure were made against the transection-only group (Tx þ 30D). At 10 Hz, the Tx þ 30D group differed (p < 0.01) from the Tx þ 7D, and control groups. These results suggest that the onset of hyperreflexia may occur between 7 and 14 days, although the variance of means at 14 days (note error bar) was greater than at 7 days while reflexes were normal in all animals, and 30 days when they were increased in all animals. This suggests that 14 days is a transition period during which the reflexes in some animals were only marginally increased (Yates et al., 2008a). Passive exercise for 30 days instituted 1 week after transection (Tx þ Ex) was able to prevent the loss of low frequency-dependent depression of the H-reflex (Reese et al., 2006). Further studies (data not shown) demonstrated that passive exercise instituted after hyperreflexia had set in required a longer duration of exercise therapy, and longer durations of exercise produced greater savings in the reduction of hyperreflexia once therapy ceased (Yates et al., 2008b). In addition, results from our lab indicate that oral administration of MOD (4 mg/kg, p.o.) over a period of 30 days was able to prevent the loss of frequency-dependent depression of the Hreflex when tested at 10 Hz (Tx þ MOD). At 5 and 10 Hz, the H-reflex habituation in the Tx þ 30D group differed (p < 0.01) from the control group, the MBET 30 days (Tx þ Ex), and the MOD group (Tx þ MOD). These results for the first time suggest that pharmacological intervention with an agent that increases electrical coupling may be useful in reducing hyperreflexia. These data indicate that MOD was as effective as passive exercise in normalizing low frequency-
dependent depression of the H-reflex. These results also suggest the possibility that at least some of the changes occurring between 1 and 2 weeks after transection involve changes in gap junction function (Yates et al., 2009).
Effects on the stretch reflex Using changes in stretch reflex windup to track the effects of transection, we found that the stretch reflex became hyperactive much later than the H-reflex, with the windup of plantarflexion torque and gastrocnemius EMG emerging 7 weeks after transection (Fig. 2). Plotting the responses from various stretch intervals revealed that windup peaked approximately 1 s from the first stretch with habituation occurring by 2 s. Figure 3 shows peak windup of the stretch reflex torque for three groups. Animals not receiving an intervention (Tx) demonstrated significant windup at 7 weeks (p ¼ 0.01*). In contrast, neither MBET- (Tx þ Ex) nor MOD-treated animals (Tx þ MOD) developed windup, and normalized torque at 7 weeks was significantly lower than in the animals with transection only (Tx) (p ¼ 0.01 and p ¼ 0.05, respectively). These data demonstrate that (a) the stretch reflex habituation becomes abnormal much later after transection than the H-reflex, (b) MBET nevertheless could normalize the stretch reflex, and (c) MOD was also able to prevent the changes in the stretch reflex in the absence of MBET. These findings suggest that, although the mechanisms behind hyperreflexia and spasticity undergo changes with different time courses, the stretch reflex is responsive to MBET and MOD in a manner similar to the H-reflex. Are passive exercise (MBET) and MOD addressing the same underlying mechanism? The answer is not known. We do not yet know if passive exercise has its salutary effects by modulating electrical coupling and Cx-36, along with other mechanisms. The fact that the onsets of hyperreflexia (H-reflex) and exaggerated stretch reflex are different suggests
173 Integrated EMG
Peak Torque 3
49 Days
35 Days
21 Days
7 Days
7 5
1s 0.4 s 0.25 s
2
3 1
1
7
3
5
2
3 1
1
7
3
5
2
3 1
1
7
3
5
2
3 1
1 1
2
3
4
5
6
7
8
9 10
1
2
3
4
5
6
7
8
9
10
Time (s)
Time (s)
Fig. 2. Stretch reflex windup after spinal transection in rats (n ¼ 8). A series of 10 dorsiflexion stretches (600 /s, 20 ) were performed at different intervals (1, 0.4, and 0.25 s) and overplotted to show the time course of the windup and habituation of the stretch reflex. Integrated EMG of the lateral gastrocnemius and peak plantarflexion torque responses to repeated stretch was normalized to the first stretch of the series and plotted for 7, 21, 35, and 49 days posttransection. Although some variance was observed, a consistent pattern of windup was not present until 49 days.
that they are mediated by different mechanisms. Other studies have shown differences in these two measures in regard to sensitivity to GABAergic presynaptic inhibition (Morita et al., 1998), temporal dispersion of the afferent volley (Enriquez-Denton et al., 2002), and postactivation depression (Grey et al., 2008). However, both measures are normalized in response to passive exercise and MOD, suggesting related mechanisms. Much work needs to be undertaken before we can answer these questions, but these need not await the application of these novel therapeutic avenues to the SCI patient. We were successful in demonstrating that passive exercise using MBET could normalize hyperreflexia in an
ASIA B C7 SCI patient (Kiser et al., 2005). We believe this technology may also be fruitful in restoring muscle mass after prolonged (> 6 months) MBET. Moreover, we filed a patent for the use of MOD for the treatment of hyperreflexia and spasticity and hope to institute clinical trials in the near future.
Electrical coupling Role The role of electrotonic coupling in modulating motor behavior has not been extensively studied.
174 Stretch Reflex Torque 350%
**
Percent Windup
300% 250%
*
200% 150%
*
100% 50% 0%
7
21 35 Days Post Tx Tx
Tx + MOD
49
Tx + Ex
Fig. 3. Effects of exercise and modafinil on windup of the stretch reflex. Group averages are plotted for peak windup of plantarflexion torque at 7, 21, 35, and 49 days posttransection (Tx). Mean values (S.D.) were plotted as a percentage of the baseline (first) stretch for data from trials with 0.4 s interval between stretches. Animals in the transection-only (Tx, n ¼ 8) group demonstrated significant windup at 49 days (p ¼ 0.01*). The exercise (Tx þ Ex, n ¼ 8) group did not develop significant windup but were significantly different than the Tx group (p ¼ 0.01). Animals treated with modafinil (Tx þ MOD, n ¼ 8) also failed to develop windup of the stretch reflex by 49 days, and peak torque was significantly lower than the Tx group (p ¼ 0.05).
Interestingly, if the spinal cord of the rat is transected before 15 days of age, the animal regains the ability to generate locomotor movements on a treadmill, but if the Tx is performed after 15 days of age, this function is lost (Stelzner et al., 1975). The period in which the spinal cord is optimal for plastic changes in the rat has been determined to be 18 days of age (Fawcett, 2006). What is occurring in development that could contribute to the changes in plasticity at the end of the critical period? Electrical coupling of neurons via gap junctions has been shown to exist in early postnatal development of spinal motoneurons (Fulton et al., 1980). Motoneuron electrotonic coupling has been shown to decrease with postnatal age in the rat from postnatal day 0 to 14
(Walton and Navarrete, 1991). Such coupling may allow stronger synchronized contractions in weak muscles during development, for example, when chicks must break the eggshell at hatching (Wenner and O'Donovan, 2001). Does a correlation exist between the end of the critical period of developmental plasticity in the rat spinal cord and the decreased gap junction coupling observed? The decrease in coupling has been attributed to the development of fine motor control by independent recruitment of motor units (Bradley, 1990). These authors suggested that coupling of motoneurons by gap junctions is function specific with the most prevalent coupling between motoneurons innervating the same muscle, reduced with synergistic muscles, and absent with antagonist muscles (Bradley, 1990). A recent important study showed that there are populations of locomotion-related interneurons in the ventromedial gray matter that are electrically coupled (Hinckley and ZiskindConhaim, 2006), suggesting that not only motoneuron pools may be coupled. In general, electrical synapses allow the reciprocal flow of ionic currents and small molecules between neurons, often providing synchronization of subthreshold and spiking activity (Connors and Long, 2004). Connexins form clusters of channels that allow direct cell to cell communication. Electrical communication between neurons has been attributed to gap junctions made up of Connexin 36 (Cx-36). In mammals, Cx-36 is specifically expressed in neurons (Condorelli et al., 1998). Deans et al. (2001) studied the knockout (KO) mouse of Cx-36 and determined that many physiologic properties of the KO were similar to the wild type with the exception that electrical coupling in the KO mouse was rare and weak compared with the wild type supporting the theory that gap junctions comprised Cx-36 are responsible for electrical coupling. The best studied electrical synapses are in the inferior olive, where the oscillatory properties of single neurons endow the system with important dynamics, but it is the gap junctions that are needed for
175 control
Tx + 7D Tx + 14D Tx + 30D Tx + Ex Tx + MOD
Cx 36 5000 4500 4000
Cx 36 levels (arb units)
synchronized neuronal ensemble activity (Leznik and Llinas, 2005). What functional significance can be accorded to electrical coupling if Cx-36 KO mice breathe, chew, and walk? Several studies showed that while macroscopic motor activity patterns appeared normal in the Cx-36 KO mouse, detailed analysis of motor patterns showed a 10–20 ms degradation in coordination (Placantonakis et al., 2004) and a delay > 20 ms in the optokinetic reflex (Kistler et al., 2002). These studies taken together suggest that gap junctions confer an advantage in timing, probably due to their ability to promote coherence in brain rhythms. We analyzed the level of Cx-36 protein in cored samples from the lumbar enlargement of these animals (Yates et al., 2008a,b, 2009). Figure 4 shows that Cx-36 protein levels decreased compared to levels in intact animals at 7 days after transection, but slowly returned to control levels over the next time points. Immunoprecipitation followed by Western blot was used to quantify the amount of Cx-36 in spinal cord. A transient decrease below the level of the lesion in Cx-36 protein levels after transection was observed. Figure 4 shows the Cx-36 Western blot following immunoprecipitation from spinal cord from control rats or 7, 14, or 30 days after transection (Tx þ 7D, Tx þ 14D, or Tx þ 30D) and after passive exercise or MOD (Tx þ Ex, Tx þ MOD). This figure also shows quantification of the data shown in the upper panel. Note that Cx-36 protein level decreased significantly at 7 days (p > 0.01). We showed that after transection, Cx-36 protein transiently decreased by 7 days, but returned to control levels within 30 days. The return of Cx36 control levels paralleled the onset of hyperreflexia (Yates et al., 2008a). In another study, we showed that when spinal transected rats were passively exercised, hyperreflexia was normalized and Cx-36 levels were slightly higher than in unexercised rats, suggesting that exercise may change Cx-36 protein levels (Yates et al., 2008b). Hyperreflexia could also be normalized with MOD (Yates et al., 2009), although there
3500 3000 2500 2000 1500 1000 500 0 control
Tx + 7D Tx + 14D Tx + 30D Tx + Ex Tx + MOD
Fig. 4. Cx 36 protein levels transiently decrease after SCI. Upper panel, immunoprecipitation followed by Western blot of Cx 36 protein. Lower panel, quantification of the Western blot (arbitrary units). Adult rats underwent spinal cord transection. After 7 (Tx þ 7D), 14 (Tx þ 14D), or 30 (Tx þ 30D) days following transection, the entire lumbar region of the spinal cord was assayed for Cx 36 protein levels. Seven days following transection, one group of rats was treated with passive exercise for 30 days (Tx þ Ex), while another group was given MOD for 30 days (Tx þ MOD), and their Cx 36 protein levels assayed. There was a transient decrease in Cx 36 protein after 7 days, but it returned to normal levels over the course of a few weeks.
was no evidence for changes in protein levels. Therefore, while passive exercise and MOD could both target electrical coupling, they may do so in different ways. Another issue is the potential site of action of MOD at the level of the spinal cord. It is known that motoneurons are extensively coupled during development, but coupling decreases by 14 days postnatally in the rat (Walton and Navarrete, 1991). As mentioned above, there are a number of interneurons that were found to be electrically
176
coupled in the ventromedial region of the spinal cord and appear related to locomotor control (Hinckley and Ziskind-Conhaim, 2006). It is not known if MOD directly affects electrical coupling in the transected adult rat, but if it does, it may do so by influencing motoneurons, locomotionrelated interneurons, and/or perhaps even GABAergic spinal interneurons, as it does in the brain (Garcia-Rill et al., 2007; Heister et al., 2007; Urbano et al., 2007). We are exploring the possibility that overall levels of Cx-36 protein may not yield an accurate picture of changes after spinal transection and treatments like exercise and MOD, but regional measures within the spinal cord may reveal differential changes in different regions. While a number of new questions are raised by these studies, our results strongly suggest that MOD may represent a valuable therapeutic adjunct to the treatment of SCI and supports previous results in cerebral palsy. However, some of the gains noted in other studies may have been due to direct effects on spinal circuitry rather than cerebral in origin (Hurst and Cedrone, 2006; Hurst et al., 2006).
Future directions The data summarized suggest a number of experiments to confirm the role of gap junctions in chewing, breathing, and walking. While the involvement of electrical coupling has been tested in these systems using gap junction blockers, each of which has side effects, the role of MOD has not been tested. Our hypothesis would predict that MOD will increase the amplitude of oscillations in all of these systems, an effect that should be blocked by carbenoxolone and mefloquine (we recommend the use of multiple gap junction blockers that have different side effects). We would also predict that changes in Cx-36 protein should be evident in preparations in which each of these functions, such as synchronization, is dysregulated. Interestingly, a number of very
disparate pathological conditions could be amenable to MOD therapy. A more general direction is emerging from these results suggesting that it is rhythmicity that is the common factor in such disparate functions as breathing, chewing, and walking. Years ago, we proposed that regions of the brainstem that could be electrically (Atsuta et al., 1988) and chemically stimulated (Atsuta et al., 1991; Garcia-Rill et al., 1985) were not so much induced as “recruited” to result in stepping (Garcia-Rill, 1991; Garcia-Rill and Skinner, 1988). That is, these regions were proposed to be rhythmogenic rather than function specific. By this, we mean that these centers are not “locomotor” per se, but rather rhythmogenic regions that are electrically recruited, and it is the synchronization across rhythms that result in overt movement. The idea is that the oscillatory coupling between cell groups in these regions is responsible for generating various frequencies. The same can be said for regions involved in breathing and chewing, and the differences between these systems may arise because of the differing biomechanical requirements. This would help explain the considerable similarities in the structure of the premotor neurons as well as the pattern generators modulating these rhythmic functions. Conclusions The role of electrical coupling in rhythmgenerating systems is essential. The data presented suggest that electrical coupling may be dysregulated in SCI, leading to some of the symptoms observed. A novel therapy for hyperreflexia and spasticity may require modulation of electrical coupling using an agent that increases electrical coupling. Acknowledgment Supported by NIH award P20 RR20146, and NS062363
177
Abbreviations Cx-36 EMG GABA MBET MOD SCI Tx
connexin 36 electromyogram gamma amino-butyric acid motorized bicycle exercise trainer modafinil spinal cord injury transection
References Angel, R. W., & Hofmann, W. W. (1963). The H reflex in normal, spastic, and rigid subjects. Archives of Neurology, 9, 591–596. Arfaj, A., Yates, C., Reese, N. B., Ishida, K., Skinner, R. D., & Garcia-Rill, E. (2007). Changes in the H-reflex after spinal cord injury: A longitudinal study in awake rats. Neuroscience—Abstract, 33, 405–422. Atsuta, Y., Abraham, P., Iwahara, T., Garcia-Rill, E., & Skinner, R. D. (1991). Control of locomotion in vitro: II. Chemical stimulation. Somatosensory & Motor Research, 8, 55–63. Atsuta, Y., Garcia-Rill, E., & Skinner, R. D. (1988). Electrically induced locomotion in the in vitro brainstem-spinal cord preparation. Brain Research, 470, 309–312. Ballon, J. S., & Feifel, D. (2006). A systematic review of modafinil: Potential clinical uses and mechanisms of action. The Journal of Clinical Psychiatry, 67, 554–566. Bennett, D. J., Gorassini, M., Fouad, K., Sanelli, L., Han, Y., & Cheng, J. (1999). Spasticity in rats with sacral spinal cord injury. Journal of Neurotrauma, 16, 69–84. Bennett, D. J., Hultborn, H., Fedirchuk, B., & Gorassini, M. (1998). Short-term plasticity in hindlimb motoneurons of decerebrate cats. Journal of Neurophysiology, 80, 2038–2045. Bennett, D. J., Li, Y., Harvey, P. J., & Gorassini, M. (2001). Evidence for plateau potentials in tail motoneurons of awake chronic spinal rats with spasticity. Journal of Neurophysiology, 86, 1972–1982. Blatow, M., Rozov, A., Katona, I., Hormuzdi, S. G., Meyer, A. H., Whittington, M. A., et al. (2003). A novel network of multipolar bursting interneurons generates theta frequency oscillations in neocortex. Neuron, 38, 805–817. Bose, P., Parmer, R., & Thompson, F. J. (2002). Velocitydependent ankle torque in rats after contusion injury of the midthoracic spinal cord: Time course. Journal of Neurotrauma, 19, 1231–1249.
Bou-Flores, C., & Berger, A. J. (2001). Gap junctions and inhibitory synapses modulate inspiratory motoneuron synchronization. Journal of Neurophysiology, 85, 1543–1551. Bradley, N. S. (1990). Animal models offer the opportunity to acquire a new perspective on motor development. Physical Therapy, 70, 776–787. Calancie, B., Broton, J. G., Klose, K. J., Traad, M., Difini, J., & Ayyar, D. R. (1993). Evidence that alterations in presynaptic inhibition contribute to segmental hypo- and hyperexcitability after spinal cord injury in man. Electroencephalography and Clinical Neurophysiology, 89, 177–186. Calancie, B., Molano, M. R., & Broton, J. G. (2002). Interlimb reflexes and synaptic plasticity become evident months after human spinal cord injury. Brain, 125, 1150–1161. Chen, X. Y., Feng-Chen, K. C., Chen, L., Stark, D. M., & Wolpaw, J. R. (2001). Short-Term and medium-term effects of spinal cord tract transections on soleus H-reflex in freely moving rats. Journal of Neurotrauma, 18, 313–327. Christie, M. J., Williams, J. T., & North, R. A. (1989). Electrical coupling synchronizes subthreshold activity in locus coeruleus neurons in vitro from neonatal rats. The Journal of Neuroscience, 9, 3584–3589. Cifra, A., Nani, F., Sharifullina, E., & Nistri, A. (2009). A repertoire of rhythmic bursting produced by hypoglossal motoneurons in physiological and pathological conditions. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 364, 2493–2500. Condorelli, D. F., Parenti, R., Spinella, F., Trovato Salinaro, A., Belluardo, N., Cardile, V., et al. (1998). Cloning of a new gap junction gene (Cx36) highly expressed in mammalian brain neurons. The European Journal of Neuroscience, 10, 1202–1208. Connors, B. W., & Long, M. A. (2004). Electrical synapses in the mammalian brain. Annual Review of Neuroscience, 27, 393–418. Crone, C., Hultborn, H., Kiehn, O., Mazieres, L., & Wigstrom, H. (1988). Maintained changes in motoneuronal excitability by short-lasting synaptic inputs in the decerebrate cat. Journal de Physiologie, 405, 321–343. Deans, M. R., Gibson, J. R., Sellitto, C., Connors, B. W., & Paul, D. L. (2001). Synchronous activity of inhibitory networks in neocortex requires electrical synapses containing connexin36. Neuron, 31, 477–485. Dietz, V., Quintern, J., & Berger, W. (1981). Electrophysiological studies of gait in spasticity and rigidity. Evidence that altered mechanical properties of muscle contribute to hypertonia. Brain, 104, 431–449. Eken, T., Hultborn, H., & Kiehn, O. (1989). Possible functions of transmitter-controlled plateau potentials in alpha motoneurones. Progress in Brain Research, 80, 257–267. Elsen, F. P., Shields, E. J., Roe, M. T., Vandam, R. J., & Kelty, J. D. (2008). Carbenoxolone induced depression of
178 rhythmogenesis in the pre-Botzinger Complex. BMC Neuroscience, 9, 46. Enriquez-Denton, M., Morita, H., Christensen, L. O., Petersen, N., Sinkjaer, T., & Nielsen, J. B. (2002). Interaction between peripheral afferent activity and presynaptic inhibition of ia afferents in the cat. Journal of Neurophysiology, 88, 1664–1674. Faist, M., Mazevet, D., Dietz, V., & Pierrot-Deseilligny, E. (1994). A quantitative assessment of presynaptic inhibition of Ia afferents in spastics. Differences in hemiplegics and paraplegics. Brain, 117(Pt. 6), 1449–1455. Fawcett, J. W. (2006). Overcoming inhibition in the damaged spinal cord. Journal of Neurotrauma, 23, 371–383. Friedman, D., & Strowbridge, B. W. (2003). Both electrical and chemical synapses mediate fast network oscillations in the olfactory bulb. Journal of Neurophysiology, 89, 2601–2610. Fulton, B. P., Miledi, R., & Takahashi, T. (1980). Electrical synapses between motoneurons in the spinal cord of the newborn rat. Proceedings of the Royal Society of London. Series B: Biological Sciences, 208, 115–120. Garcia-Rill, E. (1991). The pedunculopontine nucleus. Progress in Neurobiology, 36, 363–389. Garcia-Rill, E., Heister, D. S., Ye, M., Charlesworth, A., & Hayar, A. (2007). Electrical coupling: Novel mechanism for sleep-wake control. Sleep, 30, 1405–1414. Garcia-Rill, E., & Skinner, R. D. (1988). Modulation of rhythmic function in the posterior midbrain. Neuroscience, 27, 639–654. Garcia-Rill, E., Skinner, R. D., & Fitzgerald, J. A. (1985). Chemical activation of the mesencephalic locomotor region. Brain Research, 330, 43–54. Grey, M. J., Klinge, K., Crone, C., Lorentzen, J., BieringSorensen, F., Ravnborg, M., et al. (2008). Post-activation depression of soleus stretch reflexes in healthy and spastic humans. Experimental Brain Research, 185, 189–197. Heister, D. S., Hayar, A., Charlesworth, A., Yates, C., Zhou, Y. H., & Garcia-Rill, E. (2007). Evidence for electrical coupling in the SubCoeruleus (SubC) nucleus. Journal of Neurophysiology, 97, 3142–3147. Hewitt, A., Barrie, R., Graham, M., Bogus, K., Leiter, J. C., & Erlichman, J. S. (2004). Ventilatory effects of gap junction blockade in the RTN in awake rats. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology, 287, R1407–R1418. Hinckley, C. A., & Ziskind-Conhaim, L. (2006). Electrical coupling between locomotor-related excitatory interneurons in the mammalian spinal cord. The Journal of Neuroscience, 26, 8477–8483. Honma, S., De, S., Li, D., Shuler, C. F., & Turman, J. E. Jr., (2004). Developmental regulation of connexins 26, 32, 36, and 43 in trigeminal neurons. Synapse, 52, 258–271. Hornby, T. G., Kahn, J. H., Wu, M., & Schmit, B. D. (2006). Temporal facilitation of spastic stretch reflexes following
human spinal cord injury. Journal de Physiologie, 571, 593–604. Hornby, T. G., Rymer, W. Z., Benz, E. N., & Schmit, B. D. (2003). Windup of flexion reflexes in chronic human spinal cord injury: A marker for neuronal plateau potentials? Journal of Neurophysiology, 89, 416–426. Hounsgaard, J., Hultborn, H., Jespersen, B., & Kiehn, O. (1988). Bistability of alpha-motoneurones in the decerebrate cat and in the acute spinal cat after intravenous 5-hydroxytryptophan. The Journal of Physiology, 405, 345–367. Hultborn, H. (2003). Changes in neuronal properties and spinal reflexes during development of spasticity following spinal cord lesions and stroke: Studies in animal models and patients. Journal of Rehabilitation Medicine, 41, 46–55. Hurst, D., & Cedrone, N. (2006). Modafinil for drooling in cerebral palsy. Journal of Child Neurology, 21, 112–114. Hurst, D. L., Lajara-Nanson, W. A., & Lance-Fish, M. E. (2006). Walking with modafinil and its use in diplegic cerebral palsy: Retrospective review. Journal of Child Neurology, 21, 294–297. Ishikawa, K., Ott, K., Porter, R. W., & Stuart, D. (1966). Low frequency depression of the H wave in normal and spinal man. Experimental Neurology, 15, 140–156. Ju, M. S., Chen, J. J., Lee, H. M., Lin, T. S., Lin, C. C., & Huang, Y. Z. (2000). Time-course analysis of stretch reflexes in hemiparetic subjects using an on-line spasticity measurement system. Journal of Electromyography and Kinesiology, 10, 1–14. Kiehn, O., & Tresch, M. C. (2002). Gap junctions and motor behavior. Trends in Neurosciences, 25, 108–115. Kiser, T. S., Reese, N. B., Maresh, T., Hearn, S., Yates, C., Skinner, R. D., et al. (2005). Use of a motorized bicycle exercise trainer to normalize frequency-dependent habituation of the H-reflex in spinal cord injury. The Journal of Spinal Cord Medicine, 28, 241–245. Kistler, W. M., De Jeu, M. T., Elgersma, Y., Van Der Giessen, R. S., Hensbroek, R., Luo, C., et al. (2002). Analysis of Cx36 knockout does not support tenet that olivary gap junctions are required for complex spike synchronization and normal motor performance. Annals of the New York Academy of Sciences, 978, 391–404. Kita, M., & Goodkin, D. E. (2000). Drugs used to treat spasticity. Drugs, 59, 487–495. Kitzman, P. (2005). Alteration in axial motoneuronal morphology in the spinal cord injured spastic rat. Experimental Neurology, 192, 100–108. Kuhn, R. A. (1950). Functional capacity of the isolated human spinal cord. Brain, 73, 1–51. Lance, J. W. (1980). Symposium synopsis. In R. G. Feldman, R. R. Young & W. P. Koella (Eds.), Spasticity: Disordered motor control (pp. 485–494). Chicago: Year Book.
179 Landau, W. M., & Clare, M. H. (1964). Fusimotor Function. Vi. H Reflex, Tendon Jerk, and Reinforcement in Hemiplegia. Archives of Neurology, 10, 128–134. Leznik, E., & Llinas, R. (2005). Role of gap junctions in synchronized neuronal oscillations in the inferior olive. Journal of Neurophysiology, 94, 2447–2456. Li, Y., & Bennett, D. J. (2003). Persistent sodium and calcium currents cause plateau potentials in motoneurons of chronic spinal rats. Journal of Neurophysiology, 90, 857–869. Li, W. C., Roberts, A., & Soffe, S. R. (2009). Locomotor rhythm maintenance: Electrical coupling among premotor excitatory interneurons in the brainstem and spinal cord of young Xenopus tadpoles. Journal de Physiologie, 587, 1677–1693. Little, J. W., Ditunno, J. F., Jr., Stiens, S. A., & Harris, R. M. (1999). Incomplete spinal cord injury: Neuronal mechanisms of motor recovery and hyperreflexia. Archives of Physical Medicine and Rehabilitation, 80, 587–599. Little, J. W., & Halar, E. M. (1985). H-reflex changes following spinal cord injury. Archives of Physical Medicine and Rehabilitation, 66, 19–22. Long, M. A., Deans, M. R., Paul, D. L., & Connors, B. W. (2002). Rhythmicity without synchrony in the electrically uncoupled inferior olive. The Journal of Neuroscience, 22, 10898–10905. Magladery, J. W., Teasdall, R. D., Park, A. M., & Languth, H. W. (1952). Electrophysiological studies of reflex activity in patients with lesions of the nervous system. I. A comparison of spinal motoneurone excitability following afferent nerve volleys in normal persons and patients with upper motor neurone lesions. Bulletin of the Johns Hopkins Hospital, 91, 219–244, (passim). Meinck, H. M. (1976). Occurrence of the H reflex and the F wave in the rat. Electroencephalography and Clinical Neurophysiology, 41, 530–533. Milanov, I. (1994). Examination of the segmental pathophysiological mechanisms of spasticity. Electromyography and Clinical Neurophysiology, 34, 73–79. Morita, H., Petersen, N., Christensen, L. O., Sinkjaer, T., & Nielsen, J. (1998). Sensitivity of H-reflexes and stretch reflexes to presynaptic inhibition in humans. Journal of Neurophysiology, 80, 610–620. Nakazawa, K., Kawashima, N., & Akai, M. (2006). Enhanced stretch reflex excitability of the soleus muscle in persons with incomplete rather than complete chronic spinal cord injury. Archives of Physical Medicine and Rehabilitation, 87, 71–75. Nielsen, J., Petersen, N., & Crone, C. (1995). Changes in transmission across synapses of Ia afferents in spastic patients. Brain, 118(Pt. 4), 995–1004. Olsen, P. Z., & Diamantopoulos, E. (1967). Excitability of spinal motor neurones in normal subjects and patients with spasticity, Parkinsonian rigidity, and cerebellar hypotonia.
Journal of Neurology, Neurosurgery and Psychiatry, 30, 325–331. Parisian, K., Wages, P., Smith, A., Jarosz, J., Hewitt, A., Leiter, J. C., et al. (2004). Ventilatory effects of gap junction blockade in the NTS in awake rats. Respiratory Physiology & Neurobiology, 142, 127–143. Pierrot-Deseilligny, E. (1990). Electrophysiological assessment of the spinal mechanisms underlying spasticity. Electroencephalography and Clinical Neurophysiology. Supplement, 41, 64–273. Placantonakis, D. G., Bukovsky, A. A., Zeng, X. H., Kiem, H. P., & Welsh, J. P. (2004). Fundamental role of inferior olive connexin 36 in muscle coherence during tremor. Proceedings of the National Academy of Sciences of the United States of America, 101, 7164–7169. Powers, R. K., & Rymer, W. Z. (1988). Effects of acute dorsal spinal hemisection on motoneuron discharge in the medial gastrocnemius of the decerebrate cat. Journal of Neurophysiology, 59, 1540–1556. Reese, N. B., Skinner, R. D., Mitchell, D., Yates, C., Barnes, C. N., Kiser, T. S., et al. (2006). Restoration of frequency-dependent depression of the H-reflex by passive exercise in spinal rats. Spinal Cord, 44, 28–34. Rekling, J. C., & Feldman, J. L. (1997). Bidirectional electrical coupling between inspiratory motoneurons in the newborn mouse nucleus ambiguus. Journal of Neurophysiology, 78, 3508–3510. Roopun, A. K., Middleton, S. J., Cunningham, M. O., LeBeau, F. E., Bibbig, A., Whittington, M. A., et al. (2006). A beta2-frequency (20–30 Hz) oscillation in nonsynaptic networks of somatosensory cortex. Proceedings of the National Academy of Sciences of the United States of America, 103, 15646–15650. Schindler-Ivens, S., & Shields, R. K. (2000). Low frequency depression of H-reflexes in humans with acute and chronic spinal-cord injury. Experimental Brain Research, 133, 233–241. Schindler-Ivens, S. M., & Shields, R. K. (2004). Soleus Hreflex recruitment is not altered in persons with chronic spinal cord injury. Archives of Physical Medicine and Rehabilitation, 85, 840–847. Schmit, B. D., Benz, E. N., & Rymer, W. Z. (2002). Afferent mechanisms for the reflex response to imposed ankle movement in chronic spinal cord injury. Experimental Brain Research, 145, 40–49. Schmit, B. D., Dhaher, Y., Dewald, J. P., & Rymer, W. Z. (1999). Reflex torque response to movement of the spastic elbow: Theoretical analyses and implications for quantification of spasticity. Annals of Biomedical Engineering, 27, 815–829. Sinfield, J. L., & Collins, D. R. (2006). Induction of synchronous oscillatory activity in the rat lateral amygdala in vitro
180 is dependent on gap junction activity. The European Journal of Neuroscience, 24, 3091–3095. Sinkjaer, T., Toft, E., Larsen, K., Andreassen, S., & Hansen, H. J. (1993). Non-reflex and reflex mediated ankle joint stiffness in multiple sclerosis patients with spasticity. Muscle & Nerve, 16, 69–76. Skinner, R. D., Houle, J. D., Reese, N. B., Berry, C. L., & Garcia-Rill, E. (1996). Effects of exercise and fetal spinal cord implants on the H-reflex in chronically spinalized adult rats. Brain Research, 729, 127–131. Soderberg, G. L., & Knutson, L. M. (2000). A guide for use and interpretation of kinesiologic electromyographic data. Physical Therapy, 80, 485–498. Stelzner, D. J., Ershler, W. B., & Weber, E. D. (1975). Effects of spinal transection in neonatal and weanling rats: Survival of function. Experimental Neurology, 46, 156–177. Thompson, F. J., Browd, C. R., Carvalho, P. M., & Hsiao, J. (1996). Velocity-dependent ankle torque in the normal rat. NeuroReport, 7, 2273–2276. Thompson, F. J., Parmer, R., Reier, P. J., Wang, D. C., & Bose, P. (2001). Scientific basis of spasticity: Insights from a laboratory model. Journal of Child Neurology, 16, 2–9. Thompson, F. J., Reier, P. J., Lucas, C. C., & Parmer, R. (1992). Altered patterns of reflex excitability subsequent to contusion injury of the rat spinal cord. Journal of Neurophysiology, 68, 1473–1486. Traub, R. D., Pais, I., Bibbig, A., LeBeau, F. E., Buhl, E. H., Hormuzdi, S. G., et al. (2003). Contrasting roles of axonal (pyramidal cell) and dendritic (interneuron) electrical coupling in the generation of neuronal network oscillations. Proceedings of the National Academy of Sciences of the United States of America, 100, 1370–1374. Tresch, M. C., & Kiehn, O. (2000). Motor coordination without action potentials in the mammalian spinal cord. Nature Neuroscience, 3, 593–599.
Urbano, F. J., Leznik, E., & Llinas, R. R. (2007). Modafinil enhances thalamocortical activity by increasing neuronal electrotonic coupling. Proceedings of the National Academy of Sciences of the United States of America, 104, 12554–12559. Verdier, D., Lund, J. P., & Kolta, A. (2004). Synaptic inputs to trigeminal primary afferent neurons cause firing and modulate intrinsic oscillatory activity. Journal of Neurophysiology, 92, 2444–2455. Walton, K. D., & Navarrete, R. (1991). Postnatal changes in motoneurone electrotonic coupling studied in the in vitro rat lumbar spinal cord. Journal de Physiologie, 433, 283–305. Wenner, P., & O'Donovan, M. J. (2001). Mechanisms that initiate spontaneous network activity in the developing chick spinal cord. Journal of Neurophysiology, 86, 1481–1498. Woods, A. J., Mennemeier, M., Garcia-Rill, E., Meythaler, J., Mark, V. W., Jewel, G. R., et al. (2006). Bias in magnitude estimation following left hemisphere injury. Neuropsychologia, 44, 1406–1412. Yablon, S. A., & Stokic, D. S. (2004). Neurophysiologic evaluation of spastic hypertonia: Implications for management of the patient with the intrathecal baclofen pump. American Journal of Physical Medicine & Rehabilitation, 83, S10–S18. Yates, C., Charlesworth, A., Allen, S. R., Reese, N. B., Skinner, R. D., & Garcia-Rill, E. (2008). The onset of hyperreflexia in the rat following complete spinal cord transection. Spinal Cord, 46, 798–803. Yates, C. C., Charlesworth, A., Reese, N. B., Ishida, K., Skinner, R. D., & Garcia-Rill, E. (2009). Modafinil normalized hyperreflexia after spinal transection in adult rats. Spinal Cord, 47, 481–485. Yates, C. C., Charlesworth, A., Reese, N. B., Skinner, R. D., & Garcia-Rill, E. (2008). The effects of passive exercise therapy initiated prior to or after the development of hyperreflexia following spinal transection. Experimental Neurology, 213, 405–409.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 12
Modulation of rhythmic movement: Control of coordination Larry M. Jordan{,* and Urszula Sławi nska{ {
Department of Physiology, Spinal Cord Research Centre, University of Manitoba, Winnipeg MB, Canada { Laboratory of Neuromuscular Plasticity, Department of Neurophysiology, Nencki Institute of Experimental Biology PAS, Warsaw, Poland
Abstract: Three rhythmic movements, breathing, walking, and chewing, are considered from the perspective of the emerging factors that control their coordination. This takes us beyond the concept of a core excitatory kernel and into the common principles that govern the interaction between components of the neural networks that must be orchestrated properly to produce meaningful movement beyond the production of the basic rhythm. We focus on the role of neuromodulators, especially 5-hydroxytryptamine (5-HT), in the production of coordinated breathing, walking, and chewing, and we review the evidence that at least in the case of breathing and walking, 5-HT input to the CPGs acts through the selection of inhibitory interneurons that are essential for coordination. We review data from recently developed mouse models that offer insight into the contributions of inhibitory coordinating neurons, including the development of a new model that has allowed the revelation that there are glycinergic pacemaker neurons that likely contribute to the production of the respiratory rhythm. Perhaps walking and chewing will not be far behind. Keywords: respiration; mastication; locomotion; spinal cord injury; serotonine; glycine; GABA.
This statement was inspired by an emerging appreciation of the multiple rhythmogenic mechanisms that are present in the neural systems underlying respiration, locomotion, and mastication, and the wealth of processes whereby they can be initiated, modified, and controlled. The papers in this section incorporate initiation/ induction of the three types of rhythmic movement, sensory and supraspinal modulation,
Every breath is a new breath,1 every step is a new step, every bite is a new bite (S. Rossignol, Montreal, May 5, 2009) *Corresponding author. Tel.: þ1-204-789-3534; Fax: þ1-204-789-3934 1
With apologies to Nino Ramirez.
DOI: 10.1016/B978-0-444-53825-3.00017-6
181
182
neuromodulation of channels and other cellular properties, coordination of activity in functionally distinct groups of neurons, developmental changes in and modulation of synaptic transmission, cortical control of orofacial movements, and cortical control of precision locomotion, to name a few. Many of these features of the modulation of rhythmic movement are implicated in pathological conditions, as well, and a few have been chosen for inclusion into this volume. There are many common themes that emerge when one considers these three rhythmic motor systems. A major one is the issue of coordination of activity among various components of the networks controlling these behaviors. This leads immediately to the roles of inhibitory interneurons and transmitters in coordination. When the contribution of inhibitory interneurons to these behaviors in mammals is considered, it becomes apparent that these coordinating inhibitory interneurons are under potent neuromodulatory control. All three of the scientists being honored in this volume have made major contributions to the understanding of coordination of the three motor patterns under consideration, and it seems appropriate to attempt to draw together the concepts that have emerged from their work and the work of others who have addressed this issue. Two major features that emerge when the question of coordination is considered are: (1) the participation of inhibitory interneurons in the CPGs of these three motor systems, and (2) the importance of neuromodulation, especially the role of serotonin or 5-hydroxytryptamine (5-HT), in the production of coordinated activity in each of these systems. We discuss the contributions of 5-HT to the control of breathing, walking, and chewing and attempt to point out the many common neuromodulatory processes that these rhythmic systems share. Then we discuss the essential contributions of inhibitory interneurons in the control of these behaviors and their control by 5-HT.
5-HT control of respiration In the case of breathing, the context of the rhythmic behavior dynamically alters the underlying processes for rhythm generation—contexts such as exercise, mastication, swallowing, speech, and several others—so that modes of coordinating inspiration and expiration exist that are state-dependent and mutable. Thus a hierarchical arrangement of neural circuits in the brainstem has evolved that can be re-organized to produce a complex array of motor behaviors (Smith et al., 2009). A prevailing hypothesis guiding research in the control of respiration (Janczewski and Feldman, 2006) is that “. . .inspiration and expiration are generated by coupled, anatomically separate rhythm generators, one generating active expiration located close to the facial nucleus in the region of the retrotrapezoid nucleus/parafacial respiratory group, the other generating inspiration located more caudally in the pre-Bötzinger Complex (PBC).” So inspiration and expiration are separately controlled. . .. But how are they normally coordinated? How can the patterns of activity that account for eupnea, sighing and gasping be achieved? What goes wrong in sudden infant death syndrome (SIDS) or sleep apnea? Breathing is produced by a respiratory rhythmgenerating network located in the brainstem that produces a synchronous bilateral drive onto cranial and spinal populations of motoneurons that innervate cranial, thoracic, and abdominal respiratory muscles (Feldman and Del Negro, 2006). Recent data shows (Bouvier et al., 2010) that Dbx1-derived interneurons, thought to be the core rhythmogenic elements of the pre-BötC oscillator, require Robo3-dependent guidance signaling for development of bilaterally synchronous activity. In the absence of commissural interconnections of Dbx1-derived interneurons, the two sides of the diaphragm contract independently and ineffectively. This control of coordinating interconnections in the respiratory circuit is analogous to the role of commissural Dbx1-derived interneurons (Lanuza et al., 2004) and the interneurons that activate them (Crone
183
et al., 2008) in the spinal cord for the control of the two hindlimbs. This is just one of the many striking similarities between respiration and locomotion that have recently become apparent. All three of the rhythmic systems under discussion are subject to powerful control from the brainstem 5-HT system. The importance of 5-HT in respiratory control, including recent results using transgenic mice, has been recently reviewed (Hodges and Richerson, 2010). Defects in the 5-HT system are implicated in SIDS (Feldman et al., 2003) due to the chemosensitivity (CO2 responses) of 5-HT neurons in the medulla, as well as to serotonin-dependent developmental plasticity. Reconfiguration of the respiratory network to produce eupnea, sighs, or gasps can be achieved by controlling glycine inhibition (Lieske et al., 2000). Such a dependence of 5-HT action on controlling glycine inhibition will be discussed below for other rhythms as well. 5-HT is implicated in pacemaker activity, and fictive gasping is eliminated by 5-HT2A receptor antagonists, suggesting that endogenous 5-HT2A receptor activation is essential for fictive gasping rhythm generation in vitro (Tryba et al., 2006). 5-HT neurons excite key circuit components required for generation of respiratory motor output, such that raphe obscurus activation increases motor output by exciting both pre-BötC and motor neurons through an action on a nonselective cation leak current that produces neuronal depolarization. In addition, 5-HT transforms some pre-BötC neurons into bursting neurons (Ptak et al., 2009). These effects were blocked by the 5-HT2 antagonist ketanserin. More recently it has been demonstrated that deficiencies in neuromodulators acting on neurokinin1 (NK1), a1 noradrenergic, and 5-HT2 serotonin receptors within the preBötzinger complex are immediately compensated by the action of other neuromodulators to accommodate state-dependent needs for respiratory control (Doi and Ramirez, 2010). Serotonergic control of inhibitory transmission in the PBC has been demonstrated in the adult
cat (Pierrefiche et al., 1998). Bilateral blockade of both GABAergic and glycinergic inhibition abolished rhythmic burst discharges and only tonic phrenic activity remained. Potentiation of synaptic inhibition by 8-OH-DPAT, a 5-HT1A/7 agonist restored rhythmic activity when given shortly after strychnine and bicuculline applications. An examination of reciprocal synaptic inhibition in the respiratory network (Manzke et al., 2009) showed that 5-HT potentiates glycinergic inhibition of both excitatory and inhibitory neurons, and glycinergic PBC neurons possess 5-HT7 and 5-HT1A receptors. Nevertheless, the respiratory rhythm persists in animals deficient in glycine receptors (Busselberg et al., 2001).
5-HT and locomotion It has been more than a decade since a comprehensive review (Schmidt and Jordan, 2000) of the contribution of 5-HT to the control of locomotion appeared. Since that time, there has been a great deal of progress in deciphering the role of 5-HT is the control of all through of the rhythmic motor tasks under consideration here. In fact, it should be recognized that 5-HT and its control of coordination in these behaviors has become a central theme in each of them. We have seen the progress in understanding the role of 5-HT in the control of respiration above. In a recent review, Grillner and Jessell (2009) exclude the “neuromodulators” (including norepinephrine, dopamine, 5-HT, mGluR1, GABAB, and others) from the “core features of the locomotor network,” and describe them as providing “more refined aspects of circuit function and motor output” rather than being integral to the function of the network. This view needs revision to account for the various roles of “neuromodulators” in the networks for mammalian locomotion, breathing, and mastication. As pointed out by Ramirez (Chapter 3), modulatory processes are critical for the neuronal control of mammalian breathing that goes beyond providing “more refined aspects
184
of circuit function and motor output”. With respect to 5-HT in particular, there is good evidence that this neurotransmitter provides an essential excitatory drive to the respiratory network (Hodges and Richerson, 2010; Hodges et al., 2009). Similar data is available for the initiation and control of locomotion. Schmidt and coworkers demonstrated that 5-HT is the most effective agent for production of well-coordinated locomotion in the isolated neonatal rat spinal cord (see Schmidt and Jordan, 2000, for review). Jacobs and Fornal (1993) concluded from work of chronic recordings from 5-HT neurons in behaving animals that 5-HT neuron activity is correlated with locomotor activity. Subsequent work has demonstrated that stimulation of a discrete site in the rat brainstem that is rich in descending 5-HT neurons gives rise to coordinated locomotion (Liu and Jordan, 2005). The effects of stimulation of these 5-HT neurons, located within the parapyramidal region of the medulla, can be abolished by spinal application of 5-HT2 and 5-HT7 antagonists. In mice lacking 5-HT7 receptors, applications of 5-HT produce rhythmic activity, but inter- and intralimb coordination is disrupted (Liu et al., 2009). As adults these mice have locomotor deficits consistent with the conclusion that the 5-HT descending system, acting through 5-HT7 receptors, is critical for well-coordinated locomotion. Noga et al. (2009) have demonstrated that neurons that express the activity-dependent marker c-fos during locomotion have close appositions with 5-HT fibers and possess 5-HT receptors. 5-HT7, 5-HT2A, or 5-HT1A receptor subtypes have been implicated in the production of locomotion (Antri et al., 2003; Beato and Nistri, 1998; Hochman et al., 2001; Landry et al., 2006; Liu and Jordan, 2005; Liu et al., 2009; Madriaga et al., 2004; Pearlstein et al., 2005; Ung et al., 2008). In chronic spinal rats, a major deficit that can be readily detected from EMG recordings is the absence of inter- and intralimb coordination. Spinal cord injury in rodents is associated with altered coordination such that proper movement
at all joints and placement of the foot (plantar stepping) are impaired. The involvement of inhibitory interneurons of the central pattern generator for locomotion seems obvious from the description above of their contribution to the control of inter- and intralimb coordination. This change in coordination can be reversed by treatments that increase the available 5-HT in the spinal cord (Majczynski et al., 2005; Slawinska et al., 2000). Pharmacological blockade or knockout of specific 5-HT receptors impairs coordination during locomotion (Jordan et al., 2008; Liu et al., 2009). As is the case for the spinal interneurons participating in the control of locomotion (Goulding, 2009; Goulding and Pfaff, 2005; Kiehn and Butt, 2003), new knowledge about the development of brainstem 5-HT neurons is providing novel approaches to the study of the importance of 5-HT is the control of a variety of physiological functions, including respiration and locomotion (Jensen et al., 2008; Wylie et al., 2010). It has been suggested (van Doorninck et al., 1999) that absence of GATA-3, which is required the development of the 5-HT neurons of the brainstem that descend to the spinal cord, leads to locomotor defects. In LIM homeobox transcription factor 1b conditional knock-out mice (Lmx1bf/f/p) in which Lmx1b was deleted in Pet-1 (pheochromocytoma 12 ETS factor-1)-expressing 5-HT neurons, 5-HT neurons of the brainstem raphe system were completely absent (Zhao et al., 2006). These knock-out mice exhibited overtly normal locomotor activity based upon the tests used, but the outcome measures used in this case was performance on an accelerating rotarod treadmill and total ambulations in an open-field test. Neither of these tests provides an assessment of locomotor coordination, a variable of locomotion controlled by the 5-HT system. In fact, there is evidence that the neurons activated in the rotarod test differ from those active during activation of the locomotor CPG (Dai et al., 2005). A detailed EMG and kinematic analysis of these animals might provide more definitive results.
185
In another study, Pet-1 null mice, which also lack 5-HT neurons (Hendricks et al., 2003), exhibited significantly shorter durations on the rotarod in early trials as compared to wild-type littermates. Pet-1 null mice also showed a lower level of total activity in an open-field test, but the difference was not significant. Again, analysis of coordination during locomotion would have been informative. It is also important to remember that the significant alterations produced in knockout animals often occur in the developing animals, prior to the completion of adaptive processes that allow for development of what Serge Rossignol might refer to as “new steps.” In lamprey swimming, unlike in mammals, 5-HT has no effects on the in vitro swimming rhythm on its own, but modulates effects induced by NMDA (Harris-Warrick and Cohen, 1985). Grillner and coworkers attributed this to 5-HT1A receptors, and showed that the effects were due to decreased AHP due to an action of the Ca dependent K channel. It increased firing, decreased AHP, and lowered the burst frequency. In mammals, the role of 5-HT is much more complex and much more important in the control of the locomotor rhythm. When the role of 5-HT in locomotion was last reviewed in detail (Schmidt and Jordan, 2000), the actions of 5-HT that could impact the control of locomotion that were recognized included motoneuron depolarization and EPSPs from raphe stimulation, primary afferent depolarization, increases in step length, increases in duration and amplitude of hindlimb flexor and extensor EMG activity, increases in the excursions of the hip, knee, and ankle joints, and induction of locomotion in rodent spinal cord in vitro. Release of 5-HT in the spinal cord upon stimulation of brainstem locomotor areas had also been demonstrated. Normal development of the spinal 5-HT system had been shown to be critical for proper operation of the locomotor network during the first 2 weeks of life. 5-HT was also known to induce plateau properties, reduce AHPs, and reduced glycinergic inhibition during swimming in Xenopus tadpoles. Since that time, we have learned that a major command pathway
for the initiation of locomotion originates in a particular group of medullary 5-HT neurons (Jordan et al., 2008; Liu and Jordan, 2005). It is now known that 5-HT produces oscillatory behavior in interneurons that are involved in the control of locomotion (Carlin et al., 2006; Zhong et al., 2006; Dai et al., 2009), and that 5-HT modulates a number of membrane properties that contribute to the control of rhythmicity, particularly in neurons that are active during locomotion. These include modulation of Ih (Dai and Jordan, 2010b; Dai et al., 2009), lowering of voltage threshold in motoneurons as well as in locomotor interneurons (Dai et al., 2009; Gilmore and Fedirchuk, 2004), an increase in Rin, and a reduction of the AHP. 5HT enhances persistent inward currents (PIC) in locomotor interneurons (Dai and Jordan, 2010a; Dai et al., 2009). In locomotor interneurons it does so by hyperpolarizing PIC onset and increasing PIC amplitude. The 5-HT-increased PIC amplitude seemed to be significantly larger in lamina VII neurons. According to Gossard (Chapter 2), rhythmic pattern generators must (by definition) control cycle period, phase durations, and phase transitions. He makes a case for separate generators for locomotion and scratch based upon this tenant. According to this view, the predominant relationship between cycle duration and burst duration of antagonist motoneurons groups is defined by the network. Frigon (Chapter 7) discusses the implications of inter-subject variability for locomotor adaptation, and points out that this is particularly relevant to recovery of locomotor capability after injury. After spinal cord injury in rats the usual form of coordination that is a signature for functional locomotor output emphasized by Gossard, with extensor burst duration increasing with step cycle duration, can be altered so that the duration of activity in ankle extensor muscles is not sufficient to provide for a functional stance phase. This relationship can be shifted to a more functional pattern of coordination by activation of 5-HT2 receptors with quipazine (Fig. 1). A “flexor-interval,” the silent interval between
186 Pre-drug 1000 Burst duration [ms]
L Sol L TA R Sol R TA
L Sol L TA R Sol R TA
800 600 400 200 0 0
1s
200 400 600 800 1000 1200 1400
Cycle duration [ms]
After quipazine 1000 Burst duration [ms]
L Sol L TA R Sol R TA
800 600 400 200 0
1s
0
200 400 600 800 1000 1200 1400
Cycle duration [ms]
Fig. 1. Intra- and interlimb coordination is improved in spinal rats after activation of 5-HT2 receptors. Recordings were taken from bipolar EMG electrodes chronically implanted into the left (L) and right (R) soleus (Sol) and tibialis anterior (TA) muscles. Several weeks after complete spinal cord transection, with no prior training, the rat was placed with its forepaws on a platform and its hindlimbs over a moving treadmill belt. Manual stimulation of the tail (as in Slawinska et al., 2000) induced rapid, un-coordinated movements of the hindlimbs, often with little sustained rhythmicity. Here we selected an episode of repetitive activity that was sustained over a sufficient period of time to allow analysis of burst duration as a function of cycle duration. As the burst/cycle plots for each muscle illustrate (top panel), there was little or no change in burst duration with increasing cycle duration, in contrast to the normal situation in intact animals. The durations of the bursts of activity in both the flexor and extensor muscles were uniformly short, with a marked “flexor-interval” (Chapter 2), so that the duration of the extensor burst was insufficient for a proper stance phase. The animal received an IP injection of 0.25 mg/kg quipazine, a 5-HT2 receptor agonist, and 30 min later was again tested on the treadmill (lower panels). The “flexor-interval” disappeared, so that the extensor burst was of normal duration, occupying the entire period between flexor bursts, and the normal relationship between burst and cycle duration for extensor muscles was restored (burst duration/cycle duration plots). During the quipazine trial the animal displayed good plantar stepping with weight support.
two flexor bursts (see Gossard, Chapter 2) appears after spinal cord injury, with the extensor burst duration shortened so that it no longer occupies the entire period between flexor bursts. The normal prolonged extensor activity is required for a functional stance phase of the locomotor cycle. Its absence is associated with poor weight support and the absence of effective plantar walking. The latter is a major hallmark of
disturbed locomotion after spinal cord injury (Basso et al., 1995). The examples shown in Fig. 1 are consistent with the view that after spinal cord injury the CPG for locomotion can still be induced to produce rhythmic activity, but the movements are fast and uncoordinated, with extensor bursts that are too brief to effectively support the body throughout the stance phase. After activation of
187
5-HT2 receptors with quipazine, the rhythm is slowed, coordination is dramatically improved, and the duration of extensor bursts is increased to allow for effective support during stance. We interpret these changes as the result of recruitment of coordinating inhibitory interneurons.
5-HT and chewing The late Jim Lund and colleagues were instrumental in showing that 5-HT is an important contributor to the control of mastication (Kolta et al., 1993). 5-HT neurons are known to be active during mastication (Ribeiro-do-Valle, 1997). Trigeminal motoneurons receive 5-HT terminals, and stimulation of 5-HT neurons in the raphe produces EPSPs. The 5-HT innervation and raphe effects on jaw closer muscles are greater than on jaw opener muscles (Nagase et al., 1997), and 5HT facilitates trigeminal rhythmic activity (Mori et al., 2002). 5-HT facilitates rhythmic jaw movements and increases bruxism (reviewed in Lavigne et al., 2003) and induces bistable membrane properties in trigeminal motoneurons (Hsiao et al., 1998). Further investigation into the role of 5-HT in the control of chewing seems warranted.
Inhibitory interneurons in the control of rhythmic movement The normal inter- and intralimb coordination that occurs during locomotion can be considered to be due to the activity of several classes of inhibitory interneurons. Figure 2 illustrates some of those that might account for intralimb coordination, providing for the alternating activity between flexor and extensor motoneurons groups. The diagrams are based on the Rybak-McCrea CPG model (McCrea and Rybak, 2008) in which a two-level CPG has a common rhythm generator (RG) that controls the operation of the pattern formation (PF) circuitry responsible for
motoneuron activation. Similar groups of reciprocally organized inhibitory interneurons are proposed at the RG, PF, and motoneuron (Mn) levels. An identical system would control the activity in the two limbs such that when the entire system is active, well-coordinated locomotion occurs. Such a case, produced in vitro by application of 5-HT, is illustrated for left and right flexor and extensor nerves (Fig. 2C). Also illustrated is the locomotor pattern produced in a spinal rat with quipazine (Fig. 2D), and the alternating excitation and inhibition observed in individual motoneurons during fictive locomotion (Fig. 2E). After blockage of glycinergic transmission with strychnine, there is loss of the hyperpolarized phase of the locomotor drive potential recorded from motoneurons (Fig. 2H) (Pratt and Jordan, 1987). Also the rhythmic activity recorded in chronic spinal rats becomes uncoordinated, with co-contractions of ipsilateral flexors and extensors and an uncoupling of the EMG discharges on the two sides (Fig. 2G). This loss of coordination eventually leads to synchronous activity among all the recorded EMGs, as also occurs in the neonatal rat (Cowley and Schmidt, 1995) after strychnine administration (Fig. 2F). We propose that the central pattern generator for locomotion can be described as an excitatory kernel for producing rhythmic activity, plus an inhibitory component that is responsible for intraand interlimb coordination (see Cowley and Schmidt, 1995). This inhibitory component can be inactive (e.g., in the presence of strychnine) while the excitatory component continues its rhythmic bursting, leading to synchronous activity throughout the motoneuron pools of the spinal cord (Fig. 2). The inhibitory interneurons can be controlled separately from the excitatory kernel by different spinal and descending systems, and they appear to be more dependent upon the descending 5-HT system for their activity than the excitatory component. This explains the observation that uncoordinated locomotion occurs in chronic spinal rats, and proper interlimb coordination can be restored by 5-HT agonists (Antri
188 (a)
RG-E
RG-F
(c)
5-HT
R Tib
Rhythm Generator
R Per L Tib
RGI-E
L Per
RGI-F
PFE-E
5s
PFE-F
(d)
Pattern Formation
Quipazine
L Sol PFI-E
L TA
PFI-F
R Sol R TA Mn I-E
Mn I-F
Mn-E
(e)
Mn-F
1s MLR evoked locomotion
Motoneurons
20 mV Extensor
Flexor
1s
STRYCHNINE glycine receptor antagonist
(b) RG-E
RG-F
(f)
Strychnine
R Tib Rhythm Generator
R Per L Tib L Per RGI-E
RGI-F
PFE-E
Pattern Formation
25 s
PFE-F
(g)
Quipazine + strychnine
L Sol PFI-E
L TA
PFI-F
R Sol R TA Mn I-E
Mn I-F
1s Mn-E
Mn-F
(h)
MLR evoked locomotion + strychnine
Motoneurons
20 mV Extensor
Flexor
1s
Fig. 2. Suppression of activity in inhibitory interneurons interferes with intralimb coordination that provides for the alternating activity between flexor and extensor motoneurons groups. These diagrams are based on the Rybak-McCrea CPG model (McCrea and Rybak, 2008) in which a two-level CPG has a common rhythm generator (RG) controlling a pattern formation (PF) layer with
189
et al., 2002; Courtine et al., 2009; Fig. 1) or by transplanting 5-HT neurons into the spinal cord below the lesion (Majczynski et al., 2005; Slawinska et al., 2000). Coordination is known to be under the control of inhibitory interneuron groups of species from primitive chordates to mammals. In the ascidian urocordate Ciona intestinalis glycinergic inhibitory interneurons have a key role in coordinating swimming movements (Nishino et al., 2010). In a reduced preparation with the motor ganglion and the tail intact, glutamate induces swimming movements with left–right phasic activity identical to the intact larvae. In the presence of the glycine antagonist strychnine there was a loss of the strict alternation between the left and right sides. When the expression of the GlyR receptor was suppressed with antisense oliogonucleotides injected into fertilized eggs, the resulting embryos were unable to swim progressively due to disrupted coordination identical to that produced by strychnine (no organized left–right alternation). Blocking the effects of GABA with picrotoxin, however, did not alter left–right coordination; instead, it increased the frequency of swimming. In zebrafish at embryonic stages, contralateral alternation of motor activity is disrupted by strychnine, demonstrating a role for glycine in coordinating left–right alternation during NMDA-induced locomotion (McDearmid and Drapeau, 2006). In adult zebrafish (Gabriel
et al., 2008), blockade of glycinergic synaptic transmission with strychnine produced synchronous bursting in the left and right sides as well as along the rostrocaudal axis, consistent with a requirement of glycinergic interneurons in the coordination of swimming in this species. In juvenile and adult zebrafish, 5-HT increases mid-cycle inhibition, facilitates the action of inhibitory interneurons, and delays the onset of the subsequent depolarization (Gabriel et al., 2009). This is consistent with an excitatory action of 5-HT on coordinating inhibitory interneurons. The situation in limbed vertebrates is similar to aquatic forms for left–right coordination with respect to the requirement for glycinergic inhibitory interneurons, but now the coordination of flexor and extensor motoneurons groups of the limbs is required for locomotor activity to give rise to progression. It is known from developmental studies and experiments on the isolated neonatal rat spinal cord that blockade of glycinergic synapses with strychnine abolishes inter- and intralimb coordination (Bracci et al., 1996; Cowley and Schmidt, 1995; Kremer and LevTov, 1997; Kudo et al., 2004; Ozaki et al., 1996). In one report (Hinckley et al., 2005) it was proposed that glycine and GABAA receptors have distinct roles in coordinating locomotor activity, but this is in contrast to previous findings (Cowley and Schmidt, 1995) that in neonates they have similar roles, such that synchronous activity is produced by blocking glycine or GABAA receptors.
circuitry responsible for motoneuron activation. The rhythm generator level consists of positive feedback-coupled excitatory RG-E and RG-F neuron populations projecting to the pattern formation layer (PFE-E and PFE-F). The PFE-E neurons excite extensor motoneurons, and the PFE-F neurons excite flexor motoneurons. Reciprocal inhibition is provided by separate groups of inhibitory interneurons as specified by the McCrea and Rybak model. These are RGI-E and RGI-F neurons for the RG layer, PFI-E and PFIF neurons for the pattern formation layer, and MnI-E and MnI-F for the motoneuron level. Well-coordinated intra- and interlimb coordination is achieved when the system is intact (top traces, C and F), as in the case of fictive locomotion in vitro produced by 5HT and quipazine induced locomotion in an adult spinal rat (D and G). Alternating excitatory and inhibitory inputs make up the locomotor drive potential (LDP) recorded intracellularly in adult motoneurons during fictive locomotion (E). After administration of the glycine receptor antagonist strychnine, the inhibitory phase of the LDP is lost (H). Coordination of left and right sides and flexor and extensor muscles becomes disrupted when the PFI-E and the PFI-F inhibitory interneurons are affected but the RG layer is still producing the basic rhythm (G). When reciprocal inhibition is completely blocked, output slows dramatically and becomes synchronous (F).
190
Near birth glycine receptors begin to predominate in the control of motoneuron output and as well as locomotion (Gao et al., 2001; Nakayama et al., 2002). In adult animals, a role in coordination is reserved for glycine receptors (Jordan and coworkers, unpublished). During MLR-evoked fictive locomotion in adult decerebrate cats the GABAA antagonist bicuculline (0.1 mM IT) increased burst frequency and ENG amplitude. At higher concentrations (1–20 mM IT) spontaneous ENG activity appeared, consisting of alternating flexor and extensor locomotor-like bursts, and brief periods of high frequency alternations (3–8 Hz) resembling scratch or pawshake rhythms. Thus, bicuculline uncovered spontaneous rhythms that are normally suppressed by GABAergic inhibition, including locomotion and other rhythms, but it was without effect on alternating activity between the two hindlimbs, and it did not change coordination among flexor and extensor motoneuron groups. Strychnine, on the other hand, did not induce spontaneous rhythmic activity, but altered the coordination among hindlimb nerves such that there was overlap between nerves normally reciprocally active. After a few minutes, this uncoordinated activity was replaced by synchronous discharge of flexor and extensor nerves on both sides of the spinal cord. Strychnine infusion also enhanced the amplitude of short-latency excitation produced in the ENG recordings by MLR stimuli. Similarly, strychnine increased the amplitude of MLR- evoked positive (P2 and/or P3) waves recorded from the surface of the spinal cord. Strychnine is known to facilitate locomotion in spinal animals (Hart, 1971), and restoration of locomotion in spinal animals by training has been associated with reduced glycinergic inhibition in the spinal cord (de Leon et al., 1999; Edgerton et al., 2008; Tillakaratne et al., 2002). Reducing glycinergic inhibition in the spinal cord has no effect on stepping in spinal cats when they are trained to step but improves stepping, including normalizing intralimb coordination, after they
are retrained to stand (de Leon et al., 1999). So how is it possible that spinal locomotion can be improved by reducing inhibition if intralimb coordination depends on activity in the inhibitory interneurons of the CPG? How can strychnine both facilitate locomotion and also produce a loss of coordinated flexor and extensor activity? Very clearly peripheral afferents have a powerful action on the inhibitory components of the CPG. For example, it has been shown that hindpaw denervation leads to coactivation of flexors and extensors at the ankle (Bouyer and Rossignol, 2003), but this does not persist during fictive locomotion. As pointed out by Viemari et al. (Chapter 1) the functions of the classical “inhibitory” amino acids are developmentally regulated due to changes in the transporters KCC2 and NKCC1, leading to a switch from high [Cl]i and a depolarizing action to low [Cl]i and a hyperpolarizing action. This switch can account for some of the changes in the responses to inhibitory amino acids and their antagonists during development of respiration (Ren and Greer, 2006) and locomotion (Nakayama et al., 2002). It has been suggested that changes in the expression of the KCC2 and/or the NKCC1 transporters in pathological conditions such as spinal cord injury can account for abnormal excitability (Boulenguez et al., 2010) or for neuropathic pain (Hasbargen et al., 2010). New genetic models continue to emerge, including several affecting the glycine receptor or its transporter (Harvey et al., 2008). In the absence of functional glycine receptors, remarkable phenotypes have been revealed, including oscillator and spasmotic. These and other similar mutants serve not only as models for investigation of the role of inhibitory processes in the control of movement, but they also serve as models for the human condition hyperekplexia, which is characterized by hypertonic motor disorders, including neonatal stiffness and an exaggerated startle reflex. Symptoms include a hesitant gait and nocturnal myoclonus (Gregory et al., 2008).
191
Oscillator mutants develop a strong tremor leading to death, a phenotype reminiscent of strychnine poisoning. Inhibitory interactions between the oscillators responsible for the respiratory rhythm is responsible for generating the basic motor patterns for breathing. Both glycine and GABAA receptors control neuronal excitability in the pre-Bötzinger complex (PBC), while only glycinergic transmission is responsible for reciprocal inhibition (Shao and Feldman, 1997). Coordination between the PBC and the parafacial respiratory group (pFRG) depends upon glycinergic and GABAergic inhibition. Blockage of these inhibitory amino acids synchronize PBC and pFRG activity and increase PBC burst frequency, amplitude and duration (Funke et al., 2008). Further review of the role of inhibition in respiration can be found in a recent paper by Feldman and coworkers (Morgado-Valle et al., 2010). They have demonstrated the presence of glycinergic pacemaker neurons in the PBC in slice preparations from transgenic mice in which neurons expressing GlyT2 coexpress enhanced green fluorescent protein (EGFP). This raises the intriguing possibility that the coupled inhibitory interneurons in a rhythmic network can function independently, without the need for drive from excitatory CPG neurons. Perhaps this finding can explain the startling persistence of rhythmic locomotor activity after VGLUT2 inactivation, even though the respiratory rhythm is disrupted in these animals (Wallen-Mackenzie et al., 2006). The suggestion that deletion of the glutamatergic signaling between excitatory interneurons (the core excitatory kernel for locomotion) does not perturb the spinal locomotor pattern (Gezelius et al., 2006; Wallen-Mackenzie et al., 2006) has generated a vigorous response in defense of the classical view. Explanations offered include redundancy in the VGLUT family of transporters, increased reliance on cholinergic transmission by these same excitatory neurons, and a quasi-normal pattern of motor activity
dependent upon direct excitation of motoneurons that is coordinated by inhibitory interneurons. It seems inescapable, however, that glutamatergic signaling is not required for intra- and interlimb coordination in these animals. It is important to recall that it is interactions among coupled inhibitory interneurons that produce coordinated rhythmic activity. Based upon the recent finding that glycinergic neurons in the respiratory system can be pacemakers, it is plausible that the explanation for persistent locomotor coordination after reduction in glutamate signaling is independent operation of the inhibitory networks responsible for coordination. The late Jim Lund has provided us with still another possible explanation for the persistence of the locomotor rhythm in VGLUT2 knockout mice (Kolta et al., 1995; Verdier et al., 2003). As reviewed in detail by Viemari et al. (Chapter 1), there is evidence that primary afferent depolarization (PAD) and consequent discharges in afferent terminals form part of the motor network for mastication. In this scenario, primary afferent terminals can function as excitatory interneurons. Based upon the superb work of Rossignol and colleagues, we now know that PAD during locomotion leads to antidromic discharges in primary afferents during locomotion. What happens at the terminals of these afferents? One of us (LJ) was involved in a lunchtime conversation with Jim Lund and Serge Rossignol at the University of Montreal, when it was decided that it seems imminently logical to presume that the terminals of the antidromically active primary afferent fibers release excitatory amino acids and serve as excitatory interneurons. As it turns out, we know that in this system VGLUT1, not VGLUT2, serves as the glutamate transporter, and so this system would be spared in the VGLUT2 knockout mice, thus leaving intact a separate set of excitatory inputs to the locomotor network and to motoneurons. It turns out that VGLUT1 knockout mice survive nicely until weaning, when they die. Could this be because they have difficulties with chewing?
192
Closing remarks Although it has been impossible within the confines of space and logic to adequately provide a summing-up of this section of this volume dedicated to Jack Feldman, Serge Rossignol and the late Jim Lund, we have made an attempt to pull together common themes and emphasize common physiological processes that govern the production of the motor rhythms associated with breathing, walking, and chewing. In the process, some new insights into the profound importance of inhibitory circuits embedded within these networks, and their control by neuromodulators, particularly 5-HT, have emerged. It is satisfying that the contributions of these three giants in the field of neural control of rhythmic movement, and those that have collected their efforts to honor them, have converged such that we now find ourselves with some new understandings of the important factors controlling these essential parts of human function. Acknowledgments The authors thank Maria Setterbom for her assistance with the preparation of the figures, and Edyta Kisielnicka and Henryk Majczy nski for assistance during the experiments carried out in the Nencki Institute of Experimental Biology PAS. Supported by the Nencki Institute of Experimental Biology (US) and by a grant (LJ) from the Canadian Institutes of Health Research (CIHR). References Antri, M., Mouffle, C., Orsal, D., & Barthe, J. Y. (2003). 5-HT1A receptors are involved in short- and long-term processes responsible for 5-HT-induced locomotor function recovery in chronic spinal rat. The European Journal of Neuroscience, 18, 1963–1972. Antri, M., Orsal, D., & Barthe, J. Y. (2002). Locomotor recovery in the chronic spinal rat: Effects of long-term treatment with a 5HT2 agonist. The European Journal of Neuroscience, 16, 467–476.
Basso, D. M., Beattie, M. S., & Bresnahan, J. C. (1995). A sensitive and reliable locomotor rating scale for open field testing in rats. Journal of Neurotrauma, 12, 1–21. Beato, M., & Nistri, A. (1998). Serotonin-induced inhibition of locomotor rhythm of the rat isolated spinal cord is mediated by the 5-HT1 receptor class. Proceedings of the Royal Society of London. Series B: Biological Sciences, 265, 2073–2080. Boulenguez, P., Liabeuf, S., Bos, R., Bras, H., Jean-Xavier, C., Brocard, C., et al. (2010). Down-regulation of the potassium-chloride cotransporter KCC2 contributes to spasticity after spinal cord injury. Natural Medicines, 16, 302–327. Bouvier, J., Thoby-Brisson, M., Renier, N., Dubreuil, V., Ericson, J., Champagnat, J., et al. (2010). Hindbrain interneurons and axon guidance signaling critical for breathing. Nature Neuroscience, 13, 1066–1074. Bouyer, L. J., & Rossignol, S. (2003). Contribution of cutaneous inputs from the hindpaw to the control of locomotion: 2. Spinal cats. Journal of Neurophysiology, 90, 3625–3639. Bracci, E., Ballerini, L., & Nistri, A. (1996). Spontaneous rhythmic bursts induced by pharmacological block of inhibition in lumbar motoneurons of the neonatal rat spinal cord. Journal of Neurophysiology, 75, 640–647. Busselberg, D., Bischoff, A. M., Becker, K., Becker, C. M., & Richter, D. W. (2001). The respiratory rhythm in mutant oscillator mice. Neuroscience Letters, 316, 99–102. Carlin, K. P., Dai, Y., & Jordan, L. M. (2006). Cholinergic and serotonergic excitation of ascending commissural neurons in the thoraco-lumbar spinal cord of the neonatal mouse. Journal of Neurophysiology, 95, 1278–1284. Courtine, G., Gerasimenko, Y., van den Brand, R., Yew, A., Musienko, P., Zhong, H., et al. (2009). Transformation of nonfunctional spinal circuits into functional states after the loss of brain input. Nature Neuroscience, 12, 1333–1342. Cowley, K. C., & Schmidt, B. J. (1995). Effects of inhibitory amino acid antagonists on reciprocal inhibitory interactions during rhythmic motor activity in the in vitro neonatal rat spinal cord. Journal of Neurophysiology, 74, 1109–1117. Crone, S. A., Quinlan, K. A., Zagoraiou, L., Droho, S., Restrepo, C. E., Lundfald, L., et al. (2008). Genetic ablation of V2a ipsilateral interneurons disrupts left-right locomotor coordination in mammalian spinal cord. Neuron, 60, 70–83. Dai, X., Noga, B. R., Douglas, J. R., & Jordan, L. M. (2005). Localisation of Spinal Neurons Activated During Locomotion Using the c-Fos Immunohistochemical Method. Journal of Neurophysiology, 93, 3442–3452. Dai, Y., Carlin, K. P., Li, Z., McMahon, D. G., Brownstone, R. M., & Jordan, L. M. (2009). Electrophysiological and pharmacological properties of locomotor activity-related neurons in cfos-EGFP mice. Journal of Neurophysiology, 102, 3365–3383. Dai, Y., & Jordan, L. M. (2010a). Multiple patterns and components of persistent inward current with serotonergic
193 modulation in locomotor activity-related neurons in cfosEGFP mice. Journal of Neurophysiology, 103, 1712–1727. Dai, Y., & Jordan, L. M. (2010b). Multiple effects of serotonin and acetylcholine on hyperpolarization-activated inward current in locomotor activity-related neurons in Cfos-EGFP mice. Journal of Neurophysiology, 104, 366–381. de Leon, R. D., Tamaki, H., Hodgson, J. A., Roy, R. R., & Edgerton, V. R. (1999). Hindlimb locomotor and postural training modulated glycinergic inhibition in the spinal cord of the adult spinal cat. Journal of Neurophysiology, 82, 359–369. Doi, A., & Ramirez, J. M. (2010). State-dependent interactions between excitatory neuromodulators in the neuronal control of breathing. The Journal of Neuroscience, 30, 8251–8262. Edgerton, V. R., Courtine, G., Gerasimenko, Y. P., Lavrov, I., Ichiyama, R. M., Fong, A. J., et al. (2008). Training locomotor networks. Brain Research Reviews, 57, 241–254. Feldman, J. L., & Del Negro, C. A. (2006). Looking for inspiration: New perspectives on respiratory rhythm. Nature Reviews Neuroscience, 7, 232–242. Feldman, J. L., Mitchell, G. S., & Nattie, E. E. (2003). Breathing: Rhythmicity, plasticity, chemosensitivity. Annual Review of Neuroscience, 26, 239–266. Funke, F., Muller, M., & Dutschmann, M. (2008). Reconfiguration of respiratory-related population activity in a rostrally tilted transversal slice preparation following blockade of inhibitory neurotransmission in neonatal rats. Pflugers Archiv, 457, 185–195. Gabriel, J. P., Mahmood, R., Kyriakatos, A., Soll, I., Hauptmann, G., Calabrese, R. L., et al. (2009). Serotonergic modulation of locomotion in zebrafish: Endogenous release and synaptic mechanisms. The Journal of Neuroscience, 29, 10387–10395. Gabriel, J. P., Mahmood, R., Walter, A. M., Kyriakatos, A., Hauptmann, G., Calabrese, R. L., et al. (2008). Locomotor pattern in the adult zebrafish spinal cord in vitro. Journal of Neurophysiology, 99, 37–48. Gao, B. X., Stricker, C., & Ziskind-Conhaim, L. (2001). Transition from GABAergic to glycinergic synaptic transmission in newly formed spinal networks. Journal of Neurophysiology, 86, 492–502. Gezelius, H., Wallen-Mackenzie, A., Enjin, A., Lagerstrom, M., & Kullander, K. (2006). Role of glutamate in locomotor rhythm generating neuronal circuitry. Journal of Physiology—Paris, 100, 297–303. Gilmore, J., & Fedirchuk, B. (2004). The excitability of lumbar motoneurones in the neonatal rat is increased by a hyperpolarization of their voltage threshold for activation by descending serotonergic fibres. Journal de Physiologie, 558, 213–224. Goulding, M. (2009). Circuits controlling vertebrate locomotion: Moving in a new direction. Nature Reviews Neuroscience, 10, 507–518.
Goulding, M., & Pfaff, S. L. (2005). Development of circuits that generate simple rhythmic behaviors in vertebrates. Current Opinion in Neurobiology, 15, 14–20. Gregory, M. L., Guzauskas, G. F., Edgar, T. S., Clarkson, K. B., Srivastava, A. K., & Holden, K. R. (2008). A novel GLRA1 mutation associated with an atypical hyperekplexia phenotype. Journal of Child Neurology, 23, 1433–1438. Grillner, S., & Jessell, T. M. (2009). Measured motion: Searching for simplicity in spinal locomotor networks. Current Opinion in Neurobiology, 19, 572–586. Harris-Warrick, R. M., & Cohen, A. H. (1985). Serotonin modulates the central pattern generator for locomotion in the isolated lamprey spinal cord. The Journal of Experimental Biology, 116, 27–46. Hart, B. L. (1971). Facilitation by strychnine of reflex walking in spinal dogs. Physiology & Behavior, 6, 627–628. Harvey, R. J., Topf, M., Harvey, K., & Rees, M. I. (2008). The genetics of hyperekplexia: More than startle!. Trends in Genetics, 24, 439–447. Hasbargen, T., Ahmed, M. M., Miranpuri, G., Li, L., Kahle, K. T., Resnick, D., et al. (2010). Role of NKCC1 and KCC2 in the development of chronic neuropathic pain following spinal cord injury. Annals of the New York Academy of Sciences, 1198, 168–172. Hendricks, T. J., Fyodorov, D. V., Wegman, L. J., Lelutiu, N. B., Pehek, E. A., Yamamoto, B., et al. (2003). Pet-1 ETS gene plays a critical role in 5-HT neuron development and is required for normal anxiety-like and aggressive behavior. Neuron, 37, 233–247. Hinckley, C., Seebach, B., & Ziskind-Conhaim, L. (2005). Distinct roles of glycinergic and GABAergic inhibition in coordinating locomotor-like rhythms in the neonatal mouse spinal cord. Neuroscience, 131, 745–758. Hochman, S., Garraway, S. M., Machacek, D. W., & Shay, B. L. (2001). 5-HT receptors and the neuromodualatory control of spinal cord function. In T. C. Cope (Ed.), Motor Neurobiology of the Spinal Cord. CRC Press, New York, pp. 48–87. Hodges, M. R., & Richerson, G. B. (2010). The role of medullary serotonin (5-HT) neurons in respiratory control: Contributions to eupneic ventilation, CO2 chemoreception, and thermoregulation. Journal of Applied Physiology, 108, 1425–1432. Hodges, M. R., Wehner, M., Aungst, J., Smith, J. C., & Richerson, G. B. (2009). Transgenic mice lacking serotonin neurons have severe apnea and high mortality during development. The Journal of Neuroscience, 29, 10341–10349. Hsiao, C.-F., Del Negro, C. A., Trueblood, P. R., & Chandler, S. H. (1998). Ionic basis for serotonin-induced bistable membrane properties in guinea pig trigeminal motoneurons. Journal of Neurophysiology, 79, 2847–2856. Jacobs, B. L., & Fornal, C. A. (1993). 5-HT and motor control: A hypothesis. Trends in Neurosciences, 16, 346–352.
194 Janczewski, W. A., & Feldman, J. L. (2006). Distinct rhythm generators for inspiration and expiration in the juvenile rat. Journal de Physiologie, 570, 407–420. Jensen, P., Farago, A. F., Awatramani, R. B., Scott, M. M., Deneris, E. S., & Dymecki, S. M. (2008). Redefining the serotonergic system by genetic lineage. Nature Neuroscience, 11, 417–419. Jordan, L. M., Liu, J., Hedlund, P. B., Akay, T., & Pearson, K. G. (2008). Descending command systems for the initiation of locomotion in mammals. Brain Research Reviews, 57, 183–191. Kiehn, O., & Butt, S. J. (2003). Physiological, anatomical and genetic identification of CPG neurons in the developing mammalian spinal cord. Progress in Neurobiology, 70, 347–361. Kolta, A., Dubuc, R., & Lund, J. P. (1993). An immunocytochemical and autoradiographic investigation of the serotoninergic innervation of trigeminal mesencephalic and motor nuclei in the rabbit. Neuroscience, 53, 1113–1126. Kolta, A., Lund, J. P., Westberg, K. G., & Clavelou, P. (1995). Do muscle-spindle afferents act as interneurons during mastication? Trends in Neurosciences, 18, 441. Kremer, E., & Lev-Tov, A. (1997). Localization of the spinal network associated with generation of hindlimb locomotion in the neonatal rat and organization of its transverse coupling system. Journal of Neurophysiology, 77, 1155–1170. Kudo, N., Nishimaru, H., & Nakayama, K. (2004). Developmental changes in rhythmic spinal neuronal activity in the rat fetus. Progress in Brain Research, 143, 49–55. Landry, E. S., Lapointe, N. P., Rouillard, C., Levesque, D., Hedlund, P. B., & Guertin, P. A. (2006). Contribution of spinal 5-HT1A and 5-HT7 receptors to locomotor-like movement induced by 8-OH-DPAT in spinal cord-transected mice. The European Journal of Neuroscience, 24, 535–546. Lanuza, G. M., Gosgnach, S., Pierani, A., Jessell, T. M., & Goulding, M. (2004). Genetic Identification of Spinal Interneurons that Coordinate Left-Right Locomotor Activity Necessary for Walking Movements. Neuron, 42, 375–386. Lavigne, G. J., et al. (2003). Neurobiological mechanisms involved in sleep bruxism. Crit Rev Oral Biol Med, 14, 30–46. Lieske, S. P., Thoby-Brisson, M., Telgkamp, P., & Ramirez, J. M. (2000). Reconfiguration of the neural network controlling multiple breathing patterns: Eupnea, sighs and gasps. Nature Neuroscience, 3, 600–607. Liu, J., Akay, T., Hedlund, P. B., Pearson, K. G., & Jordan, L. M. (2009). Spinal 5-HT7 receptors are critical for alternating activity during locomotion: In vitro neonatal and in vivo adult studies using 5-HT7 receptor knockout mice. Journal of Neurophysiology, 102, 337–348. Liu, J., & Jordan, L. M. (2005). Stimulation of the parapyramidal region of the neonatal rat brain stem produces
locomotor-like activity involving spinal 5-HT7 and 5-HT2A receptors. Journal of Neurophysiology, 94, 1392–1404. Madriaga, M. A., McPhee, L. C., Chersa, T., Christie, K. J., & Whelan, P. J. (2004). Modulation of locomotor activity by multiple 5-HT and dopaminergic receptor subtypes in the neonatal mouse spinal cord. Journal of Neurophysiology, 92, 1566–1576. Majczynski, H., Maleszak, K., Cabaj, A., & Slawinska, U. (2005). Serotonin-related enhancement of recovery of hind limb motor functions in spinal rats after grafting of embryonic raphe nuclei. Journal of Neurotrauma, 22, 590–604. Manzke, T., Dutschmann, M., Schlaf, G., Morschel, M., Koch, U. R., Ponimaskin, E., et al. (2009). Serotonin targets inhibitory synapses to induce modulation of network functions. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 364, 2589–2602. McCrea, D. A., & Rybak, I. A. (2008). Organization of mammalian locomotor rhythm and pattern generation. Brain Research Reviews, 57, 134–146. McDearmid, J. R., & Drapeau, P. (2006). Rhythmic motor activity evoked by NMDA in the spinal zebrafish larva. Journal of Neurophysiology, 95, 401–417. Morgado-Valle, C., Baca, S. M., & Feldman, J. L. (2010). Glycinergic pacemaker neurons in preBotzinger complex of neonatal mouse. The Journal of Neuroscience, 30, 3634–3639. Mori, A., Kogo, M., Ishihama, K., Tanaka, S., Enomoto, A., Koizumi, H., et al. (2002). Effect of serotonin (5-HT) on trigeminal rhythmic activities generated in in vitro brainstem block preparations. Journal of Dental Research, 81, 598–602. Nagase, Y., Moritani, M., Nakagawa, S., Yoshida, A., Takemura, M., Zhang, L. F., et al. (1997). Serotonergic axonal contacts on identified cat trigeminal motoneurons and their correlation with medullary raphe nucleus stimulation. The Journal of Comparative Neurology, 384, 443–455. Nakayama, K., Nishimaru, H., & Kudo, N. (2002). Basis of changes in left-right coordination of rhythmic motor activity during development in the rat spinal cord. The Journal of Neuroscience, 22, 10388–10398. Nishino, A., Okamura, Y., Piscopo, S., & Brown, E. R. (2010). A glycine receptor is involved in the organization of swimming movements in an invertebrate chordate. BMC Neuroscience, 11, 6. Noga, B. R., Johnson, D. M., Riesgo, M. I., & Pinzon, A. (2009). Locomotor-activated neurons of the cat. I. Serotonergic innervation and co-localization of 5-HT7, 5-HT2A, and 5-HT1A receptors in the thoraco-lumbar spinal cord. Journal of Neurophysiology, 102, 1560–1576. Ozaki, S., Yamada, T., Iizuka, M., Nishimaru, H., & Kudo, N. (1996). Development of locomotor activity induced by NMDA receptor activation in the lumbar spinal cord of the rat fetus studied in vitro. Developmental Brain Research, 97, 118–125.
195 Pearlstein, E., Mabrouk, F. B., Pflieger, J. F., & Vinay, L. (2005). Serotonin refines the locomotor-related alternations in the in vitro neonatal rat spinal cord. The European Journal of Neuroscience, 21, 1338–1346. Pierrefiche, O., Schwarzacher, S. W., Bischoff, A. M., & Richter, D. W. (1998). Blockade of synaptic inhibition within the pre-Botzinger complex in the cat suppresses respiratory rhythm generation in vivo. Journal de Physiologie, 509, 245–254. Pratt, C. A., & Jordan, L. M. (1987). Ia inhibitory interneurons and Renshaw cells as contributors to the spinal mechanisms of fictive locomotion. Journal of Neurophysiology, 57, 56–71. Ptak, K., Yamanishi, T., Aungst, J., Milescu, L. S., Zhang, R., Richerson, G. B., et al. (2009). Raphe neurons stimulate respiratory circuit activity by multiple mechanisms via endogenously released serotonin and substance P. The Journal of Neuroscience, 29, 3720–3737. Ren, J., & Greer, J. J. (2006). Modulation of respiratory rhythmogenesis by chloride-mediated conductances during the perinatal period. The Journal of Neuroscience, 26, 3721–3730. Ribeiro-do-Valle, L. E. (1997). Serotonergic neurons in the caudal raphe nuclei discharge in association with activity of masticatory muscles. Brazilian Journal of Medical and Biological Research, 30, 79–83. Schmidt, B. J., & Jordan, L. M. (2000). The role of serotonin in reflex modulation and locomotor rhythm production in the mammalian spinal cord. Brain Research Bulletin, 53, 689–710. Shao, X. M., & Feldman, J. L. (1997). Respiratory rhythm generation and synaptic inhibition of expiratory neurons in preBotzinger complex: Differential roles of glycinergic and GABAergic neural transmission. Journal of Neurophysiology, 77, 1853–1860. Slawinska, U., Majczynski, H., & Djavadian, R. (2000). Recovery of hindlimb motor functions after spinal cord transection is enhanced by grafts of the embryonic raphe nuclei. Experimental Brain Research, 132, 27–38. Smith, J. C., Abdala, A. P., Rybak, I. A., & Paton, J. F. (2009). Structural and functional architecture of respiratory networks in the mammalian brainstem. Philosophical
Transactions of the Royal Society of London. Series B: Biological Sciences, 364, 2577–2587. Tillakaratne, N. J., de Leon, R. D., Hoang, T. X., Roy, R. R., Edgerton, V. R., & Tobin, A. J. (2002). Use-dependent modulation of inhibitory capacity in the feline lumbar spinal cord. The Journal of Neuroscience, 22, 3130–3143. Tryba, A. K., Pena, F., & Ramirez, J. M. (2006). Gasping activity in vitro: A rhythm dependent on 5-HT2A receptors. The Journal of Neuroscience, 26, 2623–2634. Ung, R. V., Landry, E. S., Rouleau, P., Lapointe, N. P., Rouillard, C., & Guertin, P. A. (2008). Role of spinal 5-HT2 receptor subtypes in quipazine-induced hindlimb movements after a low-thoracic spinal cord transection. The European Journal of Neuroscience, 28, 2231–2242. van Doorninck, J. H., van Der Wees, J., Karis, A., Goedknegt, E., Engel, J. D., Coesmans, M., et al. (1999). GATA-3 is involved in the development of serotonergic neurons in the caudal raphe nuclei. In J. Neurosci.19, (p. RC12). . Verdier, D., Lund, J. P., & Kolta, A. (2003). GABAergic control of action potential propagation along axonal branches of mammalian sensory neurons. The Journal of Neuroscience, 23, 2002–2007. Wallen-Mackenzie, A., Gezelius, H., Thoby-Brisson, M., Nygard, A., Enjin, A., Fujiyama, F., et al. (2006). Vesicular glutamate transporter 2 is required for central respiratory rhythm generation but not for locomotor central pattern generation. The Journal of Neuroscience, 26, 12294–12307. Wylie, C. J., Hendricks, T. J., Zhang, B., Wang, L., Lu, P., Leahy, P., et al. (2010). Distinct transcriptomes define rostral and caudal serotonin neurons. The Journal of Neuroscience, 30, 670–684. Zhao, Z. Q., Scott, M., Chiechio, S., Wang, J. S., Renner, K. J., Gereau, R. W.4th, , et al. (2006). Lmx1b is required for maintenance of central serotonergic neurons and mice lacking central serotonergic system exhibit normal locomotor activity. The Journal of Neuroscience, 26, 12781–12788. Zhong, G., Diaz-Rios, M., & Harris-Warrick, R. M. (2006). Serotonin modulates the properties of ascending commissural interneurons in the neonatal mouse spinal cord. Journal of Neurophysiology, 95, 1545–1555.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 13
On walking, chewing, and breathing—A tribute to Serge, Jim, and Jack Sten Grillner* Nobel Institute for Neurophysiology, Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden
Abstract: This Fest-schrift honors three outstanding scientists, Serge Rossignol, James Lund, and Jack
Feldman, who have pioneered research on three complex motor systems underlying locomotion, mastication, and respiration. In this brief chapter, I highlight some of their main findings, while I also include some reminiscences of my interaction with them over more than 30 years. Keywords: locomotion; spinal cord; mastication; respiration; CPG; sensory control.
(Fig. 2). These motor patterns were “designed” by evolution to be flexible and adaptable. Although they have evolved separately in different parts of CNS to solve different tasks important for the individual, the general control structure is similar (Figs. 1 and 2), and there are many common features:
Introduction This book honors three outstanding colleagues and very dear friends, Serge Rossignol, Jim Lund, and Jack Feldman (Fig. 1). I have been asked to contribute with some reminiscences from nearly 40 years of interaction with not only a scientific but also a personal account. Their “oeuvres” address similar types of conceptual problems, applied to different patterns of rhythmic motor behavior—locomotion, chewing, and respiration in vertebrates. We now know that all three are coordinated by central pattern generator networks (CPGs), but at the same time there is a powerful sensory feedback that helps adapt each motor pattern to varying external demands
*Corresponding author. Tel.: þ46-8-52486900; Fax: þ46-8-349544 DOI: 10.1016/B978-0-444-53825-3.00018-8
199
All three motor programs are under voluntary control, since they can be initiated or stopped by will. They can furthermore be modified by will—a human, man or female, can, for instance, choose to walk on high heels, or to imitate Charlie Chaplin. All have characteristic cyclic motor programs (CPG) that can generate the appropriate timing of all the different muscles taking part in the behavior. These motor programs are designed to be flexible. The timing of different muscles can be phase shifted to allow for, for instance,
200
Fig. 1. Portraits from left to right of Serge Rossignol, James Lund, and Jack Feldman at the onset of their careers.
Sensory control
Drive & Adaptation
Conversely, the CPGs can also gate the sensory information from a given receptor type, so that it exerts one effect in one phase of the movement, and an opposite effect in another phase.
CPG Presynaptic modulation MNs
MNs
Fig. 2. The common control structure of locomotion, mastication, and respiration. For all three patterns of behavior, an adaptable and modifiable central pattern generation network (CPG) is responsible for the basic coordination. They activate in sequence different groups of motor neurons (MNs). The network is in all cases subject to sensory control from different receptors that can influence for instance the transition from one phase to another (e.g., inspiration to expiration). The sensory terminals of the afferents are also subject to phasic presynaptic inhibition, which can produce a phase-dependent gating of sensory transmission. The CPGs can receive drive from different command centers that regulate the level of activity. The drive for respiration can be influenced by CO2 levels, and the drives for locomotion and mastication are from different command centers.
walking backward, sideways, or forward or produce the different chewing motor patterns. All three CPGs are affected by movementrelated sensory information from a variety of receptors that adapt the movements to a dynamically ever-changing environment. The sensory control of the phase transitions from, for instance, inspiration to expiration is a very important type of control mechanism.
During the decades that our three key figures have been active, a dramatic increase in our knowledge of the neural bases of these three patterns of behavior has occurred. In the 1970s, the emphasis was on the actual demonstration of the existence of detailed central pattern generator networks (CPGs), and the fact that there different sensory systems are integrated in the control system. Since then the focus has shifted to the elucidation of these functions at the microcircuit, cellular, and molecular levels. Since Jack, Jim, and Serge have discussed their current research during the meeting, I have no reason to duplicate this effort, and I will instead give a brief personal account of their major contributions to our field of research. Serge Rossignol—A master of feline locomotor control After Serge's MD degree at McGill, he shifted to neuroscience, first an MSc with Collonier, and then a PhD with Melville Jones on an original project on the neural control of hopping in humans occurring at a preferred frequency of around 2 Hz—also characteristic of the beat in dance music. During a trip to Europe, he visited my newly formed laboratory in Gothenburg and
201
subsequently came to work with me (1973–1975). This was a very creative and stimulating period for both of us. At the same time, Peter Zangger, a postdoc from Zurich, was there as well as Hans Forssberg and Peter Wallén, who were beginning PhD students. Claude Perret came from Paris during extended periods, and toward the end Reggie Edgerton came for a sabbatical. Serge is friendly, thoughtful, and not only very much committed to his field of research but also very much involved in creating a good research structure in Montréal and Canada. He used to be a Quebecer with de Gaulle inspired views, as many of his colleagues at the time. I was first exposed to this aspect of Serge at a symposium in Edmonton just before Serge joined me. At the symposium dinner, when the Lieutenant Governor of Alberta entered the room, all guests stood up and sung “God save the Queen,” except one. Serge remained sitting with no intention to contribute to the national anthem of the Commonwealth. I was impressed. Now he has mellowed. Some months later when he had just arrived to Gothenburg, he explained that he badly needed to find a piano. To play the piano has, I believe, always been a way for Serge to change his mind-set, a way to maintain sanity in the hectic world of science. Some years ago, when he had received the Reeve–Irvine award for his outstanding work on spinal control of locomotion, he used the award to purchase a new grand piano. In Gothenburg, we had started to look at the ability of kittens, spinalized some weeks after birth, to produce locomotor movements. They were at the center of Serge's and my own interest over these years. We found that cats devoid of any supraspinal control could not only walk with the appropriate EMG and movement pattern but also adapt to the treadmill speed. Sensory input influenced by hip position could trigger the phase transition from extension to flexion. Another important finding was of a phase-dependent reflex reversal occurring during locomotion. A light cutaneous touch on the dorsum of the paw would give an enhanced flexion if the stimulus occurred
during the swing phase, while the identical stimulus during the support phase instead resulted in extension. This phase-dependent gating of the sensory input to different muscle groups was the result of a phasic control exerted by the CPG on alternative reflex pathways. This was a very creative period and we published seven original papers from this period—still cited today—and some reviews. In the 1970s, many European labs, including my own, published in alphabetical order (e.g., required by Journal of Physiology and Acta Physiologica Scandinavica). This was practical at the time, but the sequence of authors on a study gave no information concerning their relative contribution to the study. This made Serge end up last on most publications, which he most certainly would not have done today. In 1975, Herbert Jasper retired from the very successful MRC group that he had founded at University of Montréal (Groupe de Recherche en Sciences Neurologiques, MRC). Yves Lamarre took over the leadership and Serge joined as a junior member. Serge later became a very creative and ambitious leader of this growing constellation of research groups with a focus on motor control in vertebrates—over several decades, one of the most attractive research environment to be found anywhere in this general field of science. A hallmark of Serge's research has since been a focus on spinal locomotor mechanisms, whether fictive or real, with a detailed analysis of motor or movement pattern combined with precise analyses of neural mechanisms. Upon his return to Montreal, Serge further developed his interest in the phase-dependent reflex modulation from spinal to decerebrate and intact cats, exploring modulation in both forelimbs and hindlimbs. He also found that the same principles applied in the masticatory system in a series of studies with Jim. A general principle of CPG gating of sensory input during rhythmic behavior was established (Fig. 3). The possibility that a sensory gating would occur already at a presynaptic level (sensory terminals) was also explored with Dubuc and Gossard, and it was indeed established that a
202 Phase-dependent CPG control of reflex transmission/reflex reversals Locomotion
Mastication
Swing-enhanced flexion
Movement
CPG closing occlusal
opening
Test reflex
Support phaseenhanced extension
affer
•
F
low T
E high T
Membrane potential Digastric
Masseter A
B
Fig. 3. Phase-dependent reflex reversal during locomotion and mastication. (a) Left: the phase-dependent reflex reversals that can occur during locomotion (courtesy of KG Westberg). A tactile stimulus to the paw occurring during the swing phase elicits an enhanced flexion, as to overcome an obstacle impeding the movement, but the identical stimulus occurring during the support phase will instead elicit an enhanced extension (Forssberg et al., 1977). The same sensory stimulus can thus give opposite effects in different phases of the movement, which is due to a gaiting of the reflex pathways by the central pattern-generating network. To the right is shown the corresponding findings in the masticatory system with different effects in the masticatory cycle (Lund and Rossignol, 1981).
phasic presynaptic gating occurs in individual afferents. Moreover, the modulation occurs in different phases of the movements in different groups of afferents, which provides a selective control of the afferent input—this is now a focus for the Gossard laboratory. The resistance in the neuroscience community to accept that spinal integrated functions could play a significant role was pronounced, and even more so in clinical neurological circles. A problem that came up in the debate, at the time, was that although cats spinalized at birth could generate well-coordinated locomotor movements, the question arose if this actually could be generalized to adult cats. The view that all control
originated from the cerebral cortex was dominating (and still is in many circles)— although now the role of spinal CPGs has been accepted. At the time, it was known that adult spinal cats left alone would generate some alternating limb movements, but hardly wellcoordinated locomotion. The Rossignol laboratory firmly established that also adult cats could be made to recover walking on the treadmill belt with well-coordinated movements. They needed, however, to be trained by applying walking-like movements on the treadmill during an extended period of time. Such adult spinal cats can actually be made to perform well-coordinated walking movements over months and years.
203
Spinal section
Reflective market
(b)
Time code
Kinematic model Hip
T10
EMG electrodes Knee Ankle
Foot contact
Swing
Knee Ankle
MTP
T11 T12 T13
MTP
Foot lift
Stance
Hip
Rostral Right Left
L1 SPINAL LOCOMOTOR CPG
Head socket (EMG) Intrathecal cannula
INTACT
Cannula inlet
PARTIAL LESION
(a)
cord lesion with a unilateral section, followed with a delay by a section of the remaining part of the spinal cord (Fig. 4). The recovery of wellcoordinated locomotion was almost immediate after the second lesion—implying that there is plasticity also at the spinal CPG level—an important finding. This Rossignol model of spinal function has provided a very important impetus and rationale for the training programs of patients with spinal cord injury that now start to become developed in both North America and Europe. Another important question addressed by the Rossignol laboratory is how different descending pathways affect the locomotor movements in a phase-dependent manner. It has been addressed
COMPLETE LESION
This established an experimental model for locomotion, since then used over many years. This experimental model has also been of great importance for developing rehabilitation strategies for patients with spinal cord injury (collaboration with Barbeau). Serge and many of his collaborators have continued to explore this model in a multitude of different ways. An important aspect was to show that noradrenergic and serotonergic drugs can not only affect the locomotor movements but also reduce the time for recovery of locomotor movements after spinal cord injury. A recent conceptually important finding in this context was to show that after a two-stage spinal
L2 L3
L4
Caudal
Fig. 4. Recording strategy during in vivo recording from the cat locomotor system. (a) The complex setup for chronic recordings of intact and spinal cats during locomotion in which a large number of muscles can be recorded, which can be combined with a detailed study of the limb movements. (b) The recent results of Rossignol with a partial hemilesion (blue) of the spinal cord followed by a complete lesion after the cat has recovered locomotor function after the first lesion. After the second lesion, there is a fast recovery of locomotion on the lesion side, which is interpreted to mean that a synaptic plasticity has occurred within the spinal CPG (Barrière et al., 2008).
204
in both intact and decerebrate preparations in a series of pioneering studies of the Rossignol laboratory together with Drew and Perreault. The general question of the role of different descending systems in goal-directed behavior is now in the center of interest for the Drew laboratory in the Montreal group. Serge holds now a Canada Research Chair of research and his laboratory continues as for many years to be one major center for the study of the neural bases of locomotion. The importance of his research for the field of spinal cord injury and rehabilitation was recognized when Serge received the Christopher Reeve–Irvine award in 1999 together with Reggie Edgerton, and 9 years later, his collaborator Hughes Barbeau together with Susan Harkema received the same for the development of training programs for patients with spinal cord injury based to a large degree on the animal experiments carried out in the Rossignol laboratory. In 2003, he received another sign of international recognition, the Ipsen plasticity award, together with Francois Clarac and me.
Jim Lund—Pioneer in the neural control of chewing I have known Jim since the 1970s, primarily through my close interaction with the Montréal group but also through a number of meetings. We have never formally collaborated, but discussed a lot. Jim has always been fun to interact with; he would often provide comments with a smile that would take some time to interpret— sometimes with a hidden meaning to be unraveled behind a seemingly straight forward comment. It was great to see him during the symposium, although only too brief. It is difficult to accept that this was the very last time I would meet him—he had so much left to do. He seemed to be in good shape and had a sly smile combined with his friendly gaze. I remember nice dinners at his beautiful home on the “mountain” in Montreal in the early days, and later meetings in
Montreal and throughout the world including my home in Stockholm. Jim moved with his parents from England to Australia where he went to school and became Bachelor of Dentistry at the University of Adelaide in 1966. He then joined Peter Dellow at University of Western Ontario, who had recently moved from Adelaide to London, Ontario. He received his PhD in 1971 and subsequently moved on to the University of Montreal, where he formed his laboratory that developed into the international center for the study of the neural bases of mastication. He maintained this laboratory at UMN for nearly 40 years. In parallel from 1995 to 2008, he served as a much appreciated Dean for the Dental Faculty at McGill, a position in which he transformed the Faculty and the education, stimulated research as well as outreach activities to the society. He brought the Faculty back on track, from being proposed to be closed down by the McGill senate, to one of the leading Dental faculties in North America. In a study in 1971 with Dellow, published in the Journal of Physiology, he concluded “. . . that mastication is controlled by a brain-stem pattern generator which can be activated by adequate inputs from certain higher centres and, as concluded in other studies, from the oral cavity itself.” This important paper has provided a corner stone and much of the rationale for the scientific direction of his laboratory over several decades. It was shown that a rhythmic well-coordinated masticatory motor pattern could be elicited by stimulation of cortex, and other structures, in immobilized (by curare) rabbits. This study did away with the predominant view at the time, proposed initially by Sherrington. It suggested that rhythmic mastication would result from an alternating activation of two opposing reflexes, the jaw-opening reflex activated by tooth pressure, followed by jaw closure due to stretch activation of the jaw closer muscles. In the areas of locomotion, mastication, and respiration, there was a strong “lobby” for
205
variations of the chain reflex concept over more than half a century. Alternating movements set up by opposing reflexes were seen to generate the motor pattern without any further need for networks within the nervous system to contribute. In the locomotor area, similar views were held not only by Sherrington but very prominently also by James Gray in Cambridge, UK. This was despite the fact that von Holst, among others, had clearly demonstrated (in the 1930s) in a variety of vertebrate species that well-coordinated actual locomotor movements could remain also after a complete transection of the sensory innervation. Previously, Graham Brown (1911) had shown that if the spinal cord was pinched with a forceps, a brief period of alternating activity would occur in a pair of ankle muscles also after a complete dorsal root transection. This experiment, although very crude, showed, however, clearly that alternation could be generated by spinal networks without any sensory interaction. For almost half a century, there were two camps, those that proposed central networks (CPGs in modern terms) to be responsible, and were opposing those that with great conviction held that the important mechanism were variations of the chain reflex concept. From the perspective of history of science, the simpleminded tendency to formulate questions in an “either/or” format is not only surprising but also counterproductive, for instance, central versus reflex control or pre- versus postsynaptic with regard to synaptic plasticity. Since the 1980s, it is now generally accepted that the basic coordination underlying breathing, chewing, as well as locomotion (through the work of Jack, Jim, Serge, and others) is produced in concert by, on one hand, a CPG that can provide the correct timing for all the different muscles taking part in the behavior and, on the other hand, a sophisticated sensory control that helps adapt the movements to an ever-changing environment, be it the quality of the food subjected to chewing or walking in a complex terrain. Why do researchers have the unfortunate inclination
to pose questions as either/or rather then ask how a given function may be handled by the nervous system? From a functional point of view, a central network combined with a sophisticated sensory control seems intuitively to be the solution of choice—and it does indeed coincide with the strategy chosen by evolution. One factor contributing to the tendency for a one-eyed view of a problem may be that researchers tend to overemphasize the importance of the approach they have chosen not only in their own mind but perhaps also for making a good case in relation to the granting agencies. One may note that there are still occasionally outcries from one or the other camp. Back to the Lund laboratory, Jim's primary choice of experimental model for chewing was the rabbit, but when relevant cats, monkeys, and humans were used, and more recently rodents. After having demonstrated the presence of a masticatory CPG, he continued to work on the sensory control mechanisms involved. He detailed the different sources of sensory input during the masticatory cycle. The muscle spindle control from the jaw-closing muscles, exerted via an alpha–gamma linkage muscles, was in focus, as was the force transducers, the tendon organs. Similarly, the control of the jaw-opening muscles would occur via sensory receptors in the oral cavity and periodontal receptors sensing the biting force. The phase-dependent modulation of the reflexes was explored together with the Rossignol laboratory and Kurt Olsson in the 1980s (see above and Fig. 3). Jim has worked on all components (Fig. 5b and c) of the masticatory control system—both the CPG with its control from the frontal lobe and the different sensory components. From an evolutionary perspective, it can be noted that vertebrates from fish to reptiles (see crocodile in Fig. 5a) and birds can bite forcefully to divide the food into parts that can be instantly swallowed. They can also use the jaws to manipulate the food or other objects—that is, to maintain a well-controlled biting force for an extended
206 (a)
(c)
Only mammals chew
CPG
Motoneurons opening muscles closing muscles
Sensory feedback
(b)
Descending control
Reptiles just bite and swallow
Patterns of Mastication
Fig. 5. The neural control strategy underlying mastication. Only mammals are reported to be able to actively chew. Other classes of vertebrates, including the crocodile (a), just bite and swallow. (b) The main component of the masticatory system, a brainstem CPG, which generates the alternating generation in jaw closers (masseter muscle) and jaw openers (digastric muscle). The sensory feedback is indicated as coming primarily from periodontal receptors. The control from the cortical masticatory area is also indicated (Courtesy of K. Olsson). (c) A conceptual scheme of the masticatory control system including the variability observed in the jaw movements, which can emphasize activity on one side or the other. The CPG is driven from brainstem and cortical areas and is thought to have different modules, which allow different emphasis on the activation of different muscle groups, as discussed also in the context of the locomotor CPG. There is similarly a sensory control indicated, which can affect both the CPG level and the motorneurons. The different groups of motorneurons are indicated in blue.
time. Only mammals appear to have developed the further sophistication of grinding the food, that is, chewing, before it being swallowed. Another interesting aspect is the precise coordination required between the CPGs for chewing, swallowing, and respiration. It was detailed in the studies by McFarland and Jim. During speech and other forms of vocalization, the motoneurons controlling the jaws and the shape of the oral cavity are the same as for chewing. The question, not
yet answered, is whether parts of the modular CPG organization for mastication also contribute in speech, as the rapid sequence of phonemes (individual sounds) are formed or in syllables of the bird song or the warning calls elicited from the brainstem level. In the analysis of the central control of mastication, one important aspect was the flexibility of the jaw movements, and it was shown that different loci in the frontal lobe can elicit different
207
forms of chewing with varying lateral components, as indicated in diagram of Fig. 5c. The CPG extends from the trigeminal level to obex and is thought to be modular, in much the same way as the locomotor and respiratory CPGs. Arlette Kolta and her laboratory have interacted with Jim over more than two decades (altogether 21 papers with Jim). She has played a prominent role in the analyses of the premotor interneuron circuitry in the brainstem, feeding into the trigeminal and hypoglossal motoneurons, which in turn generate the chewing movements. These different groups of interneurons form most likely components of the masticatory CPG. To analyze components of any mammalian CPG requires patience and a long-term perspective. The interneurons need to be defined, who they talk to and their synaptic and cellular properties. One very unusual design of the trigeminal system is that the sensory cell bodies of muscle spindle afferents from the jaw-closing muscles are located within the brainstem. Moreover, their function is, however, much more sophisticated than ordinary sensory cell bodies. They receive phasic synaptic input and have pacemaker properties and are thought to contribute to the CPG activity. Jim together with Kolta and two Swedish collaborators Westberg and Olsson from Umeå (initially postdocs in the Lund laboratory) has provided much of the cellular knowledge available for the masticatory CPG. Jim received an honorary doctorate from the University of Umeå in 1995. The collaboration has continued until now, and I had the pleasure of examining an excellent PhD thesis in Umeå in 2009, jointly supervised by the Kolta–Lund laboratories and K.G. Westberg in Umeå. During the past few decades, Jim has contributed very importantly also in the clinical area, not only as a dean but also in research related to pain, dentures, implants, and dentistry in general. The clinical work has to a very large degree been carried out with his wife Jocelyn Feine, actually no less than 49 of Jims' 159 articles listed in Pub Med are together with Jocelyne—indeed a very impressive family dedication to science.
Jack Feldman—A breathing oracle Jack is a New Yorker with a Bachelor's degree from the Polytechnic Institute of NY, and a PhD from the University of Chicago; both degrees are in Physics. The PhD dealt with respiratory networks from a theoretical perspective. He subsequently went wet, with a postdoc in Paris with Gauthier and a second postdoc with Mort Cohen in New York. Mort's laboratory was at the time one of three leading laboratories in respiratory control. In 1978, he got his first academic appointment as assistant professor at Northwestern in Chicago, where he went through the ranks to full professor. In 1986, he moved to UCLA, where he is now “distinguished professor.” I got to know Jack well, when he spent a sabbatical in my laboratory in the early 1980s, and we have interacted ever since in Stockholm and at UCLA on a regular basis. Jack is fun, very intense, and once engaged he has a very strong drive to pursue an idea or a problem. Jack has a critical mind and is able to pinpoint weak and strong points, not only in the work of others but also in his own work. He is able to revise his interpretations as new data come in. He does not hesitate for demanding projects and tends always to be on the cutting edge methodologically as well as conceptually. He likes to achieve, whether in science or golf. He enjoys fast cars and does not mind a cabriolet. Jack came from a modeling environment— pools of respiratory motoneurons and models of respiratory pattern generation were part of his thesis work published in Biological Cybernetics in 1975. He must not have been satisfied with the available experimental data and therefore moved directly to experimentation, a line that he has pursued for another 35 years or so. He has developed his laboratory to the leading position in the neural control of respiration. At this time, it was known that there was a brainstem respiratory CPG—and that the level of activity was increased with increasing CO2 levels. The location of the CPG had, however,
208
not been identified and a number of nuclei were implied like those of the solitary tract, and retroambigualis, the pneumotaxic center, and the ventral and dorsal respiratory groups. The sensory control of the CPG via vagal lung volume receptors had been identified. Their level of activity would increase with the level of inspiration, and the increased vagal input would promote the termination of the inspiration. It was, at the time, referred to as an inspiratory off switch. Jack's early work carried out in vivo on the cat and dealt with the different mechanisms that control the duration of the different phases of the respiratory cycle. Subsequently, he recorded unit activity in the different respiratory nuclei and could show that units located close to each other could be tightly coupled, but further away or at the corresponding location at the two sides of the brainstem, respiratory modulated units would be only weakly correlated, if at all. To identify critical parts of the CPG, he made an extensive study combining microstimulations and microlesions, the structures thought to constitute the CPG. He concluded, however, that (Speck and Feldman, 1982) Despite the powerful short latency effects of microstimulation in VRG and DRG, extensive bilateral destruction of these neuronal populations had only modest effects on respiratory rhythm, while it decreased (or abolished) respiratory outflow in phrenic and recurrent laryngeal nerves. The combined results of the microstimulation and microlesion portions of this study suggest that a region (or regions) outside of the DRG and VRG might be important in the control of the respiratory pattern and that the DRG and VRG are important in determining the depth of inspiration; their role in generating respiratory rhythm needs to be critically re-examined.
After this time, Jack spent more than a year in my laboratory working on the spinal locomotor
CPG in the lamprey spinal cord in vitro preparation, identifying output signals to different classes of motoneurons and trying to identify premotor interneurons. We knew that there was a pattern of alternating excitatory and inhibitory drive to each motoneuron, but we had yet to identify the different interneurons responsible. We also wrote a review in which we compared in considerable detail the neural control systems responsible for respiratory and locomotor pattern generation—central and sensory mechanisms. The next major methodological step for mammalian respiratory as well as locomotor CPG control was the discovery of a brainstem—spinal cord in vitro preparation, in which both motor patterns could be produced (Smith and Feldman, 1987). To be able to work with the isolated brainstem, spinal cord allowed a great number of approaches that were unthinkable under in vivo conditions. A number of laboratories rapidly picked up on this approach. The following important approach was to make progressive lesions of the brainstem with a vibratome from the rostral and caudal levels while recording the respiratory activity. It was found that a tissue slice of around 0.6 mm contained the kernel of the respiratory pattern generation—and no other part generated respiratory activity by itself. A group of cells just caudal of the Bötzinger complex, the preBötzinger nucleus was identified. It contained cells provisionally identified as pacemaker cells that depended on glutamatergic synaptic transmission and did not require inhibitory synaptic transmission to generate the respiratory motor pattern. Moreover, the preBötzinger complex did not contain bulbospinal neurons, in contrast to most “respiratory nuclei” that serve as premotor interneurons. In a subsequent series of studies, the membrane properties of the preBötzinger neurons was studied, the presence of gap junctions was
209
established, and the modulation of different transmitters including amines, and nicotinic receptors was studied. It was shown that a proportion of the preBötzinger neuronal population contained tachykinin receptors (NK1R). This fact was used to administer a selective toxin, saponin that via the NK1 receptors was taken up into the cells to selectively kill this part of the preBötzinger cell population. When saponin was administered stereotaxically to the preBötzinger complex bilaterally in intact rodents, it led to a breakdown of the respiratory rhythm. This provided direct evidence that the preBötzinger complex is of critical importance for the generation of the normal respiratory pattern in the behaving animal (Gray et al., 2001). This was further corroborated in recent experiments in which somatostatin containing preBötzinger neurons could be transiently inactivated in animals engineered to express the allostatin receptor coupled to potassium channels. When the allostatin receptors were activated, they hyperpolarized the somatostatin containing neurons and blocked the respiratory movements in the intact rat. These experiments clarified the dominant role of the preBötzinger complex and its role for, in particular, the inspiratory drive, which dominates breathing at rest, when expiration is largely passive. In contrast, with fast and deeper breathing, the expiratory phase is actively driven. The Feldman laboratory reported evidence for another player with reciprocal activity to the preBötzinger complex in 2003: the retrotrapezoid nucleus/parafacial respiratory group (RTN/ pFRG). This cell group is located rostral to the
preBötzinger complex, and it is active in the expiratory phase—it is uncertain if it is at all active at slow breathing. Figure 6a shows the relation between the dominant preBötzinger complex and the expiratory RTN/pFRG. It also indicates that lung inflation will terminate inspiration by an action on the preBotzinger complex and facilitate the expiratory RTN/pFRG activity, while afferents activated by lung deflation have the opposite effect. The next level question concerns the intrinsic operation of the core part of the respiratory CPG, the preBötzinger complex. Feldman and Del Negro (2006) summarize their current view on this process as follows: Our favoured model for rhythmogenesis is the group-pacemaker hypothesis, in which the behaviour is emergent (Fig. 6b). We posit that periodic inspiratory bursts result from recurrent synaptic connections that combine with the intrinsic membrane properties of individual neurons 11,81,87 (Fig. 5). In this framework, excitatory interconnections between preBötC neurons initiate positive feedback through recurrent excitation. INaP and ICAN serve to amplify the synaptic depolarization and then generate high-frequency spiking to form the full inspiratory drive potential characteristic of all preBötC neurons. The small subset of bursting-pacemaker neurons, which can be rhythmically active even when synaptically isolated, participates in network activity but is not essential because most preBötC neurons discharge inspiratory bursts with the aid of INaP and ICAN induced by excitatory synaptic input.
210 (a)
(b) 1 Post-burst hyperpolarization: synapses silent
–
RTN/pFRG Expiratory activity 2 Recovery: endogenous activity resumes
+ Lung inflation – Transection Lung deflation
–
3 Recurrent excitation: positive feedback
+
SP-SAP lesion –
preBötC Inspiratory activity
–
4 Burst: synaptic excitation evokes intrinsic currents
Opiates
Fig. 6. The neural control underlying respiration (modified from Feldman and Del Negro, 2006). (a) The respiratory control consists of the preBötzinger complex (preBötC) that is responsible for driving the inspiratory activity and may be the primary rhythm generator. It interacts with the retrotrapezoid nucleus/parafacial respiratory group (RTN/pFRG), which is active during the expiratory phase. Whether it is active during quiet breathing is not clear. The reflex control for lung inflation and deflation is also indicated as well as the effect of opiates and the effect of substance P-saporin lesions (SP-SAP) (see also text). (b) The pattern generation in the preBötzinger complex is thought to depend on neurons with partial pacemaker properties. 1–4 in the graph indicates different stages occurring during one burst cycle. (1) After the end of a burst, the preBötzinger cells are hyperpolarized due to outward currents like calcium-dependent channels. (2) After the end of the hyperpolarizing phase, some of the cells resume because of activity due to partial endogenous pacemaker properties and also synaptically interact with other cells in the network. (3) The recurrent excitation within the network provides further excitation. (4) The burst terminates through the activation of activity-dependent currents that shut down the activity like calcium-dependent potassium currents.
Conclusion
Selected papers by Serge Rossignol
I have tried to highlight some of the major discoveries made of Serge, Jim, and Jack. By nature, this will be a subjective choice of areas— perhaps not the same that they would have chosen themselves. In any case, all three have made a lasting contribution to our understanding of the neural mechanisms underlying basic aspects of behavior.
Rossignol, S., & Jones, G. M. (1976). Audio-spinal influence in man studied by the H-reflex and its possible role on rhythmic movements synchronized to sound. Electroencephalography and Clinical Neurophysiology, 41, 83–92. Forssberg, H., Grillner, S., & Rossignol, S. (1977). Phasic gain control of reflexes from the dorsum of the paw during spinal locomotion. Brain Research, 132, 121–139. Forssberg, H., Grillner, S., Halbertsma, J., & Rossignol, S. (1980). The locomotion of the low spinal cat. II. Interlimb coordination. Acta Physiologica Scandinavica, 108, 283–295.
211 Drew, T., & Rossignol, S. (1984). Phase-dependent responses evoked in limb muscles by stimulation of medullary reticular formation during locomotion in thalamic cats. Journal of Neurophysiology, 52, 653–675. Barbeau, H., & Rossignol, S. (1987). Recovery of locomotion after chronic spinalization in the adult cat. Brain Research, 412, 84–95. Dubuc, R., Cabelguen, J. M., & Rossignol, S. (1988). Rhythmic fluctuations of dorsal root potentials and antidromic discharges of primary afferents during fictive locomotion in the cat. Journal of Neurophysiology, 60, 2014–2036. Perreault, M. C., Drew, T., & Rossignol, S. (1933). Activity of medullary reticulospinal neurons during fictive locomotion. Journal of Neurophysiology, 69, 2232–2247. Chau, C., Barbeau, H., & Rossignol, S. (1998). Early locomotor training with clonidine in spinal cats. Journal of Neurophysiology, 79, 392–409. Leblond, H., L'Esperance, M., Orsal, D., & Rossignol, S. (2003). Tread mill locomotion in the intact and spinal mouse. Journal of Neuroscience, 23, 11411–11419. Rossignol, S., Dubuc, R., & Gossard, J. P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86, 89–154. Barrière, G., Leblond, H., Provencher, J., & Rossignol, S. (2008). Prominent role of the spinal central pattern generator in the recovery of locomotion after partial spinal cord injuries. Journal of Neuroscience, 28, 3976–3987.
Selected papers by Jim Lund Dellow, P. G., & Lund, J. P. (1971). Evidence for central timing of rhythmical mastication. Journal of Physiology, 215, 1–13. Lund, J. P., & Lamarre, Y. (1974). Activity of neurons in the lower precentral cortex during voluntary and rhythmical jaw movements in the monkey. Experimental Brain Research, 19, 282–299. Lund, J. P., & Rossignol, S. (1981). Modulation of the amplitude of the digastric jaw opening reflex during the masticatory cycle. Neuroscience, 6, 95–98. Appenteng, K., Lund, J. P., & Séguin, J. J. (1982). Intraoral mechanoreceptor activity during jaw movement in the anesthetized rabbit. Journal of Neurophysiology, 48, 27–37. Lund, J. P., Lamarre, Y., Lavigne, G., & Duquet, G. (1983). Human jaw reflexes. Advances in Neurology, 39, 739–755. Lund, J. P., Sasamoto, K., Murakami, T., & Olsson, K. A. (1984). Analysis of rhythmical jaw movements produced by electrical stimulation of motor-sensory cortex of rabbits. Journal of Neurophysiology, 52, 1014–1029. Lavigne, G., Kim, J. S., Valiquette, C., & Lund, J. P. (1987). Evidence that periodontal pressoreceptors provide positive
feedback to jaw closing muscles during mastication. Journal of Neurophysiology, 58, 342–358. Kolta, A., Lund, J. P., & Rossignol, S. (1990). Modulation of activity of spindle afferents recorded in trigeminal mesencephalic nucleus of rabbit during fictive mastication. Journal of Neurophysiology, 64, 1067–1076. McFarland, D. H., & Lund, J. P. (1993). An investigation of the coupling between respiration, mastication, and swallowing in the awake rabbit. Journal of Neurophysiology, 69, 95–108. Kolta, A., Lund, J. P., Westberg, K. G., & Clavelou, P. (1995). Do muscle-spindle afferents act as interneurons during mastication? Trends in Neurosciences, 18, 441. Westberg, K., Clavelou, P., Sandström, G., & Lund, J. P. (1998). Evidence that trigeminal brainstem interneurons form subpopulations to produce different forms of mastication in the rabbit. Journal of Neuroscience, 18, 6466–6479. Lund, J. P., Sadeghi, S., Athanassiadis, T., Caram Salas, N., Auclair, F., Thivierge, B., et al. (2010). Assessment of the potential role of muscle spindle mechanoreceptor afferents in chronic muscle pain in the rat masseter muscle. PLoS ONE, 5, e11131.
Selected papers of Jack Feldman Smith, J. C., & Feldman, J. L. (1987). In vitro brainstem-spinal cord preparations for study of motor systems for mammalian respiration and locomotion. Journal of Neuroscience Methods, 21, 321–333. Smith, J. C., Ellenberger, H. H., Ballanyi, K., Richter, D. W., & Feldman, J. L. (1991). Pre-Bötzinger complex: A brainstem region that may generate respiratory rhythm in mammals. Science, 254, 726–729. Speck, D. F., & Feldman, J. L. (1982). The effects of microstimulation and microlesions in the ventral and dorsal respiratory groups in medulla of cat. Journal of Neuroscience, 6, 744–757. Gray, P. A., Janczewski, W. A., Mellen, N., McCrimmon, D. R., & Feldman, J. L. (2001). Normal breathing requires preBötzinger complex neurokinin-1 receptor-expressing neurons. Nature Neuroscience, 4, 927–930. Mellen, N. M., Janczewski, W. A., Bocchiaro, C. M., & Feldman, J. L. (2003). Opioid-induced quantal slowing reveals dual networks for respiratory rhythm generation. Neuron, 37, 821–826. Feldman, J. L., & Del Negro, C. A. (2006). Looking for inspiration: New perspectives on respiratory rhythm. Nature Reviews. Neuroscience, 7, 232–242. Tan, W., Janczewski, W. A., Yang, P., Shao, X. M., Callaway, E. M., & Feldman, J. L. (2008). Silencing preBötzinger complex somatostatin-expressing neurons induces persistent apnea in awake rat. Nature Neuroscience, 11, 538–540.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 14
Looking forward to breathing Jack L. Feldman* Department of Neurobiology, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, California, USA
Abstract: I provide a personal view of the developments since 1986 that underlie the contemporary view(s) about how the rhythm of breathing is generated and how the pattern of breathing is modulated. Two sites in the mammalian brainstem are likely to participate in respiratory rhythm generation: the preBötzinger Complex (preBötC), first described and intensely investigated since 1990, plays a well-documented essential role in normal breathing in mammals of all ages and may be principally involved in controlling inspiratory motor activity, and the retrotrapezoid/parafacial respiratory group (RTN/pFRG) that appears to play at least a modulatory role in neonatal and juvenile rodents and may be a conditional oscillator that controls active expiration. Keywords: preBötzinger Complex; retrotrapezoid nucleus; parafacial respiratory group; respiration; brainstem; medulla; hindbrain.
Breathing is a rhythmic, continuous behavior controlled by the brain and necessary for life in all mammals. The neural circuit underlying respiration has at its core a rhythm generator in the brainstem whose modulation affects the motor output to change pattern and timing of muscle contraction, driving the respiratory pump, for example, diaphragm, abdominals, and modulating airway resistance, for example, tongue and upper airways. Few other vital mammalian functions involve such direct control by a clearly defined relatively compact neural circuit. Breathing as a
centrally driven behavior in mammals is unique in its experimental accessibility from slices to awake/sleeping mammals. The behavior is rhythmic, continuous, and endogenous and the inputs and outputs at every level of analysis are straightforward and measurable. The importance of breathing to life is obvious. We must breathe continuously and reliably from birth, during wakefulness and sleep. Breathing is a behavior that is precisely modulated by metabolic demand that ranges over an order of magnitude, and integrated well with other behaviors such as speech, chewing, swallowing, and locomotion (this book). Disorders of breathing in humans are legion, and their consequences are significant.
*Corresponding author. Tel.: þ1-310-825-0954 DOI: 10.1016/B978-0-444-53825-3.00019-X
213
214
Understanding breathing is an essential foundation for efforts to explain more complex behaviors requiring movement of air such as vocalization, speech, and emotion (sighing, laughing, and crying are characterized by rather stereotyped breathing patterns) and provides a basis for development of rational therapeutic approaches to the host of diseases in which breathing is markedly disordered. The historical background for studies of the neural control of breathing can be traced through oldest recorded history and is thoroughly embedded in our culture. Just think of how often the verbs “inspire” and “expire” are used to express important actions. In the West, Aristotle and Galen showed great interest in breathing; in the East, the control of breathing plays a central role in several religions. The early history of these studies is briefly summarized in a previous publication (Feldman, 1986), which also covers the field until 1986. My personal motivation for studying breathing arose from my time as a graduate student in physics. I was trained in a physics culture where the most fundamental problems were solved from first principles. My interests in space and time, that is, relativity, expanded to include its perception, and I became very interested in understanding how the brain did this. My Ph.D. advisor, more aware than I was of the daunting complexity of biological problems, insisted that I start out in my studies of brain function on a much more straightforward and potentially solvable problem. So, I choose what I thought was a simple, even trivial (so goes the physicist's hubris) problem, that is, how does the mammalian nervous system generate respiratory rhythm, with the goal of solving this problem from first principles. When I started working on respiratory rhythmogenesis, the experimental model was the anesthetized cat, and the paradigm for rhythm generation was that it was generated by neurons in the dorsal medulla. I did not readily accept the premises that underlay this view, and my early career as an experimentalist was characterized by
extensive criticism of this idea. Moreover, after several years, I saw the cat being supplanted as a model as more and more interesting work, at first immunohistochemistry and later genetics and molecular biology, was being done in rodents, so we transferred our experimental model to rats. This was also a time of an explosion of in vitro mammalian (rodent) electrophysiology, for example, hippocampal slices, that greatly advanced our understanding of cellular processes, and I dreamed of extending in vitro approaches to studies of behavior, that is, breathing. With this in mind, and needing to find a kindred spirit, I did a sabbatical in Sten Grillner's laboratory (1983–1984) to learn the lamprey preparation as a foundation for developing an in vitro model for mammalian respiration. When Suzue published his important 1984 paper (Suzue, 1984) describing the isolated brainstem spinal cord preparation from neonatal rat that generated breathing rhythm, the experience in Stockholm allowed us to set up the preparation instantly at UCLA, and we started a series of experiments that revolutionized the field (Smith et al., 1991). With Jeff Smith and colleagues, we isolated a small region of the medulla that was essential for breathing rhythm and christened it the “preBötzinger Complex” (preBötC). We then made thin transverse medullary slices of this level and discovered that we could get rhythmic motor nerve output (indicating this slice was generating a bona fide respiratory-related behavior). Since then in a ongoing effort to test our ideas originating from this paper, we performed many experiments using a broad range of techniques that showed that: (i) lesions selectively targeting preBötC neurons that express the neurokinin 1 receptor (NK1R) in intact (unanesthetized and forebrain intact) adult rats induce a disturbed breathing pattern during sleep; with more extensive lesions, ataxic breathing occurs during wakefulness with apneas (complete cessation of breathing) during sleep (Gray et al., 2001); (ii) rapid silencing of preBötC neurons expressing somatostatin (Sst) produces profound apnea in
215
anesthetized or awake adult rats (Tan et al., 2008) (Fig. 1); (iii) juvenile rats with brainstem transection just rostral to the preBötC continue to generate rhythmic inspiratory-dominated breathing patterns. In quite a turnaround from the initial criticism of the preBötC and the utility of using in vitro preparations for understanding the neural control of breathing, the critical role of the preBötC in breathing is now discussed in major neuroscience and physiology textbooks and routinely taught to graduate and medical students. When doing our initial in vitro studies for breathing, we noticed unusual rhythmic patterns in the spinal cord, and this led to the development of the neonatal rodent spinal cord preparation for locomotion (Smith and Feldman, 1985, 1987; Smith et al., 1988; see pp. 89–90 in Stuart and Hultborn, 2008 for an account of our contribution). For better or worse, we did not have the resources to study locomotion, but it should be noted that our in vitro preparation is being used by > 20 laboratories worldwide to study locomotion (many in this book). At present, we are focusing on two distinct problems: 1. How do 1000 preBötC neurons generate not only a robust but also very labile respiratory rhythm?
What do we understand about rhythmogenesis? Quite frankly, very little that is testable other than issues of regional function (Fig. 1). In spite of a reasonable literature on preBötC neurons, most studies have been rather limited in scope, focusing on a small number of neuron types and never with an extensive dataset that seems truly representative. We cannot even say how many different types of preBötC neurons participate in rhythm generation or modulation, as characterized by their firing patterns under different in vitro conditions, their morphology, the receptor and channel distributions on their somatodendritic membranes, etc. Neither can we describe the relationship of these properties to physiological phenotype, for example, pacemaker versus nonpacemaker. Until recently, most if not all preBötC neurons were presumed to be glutamatergic, whereas it appears that about half are glycinergic (Winter et al., 2010) and that some of these neurons are even pacemakers (MorgadoValle et al., 2010). Perhaps most telling, we know little about the connectivity among preBötC neurons, with but a single study of ours that provides anecdotal information because the dataset was very limited (23 pairs of neurons; Rekling and Feldman, 1997). At two recent international symposia on control of breathing (St. Maximin, France, December
Anesthetized adult rat AL
mechanical ventilation (60 min) apnea recovery 10 s
Awake adult rat AL
apnea 10 s
Fig. 1. preBötC is essential for breathing in adult rats. Rapid silencing of allatostatin receptor (AlstR)-expressing preBötC somatostatin (Sst) neurons induces persistent (>45 min when mechanically ventilated) apnea in anesthetized or awake adult rats. Traces are plethysmographic recording of breathing movements. Allatostatin administered intracerebrocisternally induces a gradual decline of frequency and tidal volume until apnea develops after several minutes. After 60 min mechanical ventilation, rats resume spontaneous breathing. AAV2: adeno-associated virus 2; EGFP: enhanced green fluorescent protein. From Tan et al. (2008).
216
2008; Nara, Japan, July 2009), many models for respiratory rhythmogenesis were presented that shared the same problems: they were not testable, their modelers offered no predictions that could falsify them, they did not present any unusual predictions, and they did not provide any insight that could underlie the design of novel experiments. As I mentioned above, my research in breathing started with modeling (Feldman, 1976; Feldman and Cowan, 1975a,b), but I quickly turned to experimentation because any model developed at that time was poorly constrained (and that still seems to be the case). It remains straightforward to construct a mathematical/computational model that can reproduce a (carefully) selected subset of the experimental data. The principal reason is that many key parameters, for example, strength of synaptic connection between two neurons, are unknown, so one could always find a set of parameters, all with reasonable values within the “physiological range,” to fit the data. Unconstrained models that can accommodate new data simply by changing parameters can only establish proof of principle that a particular model can work but offer little insight into the neurobiological system. Yet, that is the “state-of-the-art” for models of respiratory rhythmogenesis. While contemporary models for generation of respiratory rhythm establish proof of principle insofar as they can mimic data, their utility in revealing biological mechanisms of respiratory rhythmogenesis is limited. The models are poorly constrained and represent mostly an exercise in parameterization, typically finding parameters that fit the data, and/or require stipulations that are simply not physiological, sometimes akin to modeling quadrapedal locomotion with wheels instead of limbs. For example, in many models of respiratory rhythmogenesis, neurons are points, ignoring data on the importance of dendrites of preBötC neurons (Morgado-Valle et al., 2008; Pace et al., 2007), and uniformly connected, ignoring data that their connections are most likely sparse (Rekling et al., 2000).
We may need to abandon the notion that this problem can be solved from first principles and accept the fact that we need to obtain tremendously detailed information to produce a dataset sufficient for representative and testable modeling, the neurobiological equivalent to sequencing the genome. The recent finding that the homeobox gene Dbx1 controls the fate of glutamatergic interneurons required for preBötC development (Bouvier et al., 2010; Gray et al., 2010) is a critical step in this direction and opens numerous possibilities for novel experiments to crack this network. 2. Testing the two-oscillator hypothesis (Feldman and Del Negro, 2006; Feldman et al., 2009; Janczewski and Feldman, 2006). When we discovered the preBötc, we hypothesized it was the sole generator of respiratory rhythm, both during development and adulthood. In 2003, we published a paper (Mellen et al., 2003) in which we hypothesized the presence of a second respiratory rhythm generator. Subsequent work in juvenile rats (Janczewski and Feldman, 2006) suggested that this oscillator was located rostral to the preBötc in the region of the retrotrapezoid nucleus/parafacial respiratory group (RTN/ pFRG) and was primarily a conditional oscillator for generating expiratory motor output; this is the core of our “two-oscillator” hypothesis. The existence of a second respiratory oscillator is still controversial (Onimaru et al., 2006), and despite developmental studies that show the critical role of RTN/pFRG in the ontogeny of respiratory rhythms (Jacquin et al., 1996; Thoby-Brisson et al., 2009), whether this second oscillator persists into adulthood is still debated (though with waning intensity). We recently obtained further evidence supportive of our two-oscillator hypothesis in adult mammals and demonstrated that indeed a conditional oscillator in the RTN/pFRG is responsible for active expiratory motor output when properly stimulated (Pagliardini et al., 2011). Furthermore, this expiratory oscillator is tightly coupled with the presumptive inspiratory oscillator in the preBötzinger Complex. We performed experiments
217
using a variety of methods, including pharmacological disinhibition, optogenetic stimulation, and single unit recordings, to show that activation of the adult RTN/pFRG turns on active expiration associated with silent or tonically firing neurons becoming rhythmic. We conclude that the induced transformation of tonic into rhythmic RTN/pFRG neurons is causal to the onset of active expiration. The most parsimonious explanation of our results is that the RTN/pFRG contains a conditional oscillator responsible for the generation of active expiratory activity in adult mammals. Much work remains to fully test this hypothesis. In summary, our understanding of the basic mechanisms underlying the neural control of breathing has substantially advanced since 1986. Two key sites are widely accepted to underlie the generation of the rhythm, but we are still in the dark about the mechanisms for rhythm generation, as well as how this rhythm gets transformed into a precisely controlled pattern of skeletal muscle activity that ultimately moves the appropriate amount of air to assure that blood gas are regulated and aerobic metabolism is continuously supported. My sense is that finding new “first principles” to understand the neural mechanisms, if they even exist, will be very difficult, and that the next substantial advance, even paradigm shift, will be delayed until we have a database sufficient to build meaningful models.
Acknowledgments This work was only possible through the generous and continuing support from the National Institutes of Health and the extraordinary scientists I have had the privilege to work with in my laboratory. Among them, with respect to the work I discussed here, I owe a special debt to Don McCrimmon, Jeffrey Smith, John Greer, Greg Funk, Jens Rekling, Paul Gray, Nick Mellen, Christopher Del Negro, Victor Janczewski, Consuelo Morgado-Valle, and Wenbin Tan.
Abbreviations AAV2 EGFP preBötC RTN pFRG
adeno-associated virus 2 enhanced green fluorescent protein pre-Bötzinger Complex retrotrapezoid nucleus parafacial respiratory group
References Bouvier, J., Thoby-Brisson, M., Renier, N., Dubreuil, V., Ericson, J., Champagnat, J., et al. (2010). Hindbrain interneurons and axon guidance signaling critical for breathing. Nature Neuroscience, 13, 1066–1074. Feldman, J. L. (1976). A network model for control of inspiratory cutoff by the pneumotaxic center with supportive experimental data in cats. Biological Cybernetics, 21, 131–138. Feldman, J. L. (1986). Neurophysiology of breathing in mammals. In F. E. Bloom (Ed.), Handbook of physiology; section I: The nervous system; volume IV: intrinsic regulatory systems of the brain (pp. 463–524). Bethesda, MD: American Physiological Society. Feldman, J. L., & Cowan, J. D. (1975a). Large-scale activity in neural nets I: Theory with application to motoneuron pool responses. Biological Cybernetics, 17, 29–38. Feldman, J. L., & Cowan, J. D. (1975b). Large-scale activity in neural nets II: A model for the brainstem respiratory oscillator. Biological Cybernetics, 17, 39–51. Feldman, J. L., & Del Negro, C. A. (2006). Looking for inspiration: New perspectives on respiratory rhythm. Nature Reviews. Neuroscience, 7, 232–242. Feldman, J. L., Kam, K., & Janczewski, W. A. (2009). Practice makes perfect, even for breathing. Nature Neuroscience, 12, 961–963. Gray, P. A., Janczewski, W. A., Mellen, N., McCrimmon, D. R., & Feldman, J. L. (2001). Normal breathing requires preBötzinger complex neurokinin-1 receptor-expressing neurons. Nature Neuroscience, 4, 927–930. Gray, P. A., Hayes, J. A., Ling, G. Y., Llona, I., Tupal, S., Picardo, M. C., Ross, S. E., Hirata, T., Corbin, J. G., Eugenin, J., & Del Negro, C. A. (2010). Developmental origin of preBotzinger complex respiratory neurons. Journal of Neuroscience, 30, 14883–14895. Jacquin, T. D., Borday, V., Schneider-Maunoury, S., Topilko, P., Ghilini, G., Kato, F., et al. (1996). Reorganization of pontine rhythmogenic neuronal networks in Krox20 knockout mice. Neuron, 17, 747–758. Janczewski, W. A., & Feldman, J. L. (2006). Distinct rhythm generators for inspiration and expiration in the juvenile rat. Journal de Physiologie, 570, 407–420.
218 Mellen, N. M., Janczewski, W. A., Bocchiaro, C. M., & Feldman, J. L. (2003). Opioid-induced quantal slowing reveals dual networks for respiratory rhythm generation. Neuron, 37, 821–826. Morgado-Valle, C., Beltran-Parrazal, L., Difranco, M., Vergara, J. L., & Feldman, J. L. (2008). Somatic Ca2þ transients do not contribute to inspiratory drive in preBötzinger Complex neurons. Journal de Physiologie, 586, 4531–4540. Morgado-Valle, C., Baca, S. M., & Feldman, J. L. (2010). Glycinergic pacemaker neurons in preBötzinger complex of neonatal mouse. The Journal of Neuroscience, 30, 3634–3639. Onimaru, H., Homma, I., Feldman, J. L., & Janczewski, W. A. (2006). The para-facial respiratory group (pFRG)/ preBötzinger Complex (preBötC) is the primary site of respiratory rhythm generation in the mammal. Journal of Applied Physiology, 100, 2094–2098. Pace, R. W., Mackay, D. D., Feldman, J. L., & Del Negro, C. A. (2007). Inspiratory bursts in the preBotzinger complex depend on a calcium-activated non-specific cation current linked to glutamate receptors in neonatal mice. Journal de Physiologie, 582, 113–125. Pagliardini, S., Janczewski, W. A., Tan, W., Dickson, C. T, Deisseroth, K., & Feldman, J. L. (2011). Active expiration induced by excitation of ventral medulla in adult anesthetized rats. Journal of Neuroscience. (In Press). Rekling, J. C., & Feldman, J. L. (1997). Bidirectional electrical coupling between inspiratory motoneurons in the newborn mouse nucleus ambiguus. Journal of Neurophysiology, 78, 3508–3510. Rekling, J. C., Shao, X. M., & Feldman, J. L. (2000). Electrical coupling and excitatory synaptic transmission between rhythmogenic respiratory neurons in the preBötzinger complex. Journal of Neuroscience, 20, RC113. Smith, J. C., & Feldman, J. L. (1985). Motor patterns for respiration and locomotion generated by an in vitro brainstem-
spinal cord from neonatal rat. Society for Neurocience Abstracts, 11, 24. Smith, J. C., & Feldman, J. L. (1987). In vitro brainstem-spinal cord preparations for study of motor systems for mammalian respiration and locomotion. Journal of Neuroscience Methods, 21, 321–333. Smith, J. C., Feldman, J. L., & Schmidt, B. J. (1988). Neural mechanisms generating locomotion studied in mammalian brain stem-spinal cord in vitro. The FASEB Journal, 2, 2283–2288. Smith, J. C., Ellenberger, H. H., Ballanyi, K., Richter, D. W., & Feldman, J. L. (1991). Pre-Bötzinger complex: A brainstem region that may generate respiratory rhythm in mammals. Science, 254, 726–729. Stuart, D. G., & Hultborn, H. (2008). Thomas Graham Brown (1882–1965), Anders Lundberg (1920-), and the neural control of stepping. Brain Research Reviews, 59, 74–95. Suzue, T. (1984). Respiratory rhythm generation in the in vitro brain stem-spinal cord preparation of the neonatal rat. Journal de Physiologie, 354, 173–183. Tan, W., Janczewski, W. A., Yang, P., Shao, X. M., Callaway, E. M., & Feldman, J. L. (2008). Silencing preBötzinger complex somatostatin-expressing neurons induces persistent apnea in awake rat. Nature Neuroscience, 11, 538–540. Thoby-Brisson, M., Karlen, M., Wu, N., Charnay, P., Champagnat, J., & Fortin, G. (2009). Genetic identification of an embryonic parafacial oscillator coupling to the preBotzinger complex. Nature Neuroscience, 12, 1028–1035. Winter, S. M., Fresemann, J., Schnell, C., Oku, Y., Hirrlinger, J., & Hulsmann, S. (2010). Glycinergic interneurons in the respiratory network of the rhythmic slice preparation. Advances in Experimental Medicine and Biology, 669, 97–100.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 15
Chew before you swallow James P. Lundw,1 Faculty of Dentistry, McGill University, and Groupe de Recherche sur le Système Nerveux Central, Université de Montréal, Montreal, Quebec, Canada
Abstract: The main text of this chapter, written by James P. Lund, summarizes most of the work related to the neural control of mastication that he conducted with his collaborators throughout the years. From his early PhD work showing that mastication is centrally patterned to his latest work related to the interaction between pain and movement, Lund will have addressed many essential questions regarding the organization and functioning of the masticatory central pattern generator (CPG). His earliest studies examined how the CPG modulates reflexes and the excitability of primary afferents, interneurons, and motoneurons forming their circuitry. He then tackled the question of how the CPG itself was modulated by different types of sensory and cortical inputs. Another series of studies focused on the organization of the subpopulations of neurons forming the CPG, their intrinsic and network properties. Shortly before his untimely passing, he had turned his attention to the potential contribution of muscle spindle afferents to the patterning of mastication as well as to the development of chronic muscle pain. Keywords: mastication; CPG; interneurons; reflex; brainstem; cortex; rabbit.
The title is taken from a quote by George W. Bush. He illustrated the fundamental importance of mastication to mammalian life by losing consciousness after swallowing a large piece of pretzel in 2002. In this chapter, I will outline some of the principles that govern the control of mastication and other rhythmical motor patterns and will emphasize the unique features of the masticatory system.
When I began my scientific life in the late 1960s, the mechanism by which patterned motor activity was produced by the nervous system was “one of the core questions of general neurology” (Bullock, 1961). During that decade, evidence steadily accumulated that many rhythmical motor patterns in invertebrates arose within CNS (Horridge, 1968; Wilson, 1961; Wyman, 1965) but evidence for similar control in vertebrates was not as conclusive. Ever since the pioneering work of Adrian and Buytendijk (1931), it was pretty clear that respiration was centrally patterned, but for swimming, locomotion, and mastication the jury was still out.
w
Deceased Corresponding author: Arlette Kolta, Dept. Physiology, Université de Montréal, Montréal, Quebec, Canada, H3C 3J7, Fax.: 514-343-2111, Tel.: 514-343-7112, E-mail: arette.
[email protected] 1
DOI: 10.1016/B978-0-444-53825-3.00020-6
219
220
My PhD supervisor and I were the first to prove that mastication was centrally patterned. We used an anesthetized rabbit model and induced mastication by repetitive electrical stimulation of the sensorimotor cortex, corticobulbar pathways, and other CNS sites, or by inflating a small balloon in the mouth. We showed that repetitive masticatory activity in trigeminal (V) and hypoglossal (XII) motoneurons could be evoked in decerebrate rabbits deprived of sensory feedback. Furthermore, we proved that the brainstem central pattern generator (CPG) could produce a rhythmical output from tonic synaptic inputs from higher centers (Dellow and Lund, 1971) or sensory afferents (Lund and Dellow, 1973), and that masticatory bursts were not triggered by timing cues in the stimulus train, respiration, or the heart beat. Many of the experiments on sensory feedback, reflexes, and central patterning that I will describe below used a similar preparation.
Alpha–gamma coupling Before returning to those topics, I will first describe experiments that were designed to look at the contribution of the muscle spindle loop to the control of both voluntary and rhythmical movements. These were carried out with Dr. Yves Lamarre, whose laboratory was set up for work on awake monkeys. We used the trigeminal system for these experiments because the spindle afferents and motoneurons of the jaw muscles can be recorded with microelectrodes advanced into brainstem in the awake-behaving animals. The jaw-closing muscle spindle afferents have their cell bodies centrally located in the trigeminal mesencephalic nucleus (NVmes) and not in the trigeminal equivalent of the dorsal root ganglion. NVmes extends rostrally from the trigeminal motor nucleus (NVmot) at the medulopontine junction. A stimulating electrode inserted next to the masseter nerve was used to stimulate masseter motoneurons and spindle afferents, so
as to measure their conduction velocities. This allowed the slowly conducting g-motoneurons to be identified. The monkeys were trained to bite voluntarily on a strain-gauge bar and to maintain a given force for 1 s. After a successful trial, they were given fruit juice, which they rhythmically licked from the bar. It was generally assumed that g-motoneurons (fusimotor neurons) maintain the sensitivity of muscle spindle afferents to perturbation during muscle contraction, thereby allowing modulation of central commands by feedback. The a-motoneurons that form a distinct pool are recruited in order of increasing cell size, at least during slow contractions (Gordon et al., 2004; Henneman et al., 1965). Since g-motoneurons are the smallest cells of all, they should be recruited before a-motoneurons, and it had been suggested that if g firing were high enough to augment spindle afferent output during shortening, a-motoneuron output could be controlled through a servo-assistance mechanism, “g-driving” (Stein, 1974). We were able to show that g-motoneurons of the masseter muscle were activated at least as early as the smallest a-motoneurons, but found that this did not lead to significant spindle afferent firing before contact with the bite bar stopped muscle shortening (Lund et al., 1979), suggesting that there was no g-driving of a-motoneurons. We also found that g output appeared to be less during rhythmical licking than during biting, which suggests that g and a drive can be modified independently, at least to some extent. This confirmed work by Prochazka et al. (1977), who had found evidence that g-drive seems to go up during locomotion. Later studies of limb muscle spindle afferents by this group and others lead to the concept of “fusimotor set” (Prochazka et al., 1985), which proposes that a- and g-motoneuron firing within a pool is not rigidly linked and varies according to task. However, no recordings of fusimotor neurons during the performance of a task have been published except for our paper in 1979.
221
Generation of distinct motor patterns by reorganization of subpopulations within the CPG The ordering and scaling of the output of separate motoneurons pools vary greatly between the different movement patterns that characterize mastication and locomotion. The term locomotion covers a series of named stereotypical patterns (walking, jogging, running, etc.), while the pattern of mastication varies not only between foods but also from the beginning to the end of a single sequence of movements (Peyron et al., 2002; Woda et al., 2006). The ways in which CPGs generate distinct patterns of motoneuron firing has been exhaustively investigated in invertebrates and lower vertebrates, and evidence has been found for three basic CPG structures: dedicated circuits, distributed circuits, and reorganizing circuits (Morton and Chiel, 1994). We decided to investigate this in our rabbit mastication model. Like humans, rabbits can chew on the left or right molar teeth, but they also rhythmical crop off plants with the incisor teeth, which is very different to molar chewing. We took advantage of the fact that, unlike locomotion, mastication is represented in the sensorimotor cortex: furthermore, specific patterns are represented at different sites so that a kinesiotopic map can be drawn on the surface of the brain. We had shown that repetitive stimulation of rostral sites causes low-amplitude jaw movements mainly in the vertical plane, while large amplitude movements with grinding on the contralateral molar teeth are represented at caudal sites (Lund et al., 1984). We evoked four of these motor patterns in paralyzed animals while recording from interneurons within the lateral brainstem (rostral spinal trigeminal nucleus, parvocellular reticular formation). We found that almost 50% of neurons were only active during one motor pattern, 35% during two, and most of the rest during three (Westberg et al., 1998). When a neuron fired rhythmically during more than one pattern, we found that there were small but significant changes in frequency and/or burst phase between
patterns. This data suggest that the masticatory CPG produces distinct motor patterns mainly through the reorganization of subpopulations. Later, we carried out a similar experiment while recording from the neurons in the Vth main sensory nucleus (Tsuboi et al., 2003). Despite the fact that this nucleus is usually regarded as purely sensory structure, we found a population of neurons that were rhythmically active during fictive mastication. These were confined to the dorso-rostral pole of the nucleus, a region known to project directly to the Vth motor nucleus. Like the neurons recorded in spinal V and the reticular formation (Westberg et al., 1998), neurons of the main sensory nucleus that fired phasically during mastication received sensory inputs from the oral mucosa, periodontal pressoreceptors, and muscle spindles, the sensory inputs that are crucial for continuous modulation of masticatory motor program in response for changes in the physical properties of food particles. All the subsequent work on CPG circuits and cellular mechanisms pattern generation has been carried out using the in vitro slice preparations developed by Dr. Arlette Kolta, and the results were presented in Chapter 9.
Modulation of reflexes It appears that all CPGs modify sensory feedback in some way, and we have studied this extensively in the masticatory system. We began with feedback from periodontal pressoreceptors. The teeth are anchored to the jaw bones through the fibrous periodontal ligament, and this contains Ruffini-type end organs that sense pressures applied to the crowns of the teeth (Byers, 1985). These receptors begin to fire as soon as the teeth engage the food during mastication (Appenteng et al., 1982). Studies carried out using microneurography in humans show that their thresholds are very low, and that their output saturates at forces well below the maximum exerted during chewing or biting (Trulsson,
222
2007). When these receptors are excited by tapping on teeth, they cause the jaw-opening reflex (Sherrington, 1917), the trigeminal equivalent of the flexion-withdrawal reflex of the limbs. The jaw-opening reflex can also be triggered by other orofacial mechanoreceptors and by stimulation of nociceptors. During the jaw-opening reflex, the motoneurons of jaw-closing muscles (masseter, temporalis, medial pterygoid) are strongly inhibited and, in experimental animals but not humans, the jaw-opening motoneurons (digastric, mylohyoid) are activated. Activation and inhibition are both disynaptic. We triggered this reflex by nerve stimulation before (control) and during cortically induced mastication in anesthetized rabbits (Lund et al., 1981). When we stimulated the nerve close to threshold, thereby activating only large-diameter afferents from periodontal and other low-threshold mechanoreceptors, we found that the reflex was tonically suppressed during mastication. When the nerve was stimulated at much higher intensity, enough to activate all groups of sensory afferents, the reflex was enhanced during the jaw-closing phase of mastication. When it occurred in the second half of the phase, it often terminated jaw-closing muscle EMG bursts. At the same time, there was a big increase above control levels in the amplitude of the excitatory response in the digastric muscle. As a result, very rapid jaw opening occurred. To explain an increase in digastric amplitude that was out of phase with the motoneuron burst, we proposed that the CPG must phasically increase the excitability of interneurons that receive nociceptor input and found evidence for this in a study of interneurons carried out by Olsson et al. (1986). We concluded that the masticatory CPG exerts differential control of the jaw-opening reflex circuits, tonically suppressing responses from mechanoreceptors while phasically facilitating nociceptor-triggered excitation and reciprocal inhibition. Mechanoreceptor feedback that would be disruptive is blocked, while protection from
tissue damage caused by biting the tongue or something hard is enhanced. Later, we studied the effects of tonically stimulating the periodontal receptors during the jaw-closing phase of cortically induced mastication. A small metal ball was thrust between the upper and lower teeth by a piston and then left there for several cycles. When the lower teeth first struck the ball, a jaw-opening reflex was triggered, but adaptation had occurred by the next cycle. With the ball in place, the jaw-opening reflex did not occur: instead, cycle duration increased, rather than decreased, as did the area of the EMG bursts from jaw-closing muscles. When we removed periodontal inputs by anesthetizing the nerves supplying the teeth, the area of the EMG bursts fell significantly with the ball in place. This showed that periodontal receptors appear to provide positive and not negative feedback to jaw-closing muscles during mastication (Lavigne et al., 1987). Although the jaw muscles contain a few Golgi tendon organs (Lund et al., 1978), it is probable that a similar switch occurs with the inputs from Golgi tendon organs of limb extensors during locomotion (Pearson, 2008). As could be expected, jaw-closing muscle spindle afferents also provide positive feedback during mastication, and they may provide feedforward adjustment of the CPG output (Komuro et al., 2001).
A new form of presynaptic modulation One of the ways that CPGs modify reflexes and sensory inputs is by depolarization of primary afferent terminals leading to presynaptic inhibition (Eccles et al., 1962; Rudomin and Schmidt, 1999). Serge Rossignol and his graduate students at the time, Réjean Dubuc and Jean-Pierre Gossard, were among the first to show that this can lead to bursts of antidromic action potentials in spinal afferents during fictive locomotion (Baev, 1980; Dubuc et al., 1988; Gossard et al., 1989, 1991). Because it was relatively easy to
223
record from the cell bodies of spindle afferents from jaw-closing muscle in the Vth mesencephalic nucleus, we decided to see if we could find evidence of antidromic firing in these identified afferents during fictive mastication, assuming that antidromic firing would appear as a phasic excitation at the soma. In fact, phasic excitation was seen in only 6% of cells during fictive mastication when they were silent. So we tested the effect of mastication on the firing elicited by tonic stretch of the muscle. There was no change in firing in the majority (60%), but 34% showed phasic inhibition that mainly coincided with what would be the jaw-opening phase of movement, and therefore the period of passive stretch of the spindles (Kolta et al., 1990). After a break of several years, we came back to this question by comparing the firing patterns of the somata of the spindle afferents, recorded in the mesencephalon, with that of their central axons recorded from tracks in the lateral medulla (Kolta et al., 1995; Westberg et al., 2000). Two cortical sites were stimulated so that firing during different fictive masticatory patterns could be compared. As in the 1990 study, about 32% of mesencephalic units were phasically inhibited, with very little evidence of excitation. Inhibition of firing was seen even more frequently in the central axons (83%), but the biggest difference was that 94% of central axons showed excitation in the jaw-closing phase, probably because of the addition of antidromic spikes generated at the central terminals. Furthermore, the bursts of antidromic firing varied strongly from one fictive pattern to another. Although the vast majority of these antidromic spikes seem to fail before they reach the soma, we were able to show with spike-triggered averaging that some do reach the Vth motor nucleus. These findings and our speculations on the significance of the findings are summarized in Fig. 1. It includes the hypothesis that presynaptic inhibition of muscle spindle afferent terminals leads to the antidromic firing during fictive mastication. Spindle afferents have many terminals in the
lateral pons and medulla among interneurons of the CPG. Because different CPG subpopulations are active during distinct motor patterns (see below), we proposed that differential presynaptic modulation of primary afferents is the cause of distinct patterns of antidromic firing (Kolta et al., 1995). Because the antidromic spikes appear to contribute to motoneurons output (first suggested by Eccles et al., 1961), we suggested that the central axons of these afferents act as an interneuron during mastication. We also proposed that some type of active filter blocked orthodromic spikes to the cell body and central axon during the jaw-opening phase, and also the passage of antidromic spikes to the cell body during the jaw-closing phase. We tested this hypothesis in a series of in vitro experiments using brainstem slices from young rats (Verdier et al., 2003). The slices contained the cell bodies and central axons of muscle spindle afferents that were filled with a lipid soluble carbocyanine dye (DiI), injected into the masseter muscle. We recorded from the cell bodies and evoked antidromic spikes by microstimulation along the descending axonal tract. Application of GABA from a pipette between the stimulation site and the cell body blocked the antidromic spikes about one-third of the time, and this effect was reversed by picrotoxine, suggesting that activation of GABAA receptors could block action potentials. Stimulation of local interneurons in the supratrigeminal area and in the Vth main sensory nucleus also blocked antidromic spikes. Because the antidromic spikes were completely blocked and not simply reduced in amplitude, we concluded that release of GABA caused shunting in the central axon, and that the CPG circuits used this mechanism to disconnect the rostral and caudal parts of the central axonal arbor during mastication. This would allow feedback control of motoneurons output through the rostral segment during mastication while the caudal segment carried part of the CPG output.
224 1. Synaptic inputs on the soma Control 2. Synaptic inputs on axonal terminals:
Mastication
- Presynaptic inhibition - Antidromic firing Motor nucleus
3. Synaptic inputs on the axon - Controls propagation of action potential along axonal arbors
Masticatory CPG
Control Mastication
Fig. 1. Functional model of jaw-closing muscle spindle afferent. This figure summarizes the evidence gathered over the years regarding synaptic inputs to jaw-closing muscle spindle afferents and illustrates our speculations about their impact on the function of these neurons. (1) Synaptic inputs on the soma are probably responsible for the phasic inhibition of the orthodromic spikes (black bars) elicited by tonic stretch of the muscle. This inhibition occurs in the jaw-opening phase of the masticatory cycle and is also seen in the pattern of activity recorded near the central terminals of these afferents. However, in recordings made near the terminals (traces at bottom of the figure), a phasic excitation is seen in the opposite phase (jaw closing). This excitation is thought to reflect spikes generated from primary afferent depolarization (PAD) occurring at the terminals and propagated antidromically (red bars). (2) Presynaptic GABAergic synapses on the afferent terminals presumably induced PADs. The antidromic spikes elicited at the terminals sometimes reach rostral branches innervating the motor nucleus (NVmot), but never reach the soma. (3) Later work showed that there were also GABAergic synapses on the main axonal trunk that could produce a shunt sufficient to cause failure of the antidromic spikes to propagate, but not enough to block propagation of orthodromic spikes.
Effects of chronic pain on motor systems Another sensory input that has profound effects on motor programs comes from nociceptors within the moving body part (e.g., see modulation of jawopening reflex above when elicited by nociceptive inputs). However, when pain becomes chronic, its effects on muscles and motor programs seem to change. Since the start of the twentieth century, it has been generally believed that chronic nociceptor activation cause muscle hyperactivity leading to abnormal movement patterns, spasms, and fatigue, which would lead to even more pain. It
was proposed that this perpetual pain machine, called the “Vicious Cycle” by Travell et al. (1942), underlies most chronic muscle pain conditions, including fibromyalgia, chronic lower back pain, tension-type headache, and temporomandibular dysfunction. We began to test the predictions of the Vicious Cycle hypothesis in the late 1980s and presented our findings at XIIth Symposium of the Groupe de recherche sur le système nerveux central (Lund et al., 1991). We analyzed data from studies that compared motor function of groups of controls to subjects suffering from the four chronic pain conditions listed above, plus
225
delayed onset muscle soreness. The data showed clearly the resting or postural activity of sore muscles was no greater than control levels, while the activity of agonist muscles went down, not up, during forceful contraction. However, there was a slight but significant increase in antagonist activity in some reports. The effect of these changes in motor output produced maximum voluntary contraction levels that were significantly lower in the patient groups, while the velocity and amplitude of movements went down. Clearly, this could not result from the action of a Vicious Cycle, so we proposed a new way of viewing the interaction between chronic pain and motor systems, which we named the pain-adaptation model. We and others then tested the pain-adaptation model by inducing tonic pain in normal human subjects and animals. We showed that infusions of painful hypertonic saline into the masseter muscle of normal human volunteers caused both tonic pain and a reduction of jaw-opening reflex amplitude (Lund et al., 1981), and that the same stimulus did not cause resting hyperactivity (Stohler et al., 1996). We then began to study the influence of tonic pain on CPGs by applying noxious pressure to the zygoma of decerebrate rabbits on mastication induced by stimulation of the corticobulbar tracts (Schwartz and Lund, 1995). Cycle duration was significantly increased, while the amplitude and velocity of movement went down. Arendt-Nielsen et al. (1996) found that painful infusions of hypertonic saline into the dorsal paraspinal muscles of men caused a dramatic increase in the amount of EMG activity recorded during the swing phase of locomotion, the phase in which the muscles act as antagonists and when they are usually almost silent. The EMG pattern caused by tonic nociceptor stimulation mimics the one that they recorded from chronic lower back pain patients. In order to study more directly the effects of tonic nociceptor stimulation on a CPG, we infused hypertonic saline into the masseters during fictive mastication in a decerebrate rabbit while we recorded from the digastric
motoneuron pool and from phasically active neurons in the rostral parvocellular reticular formation and adjacent rostral Vth spinal nucleus (Westberg et al., 1997). Nociceptor simulation caused a significant reduction in the area of digastric motoneuron bursts and increased the interburst interval and cycle duration. The firing pattern of the interneurons also changed significantly during nociceptor stimulation, even though only one had nociceptive sensory field. Furthermore, there was a strong temporal relationship between the change in peak firing of the interneuronal population and digastric burst termination. Thus it seems likely that tonic noxious inputs have general effects on CPGs that tend to lengthen cycle and phase duration, reduce agonist bursts, and increase antagonist activity. As we first suggested at the XIIth symposium, the effect of painful inputs on motor patterns such as mastication and locomotion is therefore not vicious but adaptive, because these changes seem to diminish the probability of self-induced tissue damage.
Do muscle spindle afferents contribute to chronic muscle pain? During our in vitro studies of masseter muscle spindle mechanoreceptors (Verdier et al., 2004), we noticed that most of them showed high-frequency membrane potential oscillations at resting membrane potentials above 53 mV. As the neurons were gradually depolarized by current injections, the oscillations increased in amplitude until spikes appeared on their peaks. Similar behavior had already been described in large DRG neurons, and Amir et al. (1999) had shown that the amplitude of the oscillations and spontaneous firing increased in nerve ligation models of neuropathic pain. We decided to find out if a similar change in phenotype occurred in muscle spindle afferents in a rat model of chronic muscle pain. Acidic saline injections into rat leg muscles have been used to induce long-lasting allodynia,
226
and we adapted this technique to the rat masseter. Responses to light pressure applied to the muscle of awake animals were above control levels (pH 7.2) in animals injected with saline at pH 4.0 for more than 5 weeks postinjection. In other animals, we showed that the expression of the early gene c-fos by the cell bodies of the muscle afferents in the Vth mesencephalic nucleus was also increased 1–4 days after the injections. We also used in vitro whole cell recording to show that the electrical properties and firing of spindle afferent somata were also radically modified by intramuscular acidic saline injections. Like the change in behavior, these were still present 5 weeks after injection. Neurons from the acidtreated group operated at more hyperpolarized potentials: resting membrane potentials were lower, as were thresholds for the appearance of the high-frequency oscillations and firing. The amplitude of the oscillations at a given membrane potential was higher in acid-treated animals, and spontaneous firing sometimes occurred (Lund et al., 2010). Sensory afferents fire orthodromically when the end organs are stimulated and antidromically when their central terminals are strongly depolarized, and in principle, firing is rarely initiated at the cell body level. Firing is considered ectopic when initiated in an abnormal zone or for an abnormal reason. Ectopic spikes initiated in the cell body can travel down the stem axon to the primary branch point, from which they can propagate in both directions. The central effects of ectopic orthodromic spikes arising from a neuroma or from dorsal root somata have been studied extensively in neuropathic pain models. Long-term changes occur in the dorsal horn, including the invasion of nociceptor territories by the terminals of largediameter afferents (Woolf et al., 1995); however, no one to date has considered the interactions that could occur in the periphery. Muscle spindles are complex sense organs that contain two types of mechanoreceptor terminals, the annulospiral endings of the group Ia afferents and the secondary endings supplied by group II
afferents. Both use glutamate as a central transmitter, but they also release glutamate from their peripheral terminals (Bewick et al., 2005). Annulospiral endings are filled with glutamatecontaining vesicles, and Pang et al. (2006) showed that they contain high levels of the glutamate transporter VGLUT1. Since they also contain some fine fibers, we decided to carry out an additional immunohistological study to see if these could be glutamate-sensitive nociceptors. We were able to show that fine afferent fibers could be often seen adjacent to VGLUT1-containing annulospiral endings. Many of these expressed the metabotropic glutamate receptor mGluR5 together with markers that are often associated with nociceptors, particularly P2X3, TRPV1, and Substance P (Lund et al., 2010). These results lead to several conclusions: first, changes in the phenotype of mechanoreceptor afferents may occur in the so-called functional pain states (fibromyalgia, myopathic pain, chronic lower back pain, irritable bowel syndrome, TMD, etc.), and not only in neuropathic pain. Second, ectopic firing in primary afferents may increase spontaneous pain and allodynia through actions in the periphery in addition to actions on pain relays within the dorsal horn (both spinal and trigeminal). Third, antidromic firing could activate nociceptors contained within complex sensory endings such as muscle spindles and Meissner's corpuscles (Paré et al., 2001) through an increase in the release of glutamate. Conclusions Very sadly, James P. Lund did not live long enough to write the conclusions of this chapter which will be his last contribution to the field. He left us with a great heritage of knowledge that brings us much closer to understanding how the masticatory CPG functions and how it is influenced by sensory inputs. A reading of this short summary of his work gives an idea of the breath of topics and subjects that he addressed.
227
There is not one aspect of the masticatory system that he has not touched, be it the role of the cortex and of different types of primary afferents or the modulation of motoneurons, interneurons, or reflexes. References Adrian, E. D., & Buytendijk, F. J. (1931). Potential changes in the isolated brain stem of the goldfish. Journal de Physiologie, 71, 121–135. Amir, R., Michaelis, M., & Devor, M. (1999). Membrane potential oscillations in dorsal root ganglion neurons: Role in normal electrogenesis and neuropathic pain. The Journal of Neuroscience, 19, 8589–8596. Appenteng, K., Lund, J. P., & Seguin, J. J. (1982). Intraoral mechanoreceptor activity during jaw movement in the anesthetized rabbit. Journal of Neurophysiology, 48, 27–37. Arendt-Nielsen, L., Graven-Nielsen, T., Svarrer, H., & Svensson, P. (1996). The influence of low back pain on muscle activity and coordination during gait: A clinical and experimental study. Pain, 64, 231–240. Baev, K. V. (1980). Polarization of primary afferent terminals in the lumbar spinal cord during fictitious locomotion. Neurophysiology, 12, 305–311. Bewick, G. S., Reid, B., Richardson, C., & Banks, R. W. (2005). Autogenic modulation of mechanoreceptor excitability by glutamate release from synaptic-like vesicles: Evidence from the rat muscle spindle primary sensory ending. Journal de Physiologie, 562, 381–394. Bullock, T. H. (1961). The origins of patterned nervous discharge. Behaviour, 17, 48–59. Byers, M. R. (1985). Sensory innervation of periodontal ligament of rat molars consists of unencapsulated Ruffini-like mechanoreceptors and free nerve endings. The Journal of Comparative Neurology, 231, 500–518. Dellow, P. G., & Lund, J. P. (1971). Evidence for central timing of rhythmical mastication. Journal de Physiologie, 215, 1–13. Dubuc, R., Cabelguen, J.-M., & Rossignol, S. (1988). Rhythmic fluctuations of dorsal root potentials and antidromic discharges of primary afferents during fictive locomotion in the cat. Journal of Neurophysiology, 60, 2014–2036. Eccles, J. C., Kozak, W., & Magni, F. (1961). Dorsal root reflexes of muscle group I afferent fibres. Journal de Physiologie, 159, 128–146. Gordon, T., Thomas, C. K., Munson, J. B., & Stein, R. B. (2004). The resilience of the size principle in the organization of motor unit properties in normal and reinnervated adult skeletal muscles. Canadian Journal of Physiology and Pharmacology, 82, 645–661.
Eccles, J. C., Schmidt, R. F., & Willis, W. D. (1962). Presynaptic inhibition of the spinal monosynaptic reflex pathway. Journal de Physiologie, 161, 282–297. Gossard, J. P., Cabelguen, J. M., & Rossignol, S. (1989). Intraaxonal recordings of cutaneous primary afferents during fictive locomotion in the cat. Journal of Neurophysiology, 62, 1177–1188. Gossard, J.-P., Cabelguen, J.-M., & Rossignol, S. (1991). An intracellular study of muscle primary afferents during fictive locomotion in the cat. Journal of Neurophysiology, 65, 914–926. Henneman, E., Somjen, G., & Carpenter, D. O. (1965). Functional significance of cell size in spinal motoneurons. Journal of Neurophysiology, 28, 560–580. Horridge, G. A. (1968). Interneurons: Their origin, action, specificity, growth and plasticity. London: W.H. Freeman and Co. 436pp. Kolta, A., Lund, J. P., & Rossignol, S. (1990). Modulation of activity of spindle afferents recorded in trigeminal mesencephalic nucleus of rabbit during fictive mastication. Journal of Neurophysiology, 64, 1067–1076. Kolta, A., Lund, J. P., Westberg, K.-G., & Clavelou, P. (1995). Do muscle-spindle afferents act as interneurons during mastication? Trends in Neurosciences, 18, 441. Komuro, A., Morimoto, T., Iwata, K., Inoue, T., Masuda, Y., Kato, T., et al. (2001). Putative feed-forward control of jaw-closing muscle activity during rhythmic jaw movements in the anesthetized rabbit. Journal of Neurophysiology, 86, 2834–2844. Lavigne, G., Kim, J. S., Valiquette, C., & Lund, J. P. (1987). Evidence that periodontal pressoreceptors provide positive feedback to jaw closing muscles during mastication. Journal of Neurophysiology, 58, 342–358. Lund, J. P., & Dellow, P. G. (1973). Rhythmical masticatory activity of hypoglossal motoneurons responding to an oral stimulus. Experimental Neurology, 40, 243–246. Lund, J. P., Donga, R., Widmer, C. G., & Stohler, C. S. (1991). The pain-adaptation model: A discussion of the relationship between chronic musculoskeletal pain and motor activity. Canadian Journal of Physiology and Pharmacology, 69, 683–694. Lund, J. P., Richmond, F. J., Touloumis, C., Patry, Y., & Lamarre, Y. (1978). The distribution of Golgi tendon organs and muscle spindles in masseter and temporalis muscles of the cat. Neuroscience, 3, 259–270. Lund, J. P., Rossignol, S., & Murakami, T. (1981). Interactions between the jaw-opening reflex and mastication. Canadian Journal of Physiology and Pharmacology, 59, 683–690. Lund, J. P., Sadeghi, S., Athanassiadis, T., Caram Salas, N., Auclair, F., Thivierge, B., et al. (2010). Assessment of the potential role of muscle spindle mechanoreceptor afferents in chronic muscle pain in the rat masseter muscle. PLoS ONE, 5, e11131.
228 Lund, J. P., Sasamoto, K., Murakami, T., & Olsson, K.Å. (1984). Analysis of rhythmical jaw movements produced by electrical stimulation of motor-sensory cortex of rabbits. Journal of Neurophysiology, 52, 1014–1029. Lund, J. P., Smith, A. M., Sessle, B. J., & Murakami, T. (1979). Activity of trigeminal alpha- and gamma-motoneurons and muscle afferents during performance of a biting task. Journal of Neurophysiology, 42, 710–725. Morton, D. W., & Chiel, H. J. (1994). Neural architectures for adaptive behavior. Trends in Neurosciences, 17, 413–420. Olsson, K.Å., Sasamoto, K., & Lund, J. P. (1986). Modulation of transmission in rostral trigeminal sensory nuclei during chewing. Journal of Neurophysiology, 55, 56–75. Pang, Y. W., Li, J. L., Nakamura, K., Wu, S., Kaneko, T., & Mizuno, N. (2006). Expression of vesicular glutamate transporter 1 immunoreactivity in peripheral and central endings of trigeminal mesencephalic nucleus neurons in the rat. The Journal of Comparative Neurology, 498, 129–141. Paré, M., Elde, R., Mazurkiewicz, J. E., Smith, A. M., & Rice, F. L. (2001). The Meissner corpuscle revised: A multiafferented mechanoreceptor with nociceptor immunochemical properties. The Journal of Neuroscience, 21, 7236–7246. Pearson, K. G. (2008). Role of sensory feedback in the control of stance duration in walking cats. Brain Research Reviews, 57, 222–227. Peyron, M. A., Lassauzay, C., & Woda, A. (2002). Effects of increased hardness on jaw movement and muscle activity during chewing of visco-elastic model foods. Experimental Brain Research, 142, 41–51. Prochazka, A., Westerman, R. A., & Ziccone, S. P. (1977). Ia afferent activity during a variety of voluntary movements in the cat. Journal de Physiologie, 268, 423–448. Prochazka, A., Hulliger, M., Zangger, P., & Appenteng, K. (1985). ‘Fusimotor set’: new evidence for alpha-independent control of gamma-motoneurones during movement in the awake cat. Brain Research, 339(1), 136–140. Rudomin, P., & Schmidt, R. F. (1999). Presynaptic inhibition in the vertebrate spinal cord revisited. Experimental Brain Research, 129, 1–37. Schwartz, G., & Lund, J. P. (1995). Modification of rhythmical jaw movements by noxious pressure applied to the periosteum of the zygoma in decerebrate rabbits. Pain, 63, 153–161. Sherrington, C. S. (1917). Reflexes elicitable in the cat from pinna vibrissae and jaws. Journal de Physiologie, 51, 404–431. Stein, R. B. (1974). The peripheral control of movement. Physiological Reviews, 54, 215–243. Stohler, C. S., Zhang, X., & Lund, J. P. (1996). The effect of experimental jaw muscle pain on postural muscle activity. Pain, 66, 215–221.
Travell, J., Rinzler, S., & Herman, M. (1942). Pain and disability of the shoulder and arm. Journal of the American Medical Association, 120, 417–422. Trulsson, M. (2007). Force encoding by human periodontal mechanoreceptors during mastication. Archives of Oral Biology, 52, 357–360. Tsuboi, A., Kolta, A., Chen, C. C., & Lund, J. P. (2003). Neurons of the trigeminal main sensory nucleus participate in the generation of rhythmic motor patterns. The European Journal of Neuroscience, 17, 229–238. Verdier, D., Lund, J. P., & Kolta, A. (2003). GABAergic control of action potential propagation along axonal branches of mammalian sensory neurons. The Journal of Neuroscience, 23, 2002–2007. Verdier, D., Lund, J. P., & Kolta, A. (2004). Synaptic inputs to trigeminal primary afferent neurons cause firing and modulate intrinsic oscillatory activity. Journal of Neurophysiology, 92, 2444–2455. Westberg, K., Clavelou, P., Sandström, G., & Lund, J. P. (1998). Evidence that trigeminal brainstem interneurons form subpopulations to produce different forms of mastication in the rabbit. The Journal of Neuroscience, 15, 6466–6479. Westberg, K.-G., Clavelou, P., Schwartz, G., & Lund, J. P. (1997). Effects of chemical stimulation of masseter muscle nociceptors on trigeminal motoneuron and interneuron activities during fictive mastication in the rabbit. Pain, 73, 295–308. Westberg, K.-G., Kolta, A., Clavelou, P., Sandström, G., & Lund, J. P. (2000). Evidence for functional compartmentalization of trigeminal muscle spindle afferents during fictive mastication in the rabbit. The European Journal of Neuroscience, 12, 1145–1154. Wilson, D. M. (1961). The central nervous control of flight in a locust. The Journal of Experimental Biology, 38, 471–490. Woda, A., Foster, K., Mishellany, A., & Peyron, M. A. (2006). Adaptation of healthy mastication to factors pertaining to the individual or to the food. Physiology and Behavior, 89, 28–35. Woolf, C. J., Shortland, P., Reynolds, M., Ridings, J., Doubell, T., & Coggeshall, R. E. (1995). Reorganization of central terminals of myelinated primary afferents in the rat dorsal horn following peripheral axotomy. The Journal of Comparative Neurology, 360, 121–134. Wyman, R. (1965). Probabilistic characterization of simultaneous nerve impulse sequences controlling dipteran flight. Biophysical Journal, 5, 447–471.
Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) Progress in Brain Research, Vol. 188 ISSN: 0079-6123 Copyright Ó 2011 Elsevier B.V. All rights reserved.
CHAPTER 16
Spinal plasticity in the recovery of locomotion Serge Rossignol*, Alain Frigon, Grégory Barrière, Marina Martinez, Dorothy Barthélemy, Laurent Bouyer, Marc Bélanger, Janyne Provencher, Connie Chau, Edna Brustein, Hugues Barbeau, Nathalie Giroux, Judith Marcoux, Cécile Langlet and Olivier Alluin Groupe de Recherche sur le Système Nerveux Central, Department of Physiology, Faculty of Medicine, Université de Montréal, Montreal and, Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury (CIHR), Station Centre-Ville Montréal, Québec, Canada
Abstract: Locomotion is a very robust motor pattern which can be optimized after different types of lesions to the central and/or peripheral nervous system. This implies that several plastic mechanisms are at play to re-express locomotion after such lesions. Here, we review some of the key observations that helped identify some of these plastic mechanisms. At the core of this plasticity is the existence of a spinal central pattern generator (CPG) which is responsible for hindlimb locomotion as observed after a complete spinal cord section. However, normally, the CPG pattern is adapted by sensory inputs to take the environment into account and by supraspinal inputs in the context of goal-directed locomotion. We therefore also review some of the sensory and supraspinal mechanisms involved in the recovery of locomotion after partial spinal injury. We particularly stress a recent development using a dual spinal lesion paradigm in which a first partial spinal lesion is made which is then followed, some weeks later, by a complete spinalization. The results show that the spinal cord below the spinalization has been changed by the initial partial lesion suggesting that, in the recovery of locomotion after partial spinal lesion, plastic mechanisms within the spinal cord itself are very important. Keywords: locomotion; plasticity; spinal cord injury; reflex; rehabilitation; training.
controls by the central nervous system (CNS) of a robust and fundamental primitive motor act. I (S.R.) personally became fascinated by the subject of locomotion when I saw a film shown by Sten Grillner of a cat that had been previously completely spinalized at the low thoracic level as a kitten (Grillner, 1973). This cat could step with
Introduction The study of locomotion offers several opportunities to investigate the various levels of
*Corresponding author. Tel.: þ1-514-343-6371; Fax: þ1-514-343-6113 DOI: 10.1016/B978-0-444-53825-3.00021-8
229
230
the hindlimbs on a small nature trail outside the lab while a research assistant held its tail to partially support the weight of its hindquarters and provide some lateral balance. Even more surprising, the spontaneous forward walking movements of the intact forelimbs were at times strong enough that the hindquarters would rise and the cat would make a few unaided steps with the hindlimbs before losing balance. Although there was a long history of the locomotor capabilities of spinal animals in many species (as well summarized by Grillner, 1981), this provided a model which could be investigated with modern tools of electrophysiology and which obviously provided a scientific framework that also had a great potential impact on spinal cord injured (SCI) patients. These observations were clearly showing that the spinal cord below a complete spinal section was capable of generating the basic pattern of locomotion with even some elaborate timing details. Therefore rehabilitation after SCI should strive to maintain or activate the sublesional spinal circuits. The concept of spinal generation of locomotion is robust (Delcomyn, 1980; Rossignol, 1995, 1996; Rossignol et al., 2000, 2002) and relevant even for other animal species such as the rat (Courtine et al., 2009; Gimenez y Ribotta et al., 2000), the mouse (Leblond et al., 2003), and humans (Bussel et al., 1988; Calancie, 2006; Dietz and Harkema, 2004; Gerasimenko et al., 2010; Harkema, 2008; Nadeau et al., 2010). The importance of spinal generation of locomotion was strongly revived more recently when we observed that, even after partial SCI (the most common lesion in humans), the recovery of hindlimb locomotion also depends to a great extent on changes that have occurred in the spinal circuits below the SCI (Barrière et al., 2008, 2010; Rossignol et al., 2009). This observation is also of clinical importance since it emphasizes the possibility of profoundly modifying the spinal cord through rehabilitation strategies in humans after SCI. The aim of the present paper is to link various observations made over the years that lead to such conclusions.
Generation of spinal locomotion Undoubtedly, the original observations on the generation of locomotion in spinal kittens were seminal (Forssberg et al., 1980a,b; Grillner, 1973). They established the very important concept that the spinal circuitry for generating locomotion was inborn (genetically determined) and that kittens could produce walking movements 1–2 days after spinalization without having had to “learn” walking. More detailed studies using electromyographic recordings (EMG) showed that several features of the muscle discharge resembled those seen in the hindlimbs of normal cats walking on the treadmill (Forssberg et al., 1980a). Other detailed features of the EMGs in spinal kittens were of interest. For instance, the recruitment of muscle discharges in extensors was quite similar to that of the intact cat with a gradual build-up of activity in knee extensors and a more abrupt onset of activity in ankle extensors. One important feature was that the discharge of ankle extensors occurred prior to foot contact, as was also seen during walking in intact cats (Engberg and Lundberg, 1969). This suggested that the onset of activity in these muscles was not triggered by foot contact that would subsequently trigger a chain of spinal reflexes but was pre-programmed centrally. A series of papers on the effects of DOPA on the spinal cord suggested that indeed central alternating rhythms in flexor and extensor muscle nerves could be evoked by stimulating peripheral nerves in otherwise curarized animals (Jankowska et al., 1967a,b; Stuart and Hultborn, 2008). The concept of central generation of locomotion was conclusively established in curarized and acutely spinalized adult cats showing that l-DOPA could evoke complex spinal rhythms that had many of the timing characteristics found in cats walking over ground or on a treadmill (Grillner and Zangger, 1979). Similar findings were made in chronic spinal cats that had been trained to walk on a treadmill for several weeks on a treadmill (Pearson and Rossignol, 1991). A simple perineal
231
stimulation in these previously trained spinal animals could evoke bilaterally organized rhythms in both hindlimbs even after curarization. In spinal kittens, the absolute amplitude values of muscle discharge could not be directly compared with those obtained in intact conditions because there was no control period as such and muscles were recorded through acutely inserted electrodes for each session so that variability of electrode position could have affected their discharge profile. In adult chronic spinal cats implanted with EMG electrodes, it was possible to evaluate muscle discharges and compare the activity before and after SCI in the same cat (Belanger et al., 1996; Chau et al., 1998a; Frigon and Rossignol, 2008; Rossignol, 1996). Although some changes in amplitude could be observed such as an overall reduction in limb extensors and an increase in limb flexors, the general pattern was overwhelmingly similar to the intact condition in the same cats taken as their own control. Of interest was the coupling of onset of activity between some muscles, especially flexor muscles. It was observed that, whereas in the intact state the knee flexor Semitendinosus was activated before foot lift and that the hip flexor Sartorius was activated after foot lift, these two muscles often tended to discharge almost synchronously after SCI whereas the ankle flexor Tibialis Anterior was activated earlier. This led in some spinal cats (especially early after spinalization) to a foot drag in the first part of swing (equivalent of foot drop in humans) because the ankle flexed prior to the foot being lifted by the knee flexors. Altogether these observations confirm that the spinal cord can centrally generate hindlimb locomotion. Much of this conclusion derives from a long history of observations dating back to Sherrington (1910) and that was well-recognized and summarized before (Grillner, 1981; Rossignol, 1996; Rossignol et al., 2006; Stuart and Hultborn, 2008). It is important to remember some of these findings to better understand how other work on spinal locomotion has progressed in later years. Indeed,
having clearly established that the lumbosacral spinal cord of adult cats can generate hindlimb locomotion it was important to understand (1) how this spinal pattern could be modulated by various neurotransmitters and sensory inputs; (2) which spinal segments and which descending pathways were important in triggering and/or modulating this spinal rhythm; (3) if lumbosacral circuits are important when the spinal cord is only partially damaged as is often the case in humans, and (4) if all this could really matter for humans with SCI.
Modulation of spinal locomotion Neurotransmitter modulation As mentioned above, early work using the noradrenaline precursor L-DOPA in acute spinal cats (Jankowska et al., 1967a,b) led to the concept of a central pattern generator (CPG) for locomotion (Grillner and Zangger, 1979). This seminal work triggered other research to determine which neurotransmitter systems and which receptors on which these can act could trigger and/or modulate the locomotor pattern.
Noradrenergic mechanisms The noradrenergic system was mostly studied in this respect. First, agonists of different subtypes of adrenergic receptors were used. Clonidine, an alpha-2 noradrenergic agonist can trigger a well-developed bilateral hindlimb walking pattern on a treadmill within minutes after the injection (i.p. or i.v.) in acutely spinalized adult cats demonstrating that specific stimulation of a noradrenergic alpha-2 receptors activates an already extant spinal locomotor circuit (Forssberg and Grillner, 1973). Later, we found that, in adult spinal cats chronically implanted with EMG electrodes, only the alpha-2 adrenergic agonists such as clonidine injected intraperitoneally (Barbeau et al., 1987) or other alpha-2 agonists
232
(tizanidine, oxymethazoline) injected intrathecally (Chau et al., 1998b) could induce locomotion in the early days following spinalization. After some weeks, when complete spinal cats had recovered spontaneous locomotion, clonidine still exerted potent effects on the spontaneously generated locomotor pattern by increasing EMG burst duration and overall step length. The effects of clonidine differ whether the cats have an intact spinal cord or whether they have a complete or partial spinal lesion (Rossignol et al., 1998, 2001). In the intact state, intrathecal injection of clonidine exerts little effect but, in the same cat, early after spinal section, it has the striking effect of evoking locomotion as described above (Giroux et al., 2001). Moreover, in cats with a large ventrolateral lesion, clonidine can decrease weight support and in the worst case can stop voluntary quadrupedal locomotion (Brustein and Rossignol, 1999). In conclusion then, the state of pre- and postsynaptic receptors may differ in different preparation and determine the effects of the neurotransmitter agonists. This is important when assessing drugs in humans since the excitability state of the receptors is unknown (Remy-Neris et al., 1999). Yohimbine, an alpha-2 adrenergic blocker, reverses the effect of clonidine on the initiation of locomotion or on the clonidine-induced change in the step cycle (Barbeau et al., 1987; Giroux et al., 2001). However, yohimbine has no effect in the chronic spinal cat trained to walk on a treadmill. This might appear obvious since the neurotransmitter is no longer present following spinalization but it is important to block these receptors to show that residual noradrenaline and other molecules that could potentially activate these receptors are not responsible for the ability of the cat to walk. However, in the intact cat, yohimbine reduced the correct coordination between the fore- and hindlimbs with the trunk often bending on one side or the other (Giroux et al., 2001) suggesting that in normal locomotion, noradrenergic neurotransmission is important for interlimb coordination (McDearmid et al., 1997).
Serotoninergic mechanisms In acute spinal cat, 5-HT2 serotoninergic agonists such as quipazine, 5-O-DMT, or the precursor 5-HTP do not initiate locomotion (Barbeau and Rossignol, 1990). This might be related to the level of the spinal section as suggested by others (Schmidt and Jordan, 2000) on the basis of the segmental distribution of 5-HT receptor subtypes that may be important for locomotion (Noga et al., 2009). Therefore, the level of spinal section should be taken into account in evaluating the effects of stimulation by certain drugs since specific subclasses of receptors may preferentially be distributed above or below the spinal lesion. Although 5-HT agonists do not initiate locomotion in acute spinal cats, they markedly alter the output amplitude of activity of hindlimb muscles (especially extensors) and paraxial muscles (Barbeau and Rossignol, 1990). In cats with bilateral ventrolateral spinal lesions, 5-HT agonists increased weight support as well as the ability of the cats to walk uninterruptedly (Brustein and Rossignol, 1999). The pharmacological effects of 5-HT agonists reinforced the voluntarily generated locomotor pattern. This is important because it shows that “voluntary” locomotion can be improved by increasing spinal excitability provided by the pharmacological stimulation. This obviously is of interest within the context of rehabilitation since one could imagine that a temporary state of enhanced excitability of the spinal cord could be induced to facilitate the expression of locomotion and therefore facilitate or accelerate the beneficial outcome of locomotor training. Although there have been some differences between the response of cats and rats to 5-HT agonists, as far as initiation of locomotion is concerned, there is no doubt that similarities on the output pattern by the spinal cord predominate. Using grafts of embryonic mesencephalic 5-HT cells below a complete spinal lesion in adult rats, we showed that rats, which cannot usually express a spontaneous locomotor pattern of the hindlimbs after spinalization, could express a full pattern of
233
hindlimb locomotion (Gimenez y Ribotta et al., 2000). Later studies also showed a beneficial effect on the expression of locomotion with chronic 5-HT injections in adult spinal rats (Antri et al., 2005). A recurrent question is the mechanism through which 5-HT could act to improve the recovery of spinal excitability and locomotion and even perhaps spasticity. The first mechanism that was described is that of hypersensitivity of 5-HT receptors (Bédard et al., 1979) in which similar doses of 5-HT agonists could exert progressively larger effects. Besides such denervation sensitivity, some remnant 5-HT fibers have been described that may originate from endogenous sources (Takeoka et al., 2009). However, other mechanisms may be at play. It has been shown that some membrane properties of spinal motoneurones such as dendritic persistent inward currents (PICs) may be important in the recovery of spinal cord excitability and its ability to express locomotion (Heckmann et al., 2005; Hultborn et al., 2004). In fact, after acute spinalization, PICs are lost and can be reinstated by 5-HT agonists and also noradrenaline (Conway et al., 1988; Hounsgaard et al., 1988). One could then infer that remnant 5-HT after spinalization could participate in the recovery of spinal excitability and locomotion (Harvey et al., 2006a,b; Li et al., 2007). Even more fascinating is the possibility that constitutive 5-HT receptors may become a key mechanism for recovery of excitability and even spasticity as suggested (Murray et al., 2010). In that context it is of interest that cyproheptadine, a 5-HT blocker, had a potent inhibitory effect on spinal locomotion in spinal cats after injection of quipazine (Barbeau and Rossignol, 1990, 1991).
Other neurotransmitters Intrathecal injections of NMDA, contrary to in vitro neonatal rats (Cazalets et al., 1992; Kiehn and Kjaerulff, 1996) or lampreys (Grillner et al., 1981)
and in contrast to decerebrate cats (Douglas et al., 1993), did not induce locomotion in adult spinal cats even though they markedly increased excitability as evidenced from hindfeet tremor and toe fanning (Chau et al., 2002; Giroux et al., 2003). In cats that just started to generate small steps (around 6–7 days) after spinalization, NMDA could boost the expression of emergent locomotor patterns for several hours (Chau et al., 2002). When the spontaneous recovered locomotor pattern was expressed several weeks after spinalization, NMDA had only little effects. In the intact cat AP-5, an NMDA blocker, was shown to influence locomotion by reducing weight support which was, however, rapidly compensated. The same drug injected in the same cat but after spinalization completely blocked the spontaneously generated locomotion (Giroux et al., 2003). This suggests that NMDA receptors play a critical role in maintaining spinal locomotion perhaps by potentializing the release of glutamate by afferents. In summary then, agonists or antagonists of various transmitters illustrate how neurotransmitters produced by cells in the brain stem, and more or less absent after spinalization, can induce changes in membrane properties and circuits within the spinal cord and enhance the expression of locomotor circuits that are already present. Recent work illustrating how 5-HT agonists can facilitate the expression of locomotion in chronic spinal rats (Courtine et al., 2009) should be interpreted in this light. The circuits already exist in the cord but because the monoaminergic and indoleaminergic inputs are severed, cell properties of these circuits are deficient. Providing exogenous neurotransmitters or agonists of their receptors could favor the spontaneous regain of function by promoting changes in membrane properties necessary for the behavior. In this context, it is also of great importance to better document changes in receptors, not only in their properties but also their number and localization (Chau et al., 2001; Giroux et al., 1999) as these may vary with time after various types of lesions.
234
Sensory modulation The field of sensorimotor interactions during locomotion has been reviewed several times and more specifically in Rossignol et al. (2006). The details of the observations will not be reviewed but only broad principles that apply to the modulation of the spinal circuits generating locomotion. The complexity of sensorimotor interactions has been well expressed earlier in clear terms: “Normally there is an interaction between the periphery and the central generator and presumably the former is of great importance although the basic structures of the cycle is laid down centrally” (Grillner, 1973). Tonic sensory stimuli can modify the state of the locomotor circuits. Thus a spinal cat injected with clonidine may appear motionless on a moving treadmill but an unspecific perineal stimulation can trigger a well-coordinated pattern of hindlimb locomotion (Barbeau et al., 1987; Belanger et al., 1996). A pinch of the skin in the dorsal lumbosacral region can immediately stop locomotion in spinal cats much as what was shown for spinal rabbits (Viala et al., 1978). Similarly, unspecific electrical stimulation of the dorsal portion of the cord or the dorsal roots can elicit locomotion in spinal cats and rats (Barthélemy et al., 2006, 2007; Courtine et al., 2009; Gerasimenko et al., 2007; Ichiyama et al., 2005). Tonic stimulation can also define the operating range of operation of the spinal circuits and inhibit their expression. The power of proprioceptive inputs especially at the hip joint was first studied in chronic spinal kittens in which it was demonstrated that hip flexion on one side stopped the walking movements on that side (Rossignol et al., 1975) while the other limb continued walking. When the hip reached a critical angle of hip extension the limb started to walk in alternation with the contralateral hindlimb. Hindlimb fictive locomotion in cats can be blocked completely by flexing the hip (Pearson and Rossignol, 1991) on one side. Similarly, the overall pattern can be biased toward flexor or extensor outputs by various degrees of hip
extension. In the forelimbs, protraction or retraction of the shoulder can change the actual structure of the locomotor pattern on the manipulated side and compensatory changes in the other forelimb (Saltiel and Rossignol, 2004). The study of phasic sensory also revealed important principles namely that reflexes are modulated in a phase-dependent manner. Initial observations in spinal kittens showed that a tap to the dorsum of the foot in various phases of the step cycle evoked short latency flexion responses during swing and short latency excitation in extensors when applied during stance (Forssberg et al., 1975). This reflex reversal made sense because the sensory evoked responses assisted the ongoing specific phase of the locomotor pattern. Using a very precise cutaneous nerve stimulation in the same cat, it was possible to essentially confirm the finding that the reflex responses were dramatically changed after spinalization (Frigon and Rossignol, 2008) and that “new” reflex responses could be evoked namely short latency extensor responses. How are reflexes controlled during dynamic rhythmic processes such as locomotion? Several mechanisms have been proposed (Frigon and Rossignol, 2006a,b; Rossignol et al., 2006). Besides the more obvious rhythmic modulation of interneuronal circuits, presynaptic mechanisms may play a crucial role. Cyclical modulation of afferent excitability has been demonstrated (Gossard et al., 1989; Rudomin et al., 1993) but also cyclical antidromic discharges in primary afferents were found (Beloozerova and Rossignol, 1999, 2004). Such antidromic discharges may also have a role in tritonia swimming (Sakurai and Katz, 2009) but also in mastication (Kolta et al., 1995). It is possible that multiple discharge sites in neurones could be implicated in regulation of sensorimotor interactions and participate in the reorganization of circuits after lesions by providing alternative modes of activation of the circuitry. The experimental approaches described above mainly served to investigate the role of sensory inputs by stimulating nerves or receptive fields.
235
Since it is known that a fictive locomotor pattern can be evoked centrally in cats being curarized or following a dorsal rhizotomy (Grillner and Zangger, 1974) afferent inputs are not critical for initiating the basic locomotor pattern but for adapting it. This is most probably the reason why locomotor training is so important in spinal cats which has otherwise lost all other inputs (Frigon and Rossignol, 2006b; Rossignol, 2006; Rossignol et al., 2006). Another observation showing the importance of sensory inputs for the correct execution of locomotor movement comes from studies on cutaneous denervation. If the cutaneous innervation of both hindfeet is severed by cutting all cutaneous nerves, the otherwise intact cat can walk rather well on the treadmill with proper foot contact but the same cats will be unable to place the feet correctly after spinalization (Bouyer and Rossignol, 2003a,b). This suggests that other sensory or descending pathways compensated for the cutaneous denervation when the cat was otherwise intact. The reduced spinal preparation which has a limited repertoire of other compensatory mechanisms shows how important the cutaneous inputs are. Other experiments have also shown that severing ankle muscles has minor effects on walking in intact cats (Carrier et al., 1997) but major effects in the same cat after spinalization. Similar findings were made after lesioning extensor nerves of the ankle in which a denervation performed in the intact state hampered significantly the recovery of locomotion after spinalization (Frigon and Rossignol, 2009). Finally, the importance of sensorimotor interactions in locomotion can be evaluated by plastic processes occurring in reflex pathways after locomotor training. It was indeed shown that locomotor training could “normalize” the amplitude of proprioceptive and cutaneous pathways after locomotor training (Côté and Gossard, 2004; Côté et al., 2003). This opens the way to other rehabilitation approaches in which the normalization of reflex pathways could help the recovery of locomotor function (Chen and Wolpaw, 2002; Frigon and Rossignol, 2006b; Wolpaw and Carp, 2006).
Segmental and suprasegmental control of locomotion The first section introduced the concept of a CPG while, in the second one, I summarized some of the observations on the modulation of this CPG by neurochemical substances or by activation of reflex pathways to mimic how this CPG could adapt to various environmental demands or states. How is this spinal locomotor pattern turned on and off or adapted for purposeful locomotion? Most of this question is outside the range of this review and has been well summarized previously (Armstrong, 1988; Drew et al., 2004; Grillner, 1981; Grillner et al., 2008a,b; Rossignol, 1996; Rossignol et al., 2006; Yakovenko and Drew, 2009). However, recent observations have re-emphasized the role of the CPG in the recovery of function after various types of spinal lesions and only these aspects will be reported here.
Segmental control Some early work suggested that intrathecal drug injections around rostral lumbar segments of the spinal cord in cats (at around L4) were very effective in triggering or modulating locomotion (Chau et al., 1998a,b; Giroux et al., 2001). Later, in cats spinalized 1 week earlier, an intraspinal injection of clonidine or yohimbine restricted to the L4 segment (i.e., rostral to the main hindlimb motoneuron pools; Vanderhorst and Holstege, 1997) initiated or blocked spinal locomotion (Marcoux and Rossignol, 2000). Experiments with serial complete sections were performed. Cats were spinalized at T13 and trained to recover locomotion of the hindlimbs on the treadmill. Thereafter, each cat was submitted to a second complete spinal section and their locomotor ability re-evaluated for several weeks (Langlet et al., 2005). Whereas a second lesion at L2 or rostral L3 did not prevent the locomotion that had recovered after the initial T13 spinalization, lesions at caudal L3 or L4 completely abolished locomotion
236
and the latter could not be reinstated even after 3–4 weeks of locomotor training. However, other rhythmic activities such as fast paw shake were not only present but enhanced showing the preserved rhythmogenic capability of the spinal cord. This also indicated that motoneurones of recorded muscles were undamaged by the lesion (because they were rhythmically active in fast paw shake) and that other neural elements important for triggering locomotion are located at or above L4. Finally, mid-lumbar segments may be important in evoking locomotion by other means. It was shown that intraspinal electrical microstimulation applied at different spinal levels could initiate locomotion in 1-week spinal cats (Barthélemy et al., 2005). After inactivating these mid-lumbar segments by yohimbine or performing a second lesion at caudal L4, locomotion could no longer be evoked by intraspinal electrical microstimulation at other more caudal lumbar levels (Barthélemy et al., 2007). We raised the question as to whether these segments were also important for locomotion in the decerebrate cat. Indeed, inactivation of L3–L4 by intraspinal yohimbine abolishes temporarily spontaneous locomotion (Delivet-Mongrain et al., 2008) and it is likely that these segments play an important role even when locomotion is triggered by brain stem pathways. Propriospinal pathways (Sherrington and Laslett, 1903) are important here as suggested by more recent experiments (Courtine et al., 2008, 2009; Cowley et al., 2008; Zaporozhets et al., 2006).
Suprasegmental control and the dual spinal lesion paradigm In the context of recovery of locomotion after lesions of the spinal cord, it is important to distinguish what part is played by supraspinal structures and by spinal structures respectively. Indeed, after partial spinal lesion, it is clear that regeneration or sprouting occur (Ghosh et al., 2009; Rossignol et al., 2007). One could envisage
that such descending inputs re-establish a new state of dynamic interactions between supraspinal structures and the spinal cord. However, we also showed that the spinal locomotor circuitry was very important for the recovery of hindlimb locomotion after a partial lesion of the spinal cord. This conclusion was reached using a double spinal lesion paradigm (Barrière et al., 2008, 2010). Firstly, a left hemisection was performed at thoracic levels (T10–T11) and cats were trained three to five times a week to walk on a treadmill until they reached a stable level of voluntary quadrupedal locomotor performance. The spinal cord was then completely sectioned at T13 (same level as in Barbeau and Rossignol, 1987; Belanger et al., 1996) thus removing all inputs from supraspinal fibers that were left intact by the previous partial lesion, as well as all fibers that may have regenerated or sprouted. Remarkably, within 24 h after spinalization, these cats could walk with the hindlimbs at high speeds and with bilateral plantar foot contact. This is in sharp contrast to the few weeks needed to reach such performance in cats without a prior partial lesion before spinalization, as previously mentioned (Barbeau and Rossignol, 1987). Therefore, during the period of locomotor training following the partial spinal lesion, plastic changes occurred in the lumbosacral spinal cord, creating a configuration of the locomotor circuitry that enabled it to operate without or with limited supraspinal influences. Mechanisms responsible for such plasticity below the lesion are still under study. The first question that may be raised is that of an unspecific priming effect of the first lesion on the subsequent spinalization. However, changes appear to have more specificity than that would imply. A recent paper described changes in cutaneous reflexes, evoked by stimulating cuff electrodes placed bilaterally around the superficial peroneal nerves innervating the dorsum of the foot (Frigon et al., 2009). After a partial spinal lesion at T10–T11, reflexes showed a marked asymmetry between the left and right hindlimbs. More specifically, the changes in short excitatory
237
(P1) responses could differ between hindlimbs. Remarkably, after complete spinalization, some reflex changes that were identifiable in the late stages of the partial spinal lesion could persist for a few days after complete section but then cutaneous reflexes on both sides increased more or less symmetrically. This provided one objective indication that changes had occurred in simple spinal circuits and that these changes could remain for some time before reflexes adapted again to the new, more symmetrical situation following complete spinal transection. Detailed analysis of the locomotor kinematics throughout the period of partial lesion and complete lesion showed several compensatory adjustments in footfall patterns, step cycle duration, duration of various sub-phases of the step cycle and interlimb coupling (Barrière et al., 2010). Even more interesting is the observation that some of the compensatory changes seen during the partial lesion period are actually reversed after a complete spinalization, a phenomenon akin to the Bechterew phenomenon seen after serial labyrinthectomy (Galiana et al., 1984). This implies that during the partial lesion period, covert changes occurred in the intrinsic spinal CPG that were revealed by the complete spinalization.
Conclusions This short review has per force concentrated mainly on some of the work performed in my group over the last 35 years. I cannot even begin to thank the numerous students, postdocs, assistants, and colleagues who have contributed to this work and the still exciting journey. They know and I know. I hope this mini-review conveys the excitement of discovering new things and rediscovering old things, of how concepts evolve and are revived by new observations, how old questions persist and continue to nag us. How is the spinal cord generating locomotion in mammals? How are simple circuits maintained while increasing their capabilities (Grillner and
Jessell, 2009)? Can we help patients with spinal cord injury regain useful function through this basic knowledge (Nadeau et al., 2010)? This symposium was meant to bring together scientists from the diverse fields of the neural control of breathing, chewing, and locomotion. The tradition in the field of motor control consisting in bringing scientists studying different types of movements in different species has been rich and productive and has spanned over several decades. Jim Lund, to whom this book is dedicated, was a firm believer in such an interdisciplinary approach to neuroscience and we will regret his friendly and enthusiastic input. Another nagging and simple question: do we know how mastication is controlled during locomotion while breathing (Lund et al., 1984)?
Acknowledgments The authors thank the Canadian Institute for Health Research (CIHR) for its continuous support through individual grants, Group grants and Team grants over the years. G. B., M. M., O. A. have been funded through fellowships of the Multidisciplinary Team in Locomotor Rehabilitation after Spinal Cord Injury. A. F. was supported by the Natural Sciences and Engineering Research Council of Canada and the Christopher and Dana Reeve Foundation. We also want to acknowledge support of the Fond de la Recherche en Santé du Québec. References Antri, M., Barthe, J.-Y., Mouffle, C., & Orsal, D. (2005). Long-lasting recovery of locomotor function in chronic spinal rat following chronic combined pharmacological stimulation of serotonergic receptors with 8-OHDAPT and quipazine. Neuroscience Letters, 384, 162–167. Armstrong, D. M. (1988). The supraspinal control of mammalian locomotion. Journal of Physiology, 405, 1–37. Barbeau, H., Julien, C., & Rossignol, S. (1987). The effects of clonidine and yohimbine on locomotion and cutaneous reflexes in the adult chronic spinal cat. Brain Research, 437, 83–96.
238 Barbeau, H., & Rossignol, S. (1987). Recovery of locomotion after chronic spinalization in the adult cat. Brain Research, 412, 84–95. Barbeau, H., & Rossignol, S. (1990). The effects of serotonergic drugs on the locomotor pattern and on cutaneous reflexes of the adult chronic spinal cat. Brain Research, 514, 55–67. Barbeau, H., & Rossignol, S. (1991). Initiation and modulation of the locomotor pattern in the adult chronic spinal cat by noradrenergic, serotonergic and dopaminergic drugs. Brain Research, 546, 250–260. Barrière, G., Frigon, A., Leblond, H., Provencher, J., & Rossignol, S. (2010). Dual spinal lesion paradigm in the cat: Evolution of the kinematic locomotor pattern. Journal of Neurophysiology, 104(2), 1119–1133. Barrière, G., Leblond, H., Provencher, J., & Rossignol, S. (2008). Prominent role of the spinal central pattern generator in the recovery of locomotion after partial spinal cord injuries. The Journal of Neuroscience, 28, 3976–3987. Barthélemy, D., Leblond, H., Mushahwar, V., & Rossignol, S. (2005). Use of intraspinal electrical stimulation for inducing locomotion in spinal cats. 35th Congress of the International Union of Physiological Sciences (IUPS), San Diego. Barthélemy, D., Leblond, H., Provencher, J., & Rossignol, S. (2006). Non-locomotor and locomotor hindlimb responses evoked by electrical microstimulation of the lumbar cord in spinalized cats. Journal of Neurophysiology, 96, 3273–3292. Barthélemy, D., Leblond, H., & Rossignol, S. (2007). Characteristics and mechanisms of locomotion induced by intraspinal microstimulation and dorsal root stimulation in spinal cats. Journal of Neurophysiology, 97, 1986–2000. Bédard, P., Barbeau, H., Barbeau, B., & Filion, M. (1979). Progressive increase of motor activity induced by 5-HTP in the rat below a complete section of the cord. Brain Research, 169, 393–397. Belanger, M., Drew, T., Provencher, J., & Rossignol, S. (1996). A comparison of treadmill locomotion in adult cats before and after spinal transection. Journal of Neurophysiology, 76, 471–491. Beloozerova, I., & Rossignol, S. (1999). Antidromic discharges in dorsal roots of decerebrate cats. I: Studies at rest and during fictive locomotion. Brain Research, 846, 87–105. Beloozerova, I., & Rossignol, S. (2004). Antidromic discharges in dorsal roots of decerebrate cats. II: Studies during treadmill locomotion. Brain Research, 996, 227–236. Bouyer, L. J. G., & Rossignol, S. (2003a). Contribution of cutaneous inputs from the hindpaw to the control of locomotion: 1. Intact cats. Journal of Neurophysiology, 90, 3625–3639. Bouyer, L. J. G., & Rossignol, S. (2003b). Contribution of cutaneous inputs from the hindpaw to the control of
locomotion: 2. Spinal cats. Journal of Neurophysiology, 90, 3640–3653. Brustein, E., & Rossignol, S. (1999). Recovery of locomotion after ventral and ventrolateral spinal lesions in the cat. II. Effects of noradrenergic and serotoninergic drugs. Journal of Neurophysiology, 81, 1513–1530. Bussel, B. C., Roby-Brami, A., Yakovleff, A., & Bennis, N. (1988). Evidences for the presence of a spinal stepping generator in patients with a spinal cord section. In B. Amblard, A. Berthoz & F. Clarac (Eds.), Posture and gait: Development, adaptation and modulation (pp. 273–278). North Holland: Elsevier. Calancie, B. (2006). Spinal myoclonus after spinal cord injury. The Journal of Spinal Cord Medicine, 29, 413–424. Carrier, L., Brustein, L., & Rossignol, S. (1997). Locomotion of the hindlimbs after neurectomy of ankle flexors in intact and spinal cats: Model for the study of locomotor plasticity. Journal of Neurophysiology, 77, 1979–1993. Cazalets, J. R., Sqalli-Houssaini, Y., & Clarac, F. (1992). Activation of the central pattern generators for locomotion by serotonin and excitatory amino acids in neonatal rat. Journal de Physiologie, 455, 187–204. Chau, C., Barbeau, H., & Rossignol, S. (1998a). Early locomotor training with clonidine in spinal cats. Journal of Neurophysiology, 79, 392–409. Chau, C., Barbeau, H., & Rossignol, S. (1998b). Effects of intrathecal a1- and a2-noradrenergic agonists and norepinephrine on locomotion in chronic spinal cats. Journal of Neurophysiology, 79, 2941–2963. Chau, C., Giroux, N., Barbeau, H., Jordan, L. M., & Rossignol, S. (2002). Effects of intrathecal glutamatergic drugs on locomotion. I. NMDA in short-term spinal cats. Journal of Neurophysiology, 88, 3032–3045. Chau, C., Giroux, N., Reader, T. A., & Rossignol, S. (2001). Ampa and alpha-2 adrenergic receptors in cat lumbo-sacral spinal cord following complete lesions. Abstracts—Society for Neuroscience, 27, 517.2. Chen, X. Y., & Wolpaw, J. R. (2002). Probable corticospinal tract control of spinal cord plasticity in the rat. Journal of Neurophysiology, 87, 645–652. Conway, B. A., Hultborn, H., Kiehn, O., & Mintz, I. (1988). Plateau potentials in alpha-motoneurones induced by intravenous injection of L-Dopa and clonidine in the spinal cat. Journal of Physiology, 405, 369–384. Côté, M.-P., & Gossard, J.-P. (2004). Step-training dependent plasticity in spinal cutaneous pathways. The Journal of Neuroscience, 24, 11317–11327. Côté, M.-P., Menard, A., & Gossard, J.-P. (2003). Spinal cats on the treadmill: Changes in load pathways. The Journal of Neuroscience, 23, 2789–2796. Courtine, G., Gerasimenko, Y., van den, B. R., Yew, A., Musienko, P., Zhong, H., et al. (2009). Transformation of
239 nonfunctional spinal circuits into functional states after the loss of brain input. Nature Neuroscience, 12, 1333–1342. Courtine, G., Song, B., Roy, R. R., Zhong, H., Herrmann, J. E., Ao, Y., et al. (2008). Recovery of supraspinal control of stepping via indirect propriospinal relay connections after spinal cord injury. Nature Medicine, 14, 69–74. Cowley, K. C., Zaporozhets, E., & Schmidt, B. J. (2008). Propriospinal neurons are sufficient for bulbospinal transmission of the locomotor command signal in the neonatal rat spinal cord. Journal of Physiology, 586, 1623–1635. Delcomyn, F. (1980). Neural basis of rhythmic behavior in animals. Science, 210, 492–498. Delivet-Mongrain, H., Leblond, H., & Rossignol, S. (2008). Effects of localized intraspinal injections of a noradrenergic blocker on locomotion of high decerebrate cats. Journal of Neurophysiology, 100, 907–921. Dietz, V., & Harkema, S. J. (2004). Locomotor activity in spinal cord-injured persons. Journal of Applied Physiology, 96, 1954–1960. Douglas, J. R., Noga, B. R., Dai, X., & Jordan, L. M. (1993). The effects of intrathecal administration of excitatory amino acid agonists and antagonists on the initiation of locomotion in the adult cat. The Journal of Neuroscience, 13, 990–1000. Drew, T., Prentice, S., & Schepens, B. (2004). Cortical and brainstem control of locomotion. Progress in Brain Research, 143, 251–261. Engberg, I., & Lundberg, A. (1969). An electromyographic analysis of muscular activity in the hindlimb of the cat during unrestrained locomotion. Acta Physiologica Scandinavica, 75, 614–630. Forssberg, H., & Grillner, S. (1973). The locomotion of the acute spinal cat injected with clonidine i.v. Brain Research, 50, 184–186. Forssberg, H., Grillner, S., & Halbertsma, J. (1980a). The locomotion of the low spinal cat. I. Coordination within a hindlimb. Acta Physiologica Scandinavica, 108, 269–281. Forssberg, H., Grillner, S., Halbertsma, J., & Rossignol, S. (1980b). The locomotion of the low spinal cat: II. Interlimb coordination. Acta Physiologica Scandinavica, 108, 283–295. Forssberg, H., Grillner, S., & Rossignol, S. (1975). Phase dependent reflex reversal during walking in chronic spinal cats. Brain Research, 85, 103–107. Frigon, A., Barriere, G., Leblond, H., & Rossignol, S. (2009). Asymmetric changes in cutaneous reflexes after a partial spinal lesion and retention following spinalization during locomotion in the cat. Journal of Neurophysiology, 102, 2667–2680. Frigon, A., & Rossignol, S. (2006a). Experiments and models of sensorimotor interactions during locomotion. Biological Cybernetics, 95, 607–627. Frigon, A., & Rossignol, S. (2006b). Functional plasticity following spinal cord lesions. Progress in Brain Research, 157, 231–260.
Frigon, A., & Rossignol, S. (2008). Adaptive changes of the locomotor pattern and cutaneous reflexes during locomotion studied in the same cats before and after spinalization. Journal of Physiology, 586, 2927–2945. Frigon, A., & Rossignol, S. (2009). Partial denervation of ankle extensors prior to spinalization in cats impacts the expression of locomotion and the phasic modulation of reflexes. Neuroscience, 158, 1675–1690. Galiana, H. L., Flohr, H., & Melvill Jones, G. (1984). A reevaluation of intervestibular nuclear coupling: Its role in vestibular compensation. Journal of Neurophysiology, 51, 242–258. Gerasimenko, Y., Gorodnichev, R., Machueva, E., Pivovarova, E., Semyenov, D., Savochin, A., et al. (2010). Novel and direct access to the human locomotor spinal circuitry. The Journal of Neuroscience, 30, 3700–3708. Gerasimenko, Y. P., Ichiyama, R. M., Lavrov, I. A., Courtine, G., Cai, L., Zhong, H., et al. (2007). Epidural spinal cord stimulation plus quipazine administration enable stepping in complete spinal adult rats. Journal of Neurophysiology, 98, 2525–2536. Ghosh, A., Sydekum, E., Haiss, F., Peduzzi, S., Zorner, B., Schneider, R., et al. (2009). Functional and anatomical reorganization of the sensory-motor cortex after incomplete spinal cord injury in adult rats. The Journal of Neuroscience, 29, 12210–12219. Gimenezyribotta, M., Provencher, J., Feraboli-Lohnherr, D., Rossignol, S., Privat, A., & Orsal, D. (2000). Activation of locomotion in adult chronic spinal rats is achieved by transplantation of embryonic raphe cells reinnervating a precise lumbar level. The Journal of Neuroscience, 20, 5144–5152. Giroux, N., Chau, C., Barbeau, H., Reader, T. A., & Rossignol, S. (2003). Effects of intrathecal glutamatergic drugs on locomotion. II. NMDA and AP-5 in intact and late spinal cats. Journal of Neurophysiology, 90, 1027–1045. Giroux, N., Reader, T. A., & Rossignol, S. (2001). Comparison of the effect of intrathecal administration of clonidine and yohimbine on the locomotion of intact and spinal cats. Journal of Neurophysiology, 85, 2516–2536. Giroux, N., Rossignol, S., & Reader, T. A. (1999). Autoradiographic study of a1-, a2-Noradrenergic and Serotonin 1A receptors in the spinal cord of normal and chronically transected cats. The Journal of Comparative Neurology, 406, 402–414. Gossard, J.-P., Cabelguen, J.-M., & Rossignol, S. (1989). Intraaxonal recordings of cutaneous primary afferents during fictive locomotion in the cat. Journal of Neurophysiology, 62, 1177–1188. Grillner, S. (1973). Locomotion in the spinal cat. In R. B. Stein, K. G. Pearson, R. S. Smith & J. B. Redford (Eds.), Control of posture and locomotion. Advances in behavioral biology, (Vol. 7, pp. 515–535). New York: Plenum Press. Grillner, S. (1981). Control of locomotion in bipeds, tetrapods, and fish. In J. M. Brookhart & V. B. Mountcastle (Eds.),
240 Handbook of physiology. The nervous system II (pp. 1179–1236). Bethesda, MD: American Physiological Society. Grillner, S., El Manira, A., Kiehn, O., Rossignol, S., & Stein, P. S. G. (2008). Networks in motion. Brain Research Reviews, 57, 1–269. Grillner, S., & Jessell, T. M. (2009). Measured motion: Searching for simplicity in spinal locomotor networks. Current Opinion in Neurobiology, 19, 572–586. Grillner, S., McClellan, A., Sigvardt, K., Wallen, P., & Wilen, M. (1981). Activation of NMDA-receptors elicits “fictive locomotion” in lamprey spinal cord in vitro. Acta Physiologica Scandinavica, 113, 549–551. Grillner, S., Wallen, P., Saitoh, K., Kozlov, A., & Robertson, B. (2008). Neural bases of goal-directed locomotion in vertebrates–an overview. Brain Research Reviews, 57, 2–12. Grillner, S., & Zangger, P. (1974). Locomotor movements generated by the deafferented spinal cord. Acta Physiologica Scandinavica, 91, 38A–39A. Grillner, S., & Zangger, P. (1979). On the central generation of locomotion in the low spinal cat. Experimental Brain Research, 34, 241–261. Harkema, S. J. (2008). Plasticity of interneuronal networks of the functionally isolated human spinal cord. Brain Research Reviews, 57, 255–264. Harvey, P. J., Li, X., Li, Y., & Bennett, D. J. (2006a). Endogenous monoamine receptor activation is essential for enabling persistent sodium currents and repetitive firing in rat spinal motoneurons. Journal of Neurophysiology, 96, 1171–1186. Harvey, P. J., Li, Y., Li, X., & Bennett, D. J. (2006b). Persistent sodium currents and repetitive firing in motoneurons of the sacrocaudal spinal cord of adult rats. Journal of Neurophysiology, 96, 1141–1157. Heckmann, C. J., Gorassini, M. A., & Bennett, D. J. (2005). Persistent inward currents in motoneuron dendrites: Implications for motor output. Muscle & Nerve, 31, 135–156. Hounsgaard, J., Hultborn, H., Jespersen, J., & Kiehn, O. (1988). Bistability of alpha-motoneurones in the decerebrate cat and in the acute spinal cat after intravenous 5-hydroxytryptophan. Journal of Physiology, 405, 345–367. Hultborn, H., Brownstone, R. B., Toth, T. I., & Gossard, J.-P. (2004). Key mechanisms for setting the input-output gain across the motoneuron pool. Progress in Brain Research, 143, 77–95. Ichiyama, R. M., Gerasimenko, Y. P., Zhong, H., Roy, R. R., & Edgerton, V. R. (2005). Hindlimb stepping movements in complete spinal rats induced by epidural spinal cord stimulation. Neuroscience Letters, 383, 339–344. Jankowska, E., Jukes, M. G. M., Lund, S., & Lundberg, A. (1967a). The effect of DOPA on the spinal cord. 5. Reciprocal organization of pathways transmitting excitatory action
to alpha motoneurones of flexors and extensors. Acta Physiologica Scandinavica, 70, 369–388. Jankowska, E., Jukes, M. G. M., Lund, S., & Lundberg, A. (1967b). The effects of DOPA on the spinal cord. 6. Half centre organization of interneurones transmitting effects from the flexor reflex afferents. Acta Physiologica Scandinavica, 70, 389–402. Kiehn, O., & Kjaerulff, O. (1996). Spatiotemporal characteristics of 5-HT and dopamine-induced rhythmic hindlimb activity in the in vitro neonatal rat. Journal of Neurophysiology, 75, 1472–1482. Kolta, A., Westberg, K. G., Clavelou, P., & Lund, J. P. (1995). Do muscle spindle afferent act as interneurons during mastication? Trends in Neurosciences, 18, 441. Langlet, C., Leblond, H., & Rossignol, S. (2005). The mid-lumbar segments are needed for the expression of locomotion in chronic spinal cats. Journal of Neurophysiology, 93, 2474–2488. Leblond, H., L'Espérance, M., Orsal, D., & Rossignol, S. (2003). Treadmill locomotion in the intact and spinal mouse. The Journal of Neuroscience, 23, 11411–11419. Li, X., Murray, K. C., Harvey, P. J., Ballou, E. W., & Bennett, D. J. (2007). Serotonin facilitates a persistent calcium current in motoneurons of rats with and without chronic spinal cord injury. Journal of Neurophysiology, 97, 1236–1246. Lund, J. P., Drew, T., & Rossignol, S. (1984). A study of jaw reflexes of the awake cat during mastication and locomotion. Brain, Behavior and Evolution, 25, 146–156. Marcoux, J., & Rossignol, S. (2000). Initiating or blocking locomotion in spinal cats by applying noradrenergic drugs to restricted lumbar spinal segments. The Journal of Neuroscience, 20, 8577–8585. McDearmid, J. R., Scrymgeour-Wedderburn, J. F., & Sillar, K. T. (1997). Aminergic modulation of glycine release in a spinal network controlling swimming in Xenopus laevis. Journal of Physiology, 503, 111–117. Murray, K., Nakae, A., Stephens, M. J., Rank, M., D'Amico, J., Harvey, P., et al. (2010). Recovery of motoneuron and locomotor function after chronic spinal cord injury depends on constitutive activity in 5-HT2c receptors. Nature Medicine, 16, 694–700. Nadeau, S., Jacquemin, G., Fournier, C., Lamarre, Y., & Rossignol, S. (2010). Spontaneous motor rhythms of the back and legs in a patient with a complete spinal cord transection. Neurorehabilitation and Neural Repair, 24, 377–383. Noga, B. R., Johnson, D. M., Riesgo, M. I., & Pinzon, A. (2009). Locomotor-activated neurons of the cat. I. Serotonergic innervation and co-localization of 5-HT7, 5-HT2A and 5-HT1A receptors in the thoraco-lumbar spinal cord. Journal of Neurophysiology, 102, 1560–1576. Pearson, K. G., & Rossignol, S. (1991). Fictive motor patterns in chronic spinal cats. Journal of Neurophysiology, 66, 1874–1887.
241 Remy-Neris, O., Barbeau, H., Daniel, O., Boiteau, F., & Bussel, B. (1999). Effects of intrathecal clonidine injection on spinal reflexes and human locomotion in incomplete paraplegic subjects. Experimental Brain Research, 129, 433–440. Rossignol, S. (1995). Recovery of locomotion in cats after lesions of the spinal cord. Acta Neurobiologiae Experimentalis, 55, 4. Rossignol, S. (1996). Neural control of stereotypic limb movements. In L. B. Rowell & J. T. Sheperd (Eds.), Handbook of physiology, section 12. Exercise: Regulation and integration of multiple systems (pp. 173–216). New York: Oxford University Press. Rossignol, S. (2006). Plasticity of connections underlying locomotor recovery after central and/ or peripheral lesions in the adult mammals. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences, 361, 1647–1671. Rossignol, S., Barriere, G., Alluin, O., & Frigon, A. (2009). Re-expression of Locomotor Function After Partial Spinal Cord Injury. Physiology (Bethesda), 24, 127–139. Rossignol, S., Bélanger, M., Chau, C., Giroux, N., Brustein, E., Bouyer, L., et al. (2000). The spinal cat. In R. G. Kalb & S. M. Strittmatter (Eds.), Neurobiology of spinal cord injury (pp. 57–87). Totowa: Humana Press. Rossignol, S., Chau, C., Brustein, E., Giroux, N., Bouyer, L., Barbeau, H., et al. (1998). Pharmacological activation and modulation of the Central Pattern Generator for locomotion in the cat. Annals of the New York Academy of Sciences, 860, 346–359. Rossignol, S., Chau, C., Giroux, N., Brustein, E., Bouyer, L., Marcoux, J., et al. (2002). The cat model of spinal injury. Progress in Brain Research, 137, 151–168. Rossignol, S., Dubuc, R., & Gossard, J. P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86, 89–154. Rossignol, S., Giroux, N., Chau, C., Marcoux, J., Brustein, E., & Reader, T. A. (2001). Pharmacological aids to locomotor training after spinal injury in the cat. Journal of Physiology, 533, 65–74. Rossignol, S., Grillner, S., & Forssberg, H. (1975). Factors of importance for the initiation of flexion during walking. Abstracts—Society for Neuroscience, 1, 181. Rossignol, S., Schwab, M., Schwartz, M., & Fehlings, M. G. (2007). Spinal cord injury: Time to move? The Journal of Neuroscience, 27, 11782–11792.
Rudomin, P., Quevedo, J., & Eguibar, J. R. (1993). Presynaptic modulation of spinal reflexes. Current Opinion in Neurobiology, 3, 997–1004. Sakurai, A., & Katz, P. S. (2009). Functional recovery after lesion of a central pattern generator. The Journal of Neuroscience, 29, 13115–13125. Saltiel, P., & Rossignol, S. (2004). Critical points in the forelimb fictive locomotor cycle and motor coordination: Evidence from the effects of tonic proprioceptive perturbations in the cat. Journal of Neurophysiology, 92, 1329–1341. Schmidt, B. J., & Jordan, L. M. (2000). The role of serotonin in reflex modulation and locomotor rhythm production in the mammalian spinal cord. Brain Research Bulletin, 53, 689–710. Sherrington, C. S. (1910). Flexion-reflex of the limb, crossed extension-reflex, and reflex stepping and standing. Journal of Physiology, 40, 28–121. Sherrington, C. S., & Laslett, E. E. (1903). Observations on some spinal reflexes and the interconnection of spinal segments. Journal of Physiology, 29, 58–96. Stuart, D. G., & Hultborn, H. (2008). Thomas Graham Brown (1882–1965), Anders Lundberg (1920-), and the neural control of stepping. Brain Research Reviews, 59, 74–95. Takeoka, A., Kubasak, M. D., Zhong, H., Roy, R. R., & Phelps, P. E. (2009). Serotonergic innervation of the caudal spinal stump in rats after complete spinal transection: Effect of olfactory ensheathing glia. The Journal of Comparative Neurology, 515, 664–676. Vanderhorst, V. G. J. M., & Holstege, G. (1997). Organization of lumbosacral motoneuronal cell groups innervating hindlimb, pelvic floor, and axial muscles in the cat. The Journal of Comparative Neurology, 382, 46–76. Viala, G., Orsal, D., & Buser, P. (1978). Cutaneous fiber groups involved in the inhibition of fictive locomotion in the rabbit. Experimental Brain Research, 33, 257–267. Wolpaw, J. R., & Carp, J. S. (2006). Plasticity from muscle to brain. Progress in Neurobiology, 78, 233–263. Yakovenko, S., & Drew, T. (2009). A motor cortical contribution to the anticipatory postural adjustments that precede reaching in the cat. Journal of Neurophysiology, 102, 853–874. Zaporozhets, E., Cowley, K. C., & Schmidt, B. J. (2006). Propriospinal neurons contribute to bulbospinal transmission of the locomotor command signal in the neonatal rat spinal cord. Journal de Physiologie, 572, 443–458.
Subject Index
Alpha–gamma coupling awake-behaving animals, 220 masseter motoneurons and spindle afferents, 220 muscle spindle afferents, 220 rhythmical licking, 220
gasping activity, 36–37 in vivo and in vitro conditions, 36 inputs, synaptic, 36 locomotion intrinsic membrane properties, 36 nonpacemakers, 36–37 pacemaker neurons, 36–37 pre-Bötzinger complex, 35 respiration possess bursting properties, 37 rhythm generation, 34 synaptic mechanisms, 34 well-oxygenated conditions, 34 networks within networks, 33 neuromodulation and rhythm generation behavioral conditions, 39 bioamine acting, 38 diverging effects, 40 diverse modulatory process, 40 homeostatic mechanisms, 40–41 neurological disorders, 38 noradrenergic neurons, 38 peptidergic modulation, 39 raphe magnus and obscurus, 38 receptor subtypes, 37 rhythm generation parameter, 39 rodent species and strains, 40 serotonin and norepinephrine levels, 40 stimulatory effect, 37–38 neuromodulatory process, 32 pre-BötC causes, 32 respiratory system, 32
Breathing disorders, 213 in vitro approaches, 214 mammalian functions, 213 mathematical/computational model, 216 neonatal rodent spinal cord preparation, 215 neural control, 214 NK1R, 214–215 pacemaker vs. nonpacemaker, 215 preBötC, 215 principles, 214 problems, 215 respiratory oscillator, 216 rhythmogenesis, 214 rhythmic inspiratory-dominated, 215 RTN/pFRG, 217 two-oscillator hypothesis, 216 Breathing, neuronal control behavioral level, 31 homeostatic plasticity, 41–42 in vitro approach, 33–34 inspiration and expiration, 32 Kölliker–Fuse nucleus, 32 mammalian respiratory network, 32 network reconfiguration depression phase, 35
Cadmium sensitive (CS) pacemaker neurons, 36 Central pattern generator (CPG) 243
244
Central pattern generator (CPG) (Continued) cycle-by-cycle repetition, 24 functional organization and reorganization, 18 locomotor and scratching rhythms, 25 locomotor rhythm, 18 networks, 19 organization and function control mechanism features, 156 Hodgkin–Huxley formalism, 156 mechanical and task demands, 155 neural circuits, 156 oscillatory dynamics, 156 oscillatory mechanism, 155 rhythm and pattern, 155 synaptic and intrinsic parameters, 156 Vérzar’s model, 155 oscillator, 17 reflexive stepping, 153 sensory and descending pathways, 153 Central timing network (CTN), 24 Cortical masticatory area (CMA) microzones, 74 neurons showing activity, 76 neuroplasticity, 77 semiautomatic orofacial movements, 76 CPG. See Central pattern generator Dorsal root ganglion (DRG) neurons chloride homeostasis, 5 homozygote mutant mice, 5 NKCC1 expression, 4 trigeminal mesencephalic nucleus, 9 DRG neurons. See Dorsal root ganglion neurons Face sensorimotor cortex neuroplasticity behavior implications altered motor behavior, 144 behavioral changes, 143 dental extraction and trimming, 144–145 limb motor skill training, 144 peripheral somatosensory feedback, 142 phantom limb pain, 143 rodents, 143 somatosensory inputs, 143 time-dependent neuroplastic changes, 144
clinical significance cortical and subcortical neuroplastic mechanisms, 146 intraoral manipulations, 145 oral tissues and modifications, 145 rehabilitative training, 146 extensive somatosensory and motor representations, 136 functions, 136 ICMS technique, 136 masticatory movements, 145 objectives and methods coronal cross sections, 140 dental procedures, 137–138 ICMS mapping, 138 intraoral manipulation, 137–138 jaw-opening muscles, 137 LAD, RAD, and GG, 138–139 masticatory functions, 137 systematic cortical mapping, 138 orofacial functions, 137 motor representations, 136–137 rat face M1 and face-S1 motor representations features, 140–141 intraoral manipulation effects, 141–142 rhythmic jaw movements, 136 somatosensory inputs reorganization, 145 Face sensorimotor cortex, orofacial movement brainstem circuits role and features, 73–74 central nervous system (CNS), 72 CMA, 76 comprehensive neural encoding, 74 infant animal matures, 73 jaw-closing muscle activities, 73 mastication and swallowing, 76 motor area (MI) and somatosensory area (SI), 74 neural and chemical regulatory processes, 73 mechanisms, 72 neurons showing activity, 76 neuroplasticity intraoral alterations, 78–80 MI and SI neurons, 76–77
245
motor skills, 77–78 structural and functional change, 76 properties and roles, 76 reflex component, 73 rhythmic masticatory-like movements, 74 Force field adaptation control parameters, 130 limb dynamics, 126 locomotor movement, 125 phase-dependent alternative hypothesis, 127 feedforward and feedback contributions, 127 force environment, 128 neural control, 128 pseudorandomly inserted catch trials, 129 robotized ankle exoskeleton, 128 velocity-dependent resistance, 127 protocol, 126 rhythmic movement, 126–127 training, 130 upper limb, 125 Homeostatic plasticity degree of influence, 42 forms, 41 glia cells activation, 41 long-term depression (LTD), 41 modulatory systems, 42 neural networks, 41 pre-and post-synaptic scaling, 41 serotonin receptors, 42 5-Hydroxytryptamine (5-HT) chewing, 187 locomotion antagonist motoneurons groups, 185 bipolar EMG electrodes, 186 EMG and kinematic analysis, 184–185 extensor activity, 186 features, 183 inhibitory interneurons, 184 knock-out mice, 184 lamprey swimming, 185 mammalian breathing, 183–184 oscillatory behavior, 185
receptor subtypes, 184 rhythmic motor tasks, 183 spinal application, 184 respiration control circuit components, 183 GABAergic and glycinergic inhibition, 183 rhythmic behavior, 182 Robo3-dependent guidance, 182 Hyperreflexia and spasticity novel mechanism electrical coupling Cx-36 protein level, 175 direct cell to cell communication, 174 locomotor movements, 167–168, 173–176 MOD, 176 motoneuron, 174 single neurons, 174–175 spinal cord level, 175 H-reflex component, 169 frequency-dependent depression, 169, 172 numerous investigators, 169 passive exercise, 172 posttransection, 171–172 spinal cord injury, 168–169 stretch reflex EMG response, 169 GABAergic presynaptic inhibition, 173 habituation, 172–173 Spinal transection, 173 windup behavior, 170 ICMS. See Intracortical microstimulation Intracortical microstimulation (ICMS) dental procedures, 137–138 electromyographic (EMG) activity, 137 face-M1 and face-S1, 141 face-sensorimotor cortex, 142 jaw-opening features, 140 mapping, 138 masticatory functions, 137 motor representations, 137 muscle motor representation, 139 onset latency, 139–140
246
Intracortical microstimulation (ICMS) (Continued) organizational features, 138 subsequent histological verification, 138 Laterodorsal tegmental nucleus (LDT), 55 Locomotion recovery, spinal plasticity adult cats, 231 ankle extensors, 230 clinical importance, 230 in vitro neonatal rats, 233 limb extensors, 231 neurotransmitter modulation, 231 noradrenergic mechanisms adrenergic receptors, 231 described, 232 interlimb coordination, 232 robust and fundamental primitive motor act, 229 segmental control intrathecal drug injections, 235 propriospinal pathways, 236 rhythmic activities, 235 sensory modulation cutaneous denervation, 235 dynamic rhythmic process, 234 phasic sensory, 234 proprioceptive and cutaneous pathways, 235 sensorimotor interactions, 234 tonic sensory stimuli, 234 tritonia swimming, 234 serotoninergic mechanisms 5-HT receptors, 233 pharmacological stimulation, 232 segmental distribution, 232 suprasegmental control cuff electrodes, 237 spinalization, 236 step cycle and interlimb coupling, 237 supraspinal structures, 236–237 timing characteristics, 230 Mastication, neural control alpha–gamma coupling, 220
chronic pain, motor systems acidic saline injections, 225–226 acid-treated animals, 226 hypertonic saline, 225 masseter muscle spindle mechanoreceptors, 225 mechanoreceptor afferents, 226 muscle hyperactivity, 224 muscle spindles, 226 nociceptors, 224 pain-adaptation model, 225 rostral parvocellular reticular formation, 225 sensory afferents, 226 tonic noxious inputs, 225 CPG, 221 muscle spindle afferents, 225–226 presynaptic modulation form antidromic action potentials, 222–223 CPG, 223 GABA, 223–224 primary afferent terminals, 222 spindle afferents, 223 reflexes modulation, 221–222 rhythmical motor patterns, 219 sensorimotor cortex, 220 sensory feedback, 220 Medial gastrocnemii (MG), 102 Medial longitudinal fasciculus (MLF), 23 Mesencephalic locomotor region (MLR) context-dependent manner, 54 electrical stimulation, 22 fictive locomotion, 21 flexor dominance, 23 hindlimb nerves and, 23 motor output phases, 52 nuclei, 52 phase-dependent response, 25 PPN, 54 reticulospinal cells, 23 RS cells, 54 salamanders, 54 spontaneous rhythm, 23 ventral pallidum, 52
247
MLR. See Mesencephalic locomotor region Motorized bicycle exercise training (MBET), 170 Neurokinin 1 receptor (NK1R), 214 Neuroplasticity, face sensorimotor cortex intraoral alterations brainstem sensory nucleus complex, 79 capsaicin-induced facial, 79 ICMS threshold changes, 78 maladaptive behaviors, 78 novel motor skills, 80 orofacial sites, 78 static and dynamic motor activities, 79 orofacial motor skills lingual mechanosensory inputs, 77 MI and SI, 77–78 semiautomatic oral behavior, 77 sensory inputs, 77 TMS-defined tongue representation, 78 structural and functional changes, 76 PAD. See Primary afferent depolarization Peripheral nerve/spinal cord injury, locomotor adaptation compensatory mechanisms, 103 CPG and reflex pathways, 101–102 implications adequate walking, 114 muscular activation, 113 remnant pathways and structures, 114 sensorimotor systems, 113 treadmill training, 113 interindividual variability factors, 108 gait transitions and walking speed, 104–105 muscle synergies, 106–107 recruitment patterns and spinal reflexes, 105–106 invertebrate motor systems, 103 locomotion, 102 neuronal circuitry, 101 spinal locomotor, 102 supraspinal signals and phasic sensory feedback, 103 tendon transfers, 102
variable adaptive mechanisms, 108–113 withdrawal reflexes, 103 Persistent inward currents (PICs), 233 Phases and cycle duration, spinal generation descending inputs, 19 fictive locomotor rhythm, 19–20 flexion-interval phase, 20 flexors and extensors, 19 MLR-induced fictive locomotion, 20–21 patterns, 18 period duration, 20 respiration and mastication, 20 rhythm generator, functional organization CPG models, 24 fictive locomotion, 25 locomotion and scratching, 25 locomotor and scratching rhythms, 25 mammalian spinal cord, 26 modulatory systems, 25 and scratch, 20 sensory inputs additive interaction, 21 ankle extensors, 22 dorsiflexion effects, 22 extension and flexion-interval, 22 flexor and extensor bursts, 24 locomotion, respiration and mastication, 21 locomotor generator, 22 MLR stimulation, 23 rhythmic movements, 22 supraspinal structures, 23 speed and stance air-scratching, 17 ankle dorsiflexion, 18 descending supraspinal signals, 16 electromyograms (EMG), 16 fictive locomotion, 16 flexion and extension, 16 half-centers oscillator model, 17 locomotion swing phase, 18 locomotor networks, 17 period relationships, 17 respiration and mastication, 17–18 rhythmic motor behaviors, 18 subcomponents, 16
248
Phases and cycle duration, spinal generation (Continued) supraspinal and sensory inputs, 16 spinal transection, 19 Posterior parietal cortex (PPC) behavioral process, 84 cortical area, 84 locomotion guide, 84 motor planning complex integration and planning process, 85 obstacle and attributes, 86–87 obstacle avoidance task, 85 reaching and grasping, 84 state estimation, 87–88 task selection, 88–89 TTC signals, 85 visually guided locomotion lesion studies, 90–92 single-unit recording studies, 92–98 PPC. See Posterior parietal cortex Pre-Bötzinger complex (preBötC) breathing in adult rats, 215 glutamatergic interneurons, 216 inspiratory activity, 210 neurons, 209 rapid silencing, 214 transverse medullary slices, 215 Primary afferent depolarization (PAD) antidromic discharge, 9 bicuculline, 9 generation, 9 muscle spindle afferents, 9–10 phasic inhibition, 9 premotor interneuron carrying signals, 10 Rhythm generation, locomotor control animal behaviors, 152 complex hierarchical system, 152 consecutive antagonistic actions, 152 CPG organization and function, 155–156 intrinsic and sensory feedback signals play, 153 intrinsic spinal rhythmogenesis, 152 neural structures, 151–152 pursuing and catching prey, 151 spinal networks
bilateral pattern, 159 Brown’s locomotor pattern generator, 157 CPG command signal, 159–160 double-stepping using velocity, 159–160 flexor and extensor phases, 157–158 half-center elements, 157 hard-wired parameters, 159 ipsilateral reticulospinal neurons, 160 model parameters, CPG, 158–159 multiple descending pathways, 160 phase-duration characteristic error, 158 phase durations, 157 task-specific goals, 158 stabilizing properties autogenic and heterogenic sensory feedback pathways, 154 conceptual framework, 153 inadequate model assumptions, 154 intrinsic, 154 inverted pendulum model, 153 length-force relationship, 154 mass-spring system, 153 muscle force-length characteristics, 154 oscillatory behavior, 153 sliding-filament theory, 154 vertebrates, 152 Rhythmic motor networks, chloride homeostasis cation-chloride cotransporters, 8 cell excitability glycine receptors, 6 neuronal excitability, 6 shunting mechanisms, 6 classical inhibitory neurotransmitters, 8 DRG neurons, 4 dysfunction, pathological conditions, 10–11 GABA, 3 GABAergic synapses, 7 gap junction coupling, 6–7 glutamatergic connections, 8 glycine mediated inhibition, 8 receptors, 3–4 inhibitory and excitatory actions, 7 KCC2, 6 late gestation, 6
249
mechanisms, 5 NKCC1 knock-out mice, 4 non-NMDA receptors, 7 PAD antidromic discharge, 9 bicuculline, 9 generation, 9 muscle spindle afferents, 9–10 phasic inhibition, 9 premotor interneuron carrying signals, 10 respiratory frequency, 8–9 serotonin application, 8 Rhythmic movement control, adaptive capacity central pattern generators, 121–122 feedforward and feedback control, 120–122 force field adaptation control parameters, 130 limb dynamics, 126 locomotor movement, 125 phase-dependent, 127–129 protocol, 126 training, 130 upper limb, 125 motor output over lifespan, 120 partial denervation model extent and limits, 123 limits, 124–125 underlying mechanisms, 123–124 tailored force fields, 127 Rhythmic movement modulation 5-HT chewing, 187 control, respiration, 182–183 locomotion, 183–187 inhibitory interneurons flexor and extensor motoneurons, 187 GABAergic inhibition, 190 glycine receptor, 190 glycinergic inhibitory interneurons, 189 glycinergic transmission, 187 inhibitory interneurons interferes, 187–188 intra and interlimb coordination, 187 left-right phasic activity, 189 motoneuron activation, 187
NMDA-induced locomotion, 189 pre-Bötzinger complex (PBC), 191 primary afferent depolarization (PAD), 191 strychnine, 190 respiration, locomotion, and mastication, 181 Rostral ventral respiratory group (rVRG), 32 Runge–Kutta (fourth order) method, 157–158 Subthalamic locomotor region (SLR), 23 Sudden infant death syndrome (SIDS), 32, 40, 182 Supraspinal control, locomotion CN activation, 53 description, 52–53 lamprey’s model bilateral MLR inputs, 58–61 forebrain projections, 55 glutamatergic and cholinergic MLR outputs, 56–57 locomotor-boosting mechanism, 61–62 MLR, 55 motor behavior, 55 nervous system, 54–55 nucleus gigantocellularis, 55 RS cells, MRRN and PRRN, 57–58 sensory inputs gating, 62–65 swimming movement velocity, 55–56 vertebrate species, 55 MLR context-dependent manner, 54 motor output phases, 52 nuclei, 52 PPN, 54 RS cells, 54 salamanders, 54 ventral pallidum, 52 supraspinal control, 52 vertebrates, 52 Tibialis anterior (TA), 102 Time-to-contact (TTC) signals avoidance behavior, 84 limb state, 85 mechanisms, 86 object approaches, 86
250
Time-to-contact (TTC) signals (Continued) obstacle collision time, 88 size/dimensions, 87 TMS. See Transcranial magnetic stimulation Transcranial magnetic stimulation (TMS) motor potentials, 78 tongue task analogous, 78 TTC signals. See Time-to-contact signals Variable adaptive mechanisms dual-lesion paradigms compensatory mechanisms, 112 motor skill level, 113 peripheral denervation, 112 peripheral nerve lesion, 112 spinal locomotion, 112 treadmill training, 112 peripheral nerve lesions electrodes and hindlimb kinematics, 108 hip and metatarsophalangeal (MTP) joints, 109 locomotor and reflex adaptation, 108 locomotor deficit, 109 methodological problems, 108 spinal cord lesions angular excursions and patterns, 110 phase-dependent modulation, 112 reflex responses, 111 short-latency inhibition, 109 structures, 109 supraspinal signals, 109 Visually guided locomotion lesion studies effects, 90 exteroceptive visual information, 90 forelimbs and hindlimbs straddling, 91 lateral sulcus, 90 voluntary gait modifications, 91 obstacle speed, 90 PPC, 89 single-unit recording studies cells population, 92
contralateral forelimb (coFL), 95 discharge characteristics, 93 forelimb-hindlimb cell, 96 hazard detector, 93 limb-independent cell, 95 limb-specific cells, 94 neural representation, 94 neuronal recordings, 92 proprioceptive and visual inputs, 94 step-related and step-advanced cell, 93 step-related neurons, 92 TTC, 97 step-by-step regulation, 89 Walking, chewing and breathing Bötzinger complex, 208 brainstem respiratory CPG, 207–208 expiratory phase, 209 features, 199 feline locomotor control CPG gating, 201 movements, 201 noradrenergic and serotonergic drugs, 203 phase-dependent reflex, 201–202 research structure, 201 Rossignol model, 203 spinal cord injury and rehabilitation, 204 spinal locomotor mechanisms, 201 locomotion structure, 200 modeling environment, 207 neural control experimental model, 205 hypoglossal motoneurons, 207 jaw movements, 206 mastication, 206 masticatory control system, 205 masticatory motor pattern, 204 NK1 receptors, 209 preBötC, 210 sensory control, 200
Other volumes in PROGRESS IN BRAIN RESEARCH Volume 149: Cortical Function: A View from the Thalamus, by V.A. Casagrande, R.W. Guillery and S.M. Sherman (Eds.) – 2005 ISBN 0-444-51679-4. Volume 150: The Boundaries of Consciousness: Neurobiology and Neuropathology, by Steven Laureys (Ed.) – 2005, ISBN 0-444-51851-7. Volume 151: Neuroanatomy of the Oculomotor System, by J.A. Büttner-Ennever (Ed.) – 2006, ISBN 0-444-51696-4. Volume 152: Autonomic Dysfunction after Spinal Cord Injury, by L.C. Weaver and C. Polosa (Eds.) – 2006, ISBN 0-444-51925-4. Volume 153: Hypothalamic Integration of Energy Metabolism, by A. Kalsbeek, E. Fliers, M.A. Hofman, D.F. Swaab, E.J.W. Van Someren and R.M. Buijs (Eds.) – 2006, ISBN 978-0-444-52261-0. Volume 154: Visual Perception, Part 1, Fundamentals of Vision: Low and Mid-Level Processes in Perception, by S. Martinez-Conde, S.L. Macknik, L.M. Martinez, J.M. Alonso and P.U. Tse (Eds.) – 2006, ISBN 978-0-444-52966-4. Volume 155: Visual Perception, Part 2, Fundamentals of Awareness, Multi-Sensory Integration and High-Order Perception, by S. Martinez-Conde, S.L. Macknik, L.M. Martinez, J.M. Alonso and P.U. Tse (Eds.) – 2006, ISBN 978-0-444-51927-6. Volume 156: Understanding Emotions, by S. Anders, G. Ende, M. Junghofer, J. Kissler and D. Wildgruber (Eds.) – 2006, ISBN 978-0-444-52182-8. Volume 157: Reprogramming of the Brain, by A.R. M ller (Ed.) – 2006, ISBN 978-0-444-51602-2. Volume 158: Functional Genomics and Proteomics in the Clinical Neurosciences, by S.E. Hemby and S. Bahn (Eds.) – 2006, ISBN 978-0-444-51853-8. Volume 159: Event-Related Dynamics of Brain Oscillations, by C. Neuper and W. Klimesch (Eds.) – 2006, ISBN 978-0-444-52183-5. Volume 160: GABA and the Basal Ganglia: From Molecules to Systems, by J.M. Tepper, E.D. Abercrombie and J.P. Bolam (Eds.) – 2007, ISBN 978-0-444-52184-2. Volume 161: Neurotrauma: New Insights into Pathology and Treatment, by J.T. Weber and A.I.R. Maas (Eds.) – 2007, ISBN 978-0-444-53017-2. Volume 162: Neurobiology of Hyperthermia, by H.S. Sharma (Ed.) – 2007, ISBN 978-0-444-51926-9. Volume 163: The Dentate Gyrus: A Comprehensive Guide to Structure, Function, and Clinical Implications, by H.E. Scharfman (Ed.) – 2007, ISBN 978-0-444-53015-8. Volume 164: From Action to Cognition, by C. von Hofsten and K. Rosander (Eds.) – 2007, ISBN 978-0-444-53016-5. Volume 165: Computational Neuroscience: Theoretical Insights into Brain Function, by P. Cisek, T. Drew and J.F. Kalaska (Eds.) – 2007, ISBN 978-0-444-52823-0. Volume 166: Tinnitus: Pathophysiology and Treatment, by B. Langguth, G. Hajak, T. Kleinjung, A. Cacace and A.R. M ller (Eds.) – 2007, ISBN 978-0-444-53167-4. Volume 167: Stress Hormones and Post Traumatic Stress Disorder: Basic Studies and Clinical Perspectives, by E.R. de Kloet, M.S. Oitzl and E. Vermetten (Eds.) – 2008, ISBN 978-0-444-53140-7. Volume 168: Models of Brain and Mind: Physical, Computational and Psychological Approaches, by R. Banerjee and B.K. Chakrabarti (Eds.) – 2008, ISBN 978-0-444-53050-9. Volume 169: Essence of Memory, by W.S. Sossin, J.-C. Lacaille, V.F. Castellucci and S. Belleville (Eds.) – 2008, ISBN 978-0-444-53164-3. Volume 170: Advances in Vasopressin and Oxytocin – From Genes to Behaviour to Disease, by I.D. Neumann and R. Landgraf (Eds.) – 2008, ISBN 978-0-444-53201-5. Volume 171: Using Eye Movements as an Experimental Probe of Brain Function—A Symposium in Honor of Jean BüttnerEnnever, by Christopher Kennard and R. John Leigh (Eds.) – 2008, ISBN 978-0-444-53163-6. Volume 172: Serotonin–Dopamine Interaction: Experimental Evidence and Therapeutic Relevance, by Giuseppe Di Giovanni, Vincenzo Di Matteo and Ennio Esposito (Eds.) – 2008, ISBN 978-0-444-53235-0. Volume 173: Glaucoma: An Open Window to Neurodegeneration and Neuroprotection, by Carlo Nucci, Neville N. Osborne, Giacinto Bagetta and Luciano Cerulli (Eds.) – 2008, ISBN 978-0-444-53256-5. Volume 174: Mind and Motion: The Bidirectional Link Between Thought and Action, by Markus Raab, Joseph G. Johnson and Hauke R. Heekeren (Eds.) – 2009, 978-0-444-53356-2. Volume 175: Neurotherapy: Progress in Restorative Neuroscience and Neurology — Proceedings of the 25th International Summer School of Brain Research, held at the Royal Netherlands Academy of Arts and Sciences, Amsterdam, The Netherlands, August 25–28, 2008, by J. Verhaagen, E.M. Hol, I. Huitinga, J. Wijnholds, A.A. Bergen, G.J. Boer and D.F. Swaab (Eds.) –2009, ISBN 978-0-12-374511-8. Volume 176: Attention, by Narayanan Srinivasan (Ed.) – 2009, ISBN 978-0-444-53426-2. Volume 177: Coma Science: Clinical and Ethical Implications, by Steven Laureys, Nicholas D. Schiff and Adrian M. Owen (Eds.) – 2009, 978-0-444-53432-3. Volume 178: Cultural Neuroscience: Cultural Influences On Brain Function, by Joan Y. Chiao (Ed.) – 2009, 978-0-444-53361-6. Volume 179: Genetic models of schizophrenia, by Akira Sawa (Ed.) – 2009, 978-0-444-53430-9. Volume 180: Nanoneuroscience and Nanoneuropharmacology, by Hari Shanker Sharma (Ed.) – 2009, 978-0-444-53431-6.
252
Other volumes in PROGRESS IN BRAIN RESEARCH
Volume 181: Neuroendocrinology: The Normal Neuroendocrine System, by Luciano Martini, George P. Chrousos, Fernand Labrie, Karel Pacak and Donald W. Pfaff (Eds.) – 2010, 978-0-444-53617-4. Volume 182: Neuroendocrinology: Pathological Situations and Diseases, by Luciano Martini, George P. Chrousos, Fernand Labrie, Karel Pacak and Donald W. Pfaff (Eds.) – 2010, 978-0-444-53616-7. Volume 183: Recent Advances in Parkinson's Disease: Basic Research, by Anders Björklund and M. Angela Cenci (Eds.) – 2010, 978-0-444-53614-3. Volume 184: Recent Advances in Parkinson's Disease: Translational and Clinical Research, by Anders Björklund and M. Angela Cenci (Eds.) – 2010, 978-0-444-53750-8. Volume 185: Human Sleep and Cognition, by Gerard A. Kerkhof and Hans P.A. Van Dongen (Eds.) – 2010, 978-0-444-53702-7. Volume 186: Sex Differences in the Human Brain, their Underpinnings and Implications, by Ivanka Savic (Ed.) – 2010, 978-0-44453630-3. Volume 187: Breathe, Walk and CHew; the Neural Challenge: Part II, by Jean-Pierre Gossard, Réjean Dubuc and Arlette Kolta (Eds.) – 2011, 978-0-444-53825-3.
