Characterization of Facial Pain Using an Operant Behavioral Testing Paradigm and Evaluating the Role of Transient Recept...

Permanent Link: http://ufdc.ufl.edu/UFE0022896/00001

Material Information

Title: Characterization of Facial Pain Using an Operant Behavioral Testing Paradigm and Evaluating the Role of Transient Receptor Potential Channels in Facial Pain
Physical Description: 1 online resource (139 p.)
Language: english
Creator: Rossi, Heather
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008


Subjects / Keywords: anktm1, behavior, capscaisin, cci, channels, chronic, cmr1, constriction, face, icilin, injury, menthol, nociception, operant, orofacial, pain, potential, rats, receptor, resiniferatoxin, rtx, transient, trigeminal, trpa1, trpm8, trpv1, vr1
Neuroscience (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Abstract: Pain normally serves an adaptive purpose, but injury and prolonged inflammation can lead to changes that produce chronic pain. In order to better understand and treat chronic pain, scientific study must use methods that model and assess pain in a manner relevant to the human condition. Operant assessments of pain accomplish this by providing the animal with a conflict between a desired state and potential pain. In these studies, an operant method was used to establish the ability of rats to obtain a milk reward while stimulating their faces in the presence of a single stimulus or two stimuli. Similarly a person with chronic pain may forgo rewarding activities to avoid pain. This method was then used to assess the effects of transient receptor potential (TRP) channel agonists and of chronic constriction injury (CCI) on successful task completion. The TRP channel melastatin 8 is activated by cold, and another TRP ankyrin 1 may also be activated by noxious cold. Both of these receptors are co-expressed with TRP vanilloid 1, activated by noxious heat. Thus, any of these channels may affect the perception of cold in vivo. Without TRP channel manipulation or injury, successful task completion declines sharply with increasing heat, but only slightly with increasing cold. Menthol, a TRPM8 agonist, and icilin, a TRPM8/TRPA1 agonist, led to a decline successful task completion to 10 degrees and enhanced cold avoidance respectively, suggesting that cold allodynia was induced. Activation and lesion of TRPV1 hindered and enhanced successful task completion with painful cold respectively, supporting a role for TRPV1-expressing afferents in cold nociception. CCI is a common model of chronic neuropathic pain. CCI-treated rats exhibited impaired success with 10, 37 degrees, and rough stimulation, with differing duration and temporal patterns. Although aversive behaviors were observed, success was not significantly affected with 48 degree stimulation in these animals. Changes in TRPV1 expression and intensity of TRPM8 expression were also observed following CCI, supporting current evidence that these channels contribute to thermal allodynia accompanying neuropathic injury. These findings provide a foundation for clinically relevant means of evaluating pain mechanisms and analgesia.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Heather Rossi.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Neubert, John K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022896:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022896/00001

Material Information

Title: Characterization of Facial Pain Using an Operant Behavioral Testing Paradigm and Evaluating the Role of Transient Receptor Potential Channels in Facial Pain
Physical Description: 1 online resource (139 p.)
Language: english
Creator: Rossi, Heather
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008


Subjects / Keywords: anktm1, behavior, capscaisin, cci, channels, chronic, cmr1, constriction, face, icilin, injury, menthol, nociception, operant, orofacial, pain, potential, rats, receptor, resiniferatoxin, rtx, transient, trigeminal, trpa1, trpm8, trpv1, vr1
Neuroscience (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Abstract: Pain normally serves an adaptive purpose, but injury and prolonged inflammation can lead to changes that produce chronic pain. In order to better understand and treat chronic pain, scientific study must use methods that model and assess pain in a manner relevant to the human condition. Operant assessments of pain accomplish this by providing the animal with a conflict between a desired state and potential pain. In these studies, an operant method was used to establish the ability of rats to obtain a milk reward while stimulating their faces in the presence of a single stimulus or two stimuli. Similarly a person with chronic pain may forgo rewarding activities to avoid pain. This method was then used to assess the effects of transient receptor potential (TRP) channel agonists and of chronic constriction injury (CCI) on successful task completion. The TRP channel melastatin 8 is activated by cold, and another TRP ankyrin 1 may also be activated by noxious cold. Both of these receptors are co-expressed with TRP vanilloid 1, activated by noxious heat. Thus, any of these channels may affect the perception of cold in vivo. Without TRP channel manipulation or injury, successful task completion declines sharply with increasing heat, but only slightly with increasing cold. Menthol, a TRPM8 agonist, and icilin, a TRPM8/TRPA1 agonist, led to a decline successful task completion to 10 degrees and enhanced cold avoidance respectively, suggesting that cold allodynia was induced. Activation and lesion of TRPV1 hindered and enhanced successful task completion with painful cold respectively, supporting a role for TRPV1-expressing afferents in cold nociception. CCI is a common model of chronic neuropathic pain. CCI-treated rats exhibited impaired success with 10, 37 degrees, and rough stimulation, with differing duration and temporal patterns. Although aversive behaviors were observed, success was not significantly affected with 48 degree stimulation in these animals. Changes in TRPV1 expression and intensity of TRPM8 expression were also observed following CCI, supporting current evidence that these channels contribute to thermal allodynia accompanying neuropathic injury. These findings provide a foundation for clinically relevant means of evaluating pain mechanisms and analgesia.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Heather Rossi.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Neubert, John K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022896:00001

This item has the following downloads:

Full Text




2008 Heather Lynn Rossi 2


To my biological family and my Florida famil y, who have kept me going, Mr. Booker, my first biology teacher, and the rats. 3


ACKNOWLEDGMENTS I would like to thank my mentor (Dr. John Neubert) and my committee members (Drs. Robert Caudle, Charles Vierck, Jianguo Gu, and Shannon Holliday) for their assistance and advice. I would also like to th ank Wendi Malphurs, for my initi al training. Alan Jenkins performed the histology, made assessment s regarding TRPM8 and TRPV1 staining, and provided the images reported in Chapters 3 and 4. Dr. Indraneel Bhattacharyya also provided the blinded assessment of nerve inflammation and in jury reported in Chapter 4. I thank Jean Kaufmann, for her technical assistance and care of our rats, as well as the blinded assessments of unlearned behaviors reported in Ch apter 4. I thank Dr. Melanie Wexe l, who also participated in the central RTX study reported in Chapter 3. Fi nally, funding from Pfizer and the National Institute of Dental and Cranio facial Research (5R21DE016704-02) made this work possible. 4


TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 LIST OF OBJECTS.......................................................................................................................11 ABSTRACT...................................................................................................................................12 CHAPTER 1 INTRODUCTION................................................................................................................. .14 Clinical Significance.......................................................................................................... .....14 Benefits of Incorporating Operant Assessments of Pain into Scientific Research.................15 Role of Transient Receptor Potentia l Channels in Thermal Processing.................................17 Vanilloid Transient Receptor Potential Channels Feel the Heat.....................................17 Transient Receptor Potential Channels Make Cold Hurt................................................18 Encoding Thermal Sensations is a Group Effort.............................................................19 Using Operant Testing to Evaluate the Mech anisms of Normal and Neuropathic Pain........20 2 CHARACTERIZATION OF OPERANT RESPONSES TO THERMAL AND MECHANICAL FACIAL STIMULATI ON WITHOUT PAIN INDUCTION.....................22 Operant Methods for Evaluating Facial Pain..........................................................................22 Animals............................................................................................................................22 Single Stimulus Task.......................................................................................................23 Stimulus Preference Task................................................................................................24 Assessment of Mechan ical Sensitivity............................................................................24 Statistical Analysis..........................................................................................................2 5 Operant Behavioral Profile under Normal Conditions...........................................................26 Operant Responses to Thermal Stimuli...........................................................................26 Individual stimuli.....................................................................................................26 Thermal preference..................................................................................................27 Modulation of successful task completi on for each stimulus in the thermal preference task......................................................................................................28 Using the thermal preference task to condition aversion or preference...................29 Operant Responses to Mechanical Stimuli......................................................................31 Individual stimuli.....................................................................................................31 Mechanical preference.............................................................................................31 Discussion of Normal Operant Responses.............................................................................32 Single Stimulus Response...............................................................................................32 Thermal Preference and Unstimulated Time...................................................................34 5


Interaction Between Stimuli............................................................................................35 Conditioned Response in the Thermal Preference Assay................................................37 Operant Assessment of Mechanical Sensitivity..............................................................38 Conclusion..............................................................................................................................39 3 PHARMACOLOGICAL MANIPULATION OF OPERANT BEHAVIOR USING TRANSIENT RECEPTOR POTEN TIAL CHANNEL AGONISTS.....................................50 Methods..................................................................................................................................51 Animals............................................................................................................................51 Preparation and Administrati on of TRP Channel Agonists............................................51 Infraorbital and Intracisternal Injections.........................................................................52 Additional Behavioral Tests............................................................................................53 Evaluation of wet dog shaking induced by icilin.....................................................53 Capsaicin eye-wipe test............................................................................................53 Statistical Analysis..........................................................................................................5 4 Results.....................................................................................................................................54 Effect of the TRPM8 Agonist Ment hol on Operant Responses to Cold.........................54 Effect of the TRPM8/TRPA1 Agonist Icilin on Wet Dog Shaking................................54 Effect of the Icilin on Thermal Preference......................................................................55 Effect of the TRPV1 Agonist Capsai cin on Operant Responses to Cold........................56 Effect of TRPV1 Lesion with Resinife ratoxin on Operant Responses to Cold..............56 Peripheral versus Central RTX and Cold Sensitivity...............................................57 Effects of TRPV1 lesion by RTX on Cold Pain and Avoidance.............................57 Effects of TRPV1 lesion by RTX on Thermogenic and Nocifensive Responses induced by TRPM8 Agonists................................................................................58 Discussion of TRP Channel Manipu lation and Sensory Processing......................................59 Effects of Menthol on Cold Sensitivity...........................................................................59 Effect of Icilin on WDS...................................................................................................60 Effect of Icilin on Thermal Preference............................................................................61 A Role for TRPV1 in Cold Pain......................................................................................63 Conclusions.............................................................................................................................65 4 EFFECTS OF NEUROPATHIC PAIN ON OPERANT RESPONSES TO THERMAL AND MECHANICAL STIMULATION................................................................................74 Neuropathic Pain....................................................................................................................74 Methods..................................................................................................................................75 Induction of Neuropathic Pain, Monitoring Recovery, and Behavioral Testing.............75 Evaluation of Innate and Aversive Behaviors during Operant Testing...........................76 Evaluating the Effect of Surgical Treatm ent, Novelty, and Pregabalin on Thermal Preference....................................................................................................................77 Histology.........................................................................................................................77 Tissue preparation....................................................................................................77 Immunohistochemistry.............................................................................................78 Assessment of nerve inflammation and injury.........................................................79 Statistical Analysis..........................................................................................................7 9 6


Results.....................................................................................................................................80 General Observation of Immediate Po st-Surgical Recovery and Behavior....................80 Effect of Surgical Treatment on Operan t Responses to Cold, Neutral, and Hot Facial Stim ulation........................................................................................................81 Effects of Surgical Treatment on Oper ant Response to a Rough Stimulus.....................83 Effect of Surgical Treatment on Innate a nd Aversive Behaviors in the Presence of Thermal and Mechanical Stimulation..........................................................................84 Effect of Surgical Treatment on Nerve Inflammation and Injury...................................86 Effect of Surgical Treatment on TRPV 1 and TRPM8 expression at Two Weeks Post-Treatment.............................................................................................................87 Effect of Pregabalin and Gabapentin Treatment on Operant Responses to Cold Facial Stimulation in Surgically-Treated and Untreated Rats.....................................88 Effect of Surgical Treatmen t on Mechanical Preference.................................................89 Effect of Surgical treatment on Therma l Preference and the Influence of Task Novelty and Drug Treatment.......................................................................................90 Discussion of Neuropathic Pa in and Operant Behavior.........................................................91 Effects of CCI on Behavioral Responses to 10, 37C and Mechanical Stimulation.......92 Effects of CCI on Hot-Mediated Behaviors....................................................................94 Effects of CCI on Cold-mediated Behaviors and TRPM8 Expression...........................95 Cold Sensitivity and TRP Expression.............................................................................96 Effect of Drug Treatment on Cold Allodynia..................................................................98 Modulation of Preference After Injury and Affec tive Aspects of Pain...........................99 Conclusion............................................................................................................................100 5 FUTURE DIRECTION........................................................................................................117 Adaptation of the Assay to Evaluate Pain in Mice...............................................................117 Thermal Preference, Conditioned Aversion, and Drug Treatment.......................................119 Stimulus Novelty and Neuropathic Injury............................................................................120 General Conclusion............................................................................................................. .120 APPENDIX DOSE DETERMINATION OF PREGABALIN USED FOR TREATMENT OF CHRONIC CONSTRICTION INJURY...............................................................................122 Evaluation of General Activity by Rearing..........................................................................122 Administration and Dose Dete rmination of Pregabalin........................................................122 Pregabalin Dose Determination Based on Rearing and Alleviation of Capsaicin-Induced Heat Hyperalgesia.............................................................................................................123 LIST OF REFERENCES.............................................................................................................126 BIOGRAPHICAL SKETCH.......................................................................................................139 7


LIST OF TABLES Table page 1-1 Summary of activation ra nge, agonists, and antagonists for thermally activated Transient Receptor Poten tial (TRP) channels....................................................................21 2-1 Number of rats exhibiting a cold, hot, or no preference at the stimulus combinations tested (bold numbers indicat e the group preference).........................................................48 2-2 Influence of previous experience on freque ncy of start side an d preference with 10 and 10C stimuli 24 hours later.........................................................................................48 2-3 Influence of previous experience on freque ncy of start side and thermal preference 24 hours with 18 and 48C stimuli 24 hours later.............................................................48 2-4 Effect of start side on th e number of rats preferring soft rough, or neither mechanical stimulus....................................................................................................................... .......49 3-2 Rate of Wet Dog Shaking (WDS per minut e, mean SEM) calculated from observed counts over thirty-minute intervals fo llowing two doses of icilin or DMSO administered intraperitoneally (i.p. ) or intracisternally (i.c.m.).........................................73 4-1 Seven day post-operative recovery progress for surgical groups as indicated by number of rats with swe lling in the surgical area ra nging from severe to none..............115 4-2 Percentage of occurrence of behaviors and behavioral scores (mean SEM) with 10C, 37C, 48C, and rough stimulation for nave rats (n = 5, across multiple sessions)...........................................................................................................................115 4-3 Percentage of CCIand Sham-treated rats exhibiting head tilting (HT) or thermode wiping (wipe) with thermal and mechanical stimulation, or biting with mechanical stimulation preand post-operatively..............................................................................116 4-4 Percentage of surgically treated rats exhibiting soft, r ough, or no stimulus preference post-operatively (n = 30 per treatment)...........................................................................116 5-1 Experimental schedule for evaluating th e effect of different classes of drugs on conditioned aversion in the thermal preference task........................................................121 8


LIST OF FIGURES Figure page 2-1 Modification made to the ope rant testing apparatus to a ssess mechanical sensitivity......40 2-2 Example of licking bout determination ba sed on compressed outputs of raw data in Windaq or Labview............................................................................................................41 2-3 Effect of single thermal stim uli on operant task completion.............................................42 2-4 Distribution of time spent on the cold, hot, or off the thermode and percentage of licks spent at either thermode when cold stimuli were paired with 45 or 48C................43 2-5 Successful task completion when st imuli are paired or presented alone. ..........................44 2-6 Preference can be conditioned or abo lished by stimulus exposure 24 hours prior............45 2-7 Effect of smooth, soft, or rough mechanical stimulation on oper ant task completion......46 2-8 Effect of starting side on licks and duration spent with the soft and rough stimuli, or on total licks and unstimulated tim e when soft is preferred..............................................47 3-1 Effect of menthol (10%, s.c.) on operant responses to cold stimuli..................................66 3-2 Effect of two doses of icilin and DMSO on thermal preference with facial stimulation at 10 and 48C.................................................................................................67 3-3 Effect of the TRPV1 agonist capsaicin on operant responses to moderate (10C) and noxious cold (-4C)............................................................................................................ 68 34 Effect of peripheral resiniferatoxin (RTX) treatment on TRPV1 expression on in the trigeminal ganglia and TRPV1 function............................................................................69 3-5 Effect of peripheral TRPV 1 lesion by resiniferatoxin (RTX) on operant response to 10, 2C stimulation............................................................................................................70 3-6 Effect of central TRPV1 lesion by resini feratoxin (RTX) on operant response to -4C stimulation.................................................................................................................... ......71 3-7 Effect of central TRPV1 lesion by resini feratoxin (RTX) on thermal preference for 4 and 48C stimulation......................................................................................................72 4-1 Explanation of injury and inflammation scores used to assess the health of surgically treated infraorbital nerves................................................................................................102 4-2 Effect of surgical treatment on operant responses with 10, 37, and 48C stimulation ...104 9


4-3 Effect of surgical treatment on ope rant responses with rough mechanical stimulation.................................................................................................................... ....105 4-4 CCI-treated rats exhibit aversive behaviors towards the stimulus not observed in Sham-treated rats.............................................................................................................1 06 4-5 Effect of surgical treatment on innate a nd aversive behavior scores with thermal and mechanical stimuluation..................................................................................................107 4-6 Presence of suture and quantification of inflammation in surgically treated infraorbital trigeminal nerves...........................................................................................108 4-7 TRPV1 immunoreactivity in trigeminal ganglia from na ve and surgically treated rats....................................................................................................................................109 4-8 TRPM8 immunoreactivity in trigeminal ga nglia from nave and surgically treated rats....................................................................................................................................110 4-9 Effect of drug treatment on operant behavior in surgically treated and nave rats with 10C stimulation at two weeks post-surgery....................................................................111 4-10 Effect of surgical treatment and starti ng stimulus on mechanical preference, total licks, and time spent unstimulated...................................................................................112 4-11 Effect of surgical treatment and experi ence on thermal preference and unstimulated time........................................................................................................................... .......113 4-12 Effect of surgical treatment and expe rience on success for each stimulus in the thermal prefer ence taks....................................................................................................114 A-1 Determination of lowest analgesic dose of pregabalin withou t significant effects on general activity, measured by rearing behavior...............................................................125 10


LIST OF OBJECTS Object page 2-1 Video clip of rats performing the single stimulus operant task.........................................49 4-1 Video clip of aversive head tilting beha vior exhibited by a CCI-treated rat at postoperative day 10, with 10C stimulation.........................................................................116 4-2 Video clip of normal operant behavior exhibited by a sham-treated rat at postoperative day 10, with 10C stimulation.........................................................................116 11


Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHARACTERIZATION OF FACIAL PA IN USING AN OPERANT BEHAVIORAL TESTING PARADIGM AND E VALUATING THE ROLE OF TRANSIENT RECEPTOR POTENTIAL CHANELS IN FACIAL PAIN By Heather Lynn Rossi December 2008 Chair: John Neubert Major: Medical Sciences Neuroscience Pain normally serves an adaptive purpose, but injury and prolonged inflammation can lead to changes that produce chronic pain. In or der to better understand and treat chronic pain, scientific study must use methods that model a nd assess pain in a manner relevant to the human condition. Operant assessments of pain accomplis h this by providing the animal with a conflict between a desired state and potenti al pain. In these studies, an operant method was used to establish the ability of rats to obtain a milk rewa rd while stimulating thei r faces in the presence of a single stimulus or two stimuli. Similarl y a person with chronic pain may forgo rewarding activities to avoid pain. This method was then us ed to assess the effects of transient receptor potential (TRP) channel agonists and of chronic constriction in jury (CCI) on successful task completion. The TRP channel melastatin 8 is ac tivated by cold, and another TRP ankyrin 1 may also be activated by noxious cold. Both of thes e receptors are co-expressed with TRP vanilloid 1, activated by noxious heat. Thus any of these channels may a ffect the perception of cold in vivo Without TRP channel manipulation or injury, successful task completion declines sharply with increasing heat, but only slightly with increasing cold. Menthol a TRPM8 agonist, and icilin, a TRPM8/TRPA1 agonist, le d to a decline successful task completion to 10 degrees and 12


enhanced cold avoidance respectively, suggesti ng that cold allodynia was induced. Activation and lesion of TRPV1 hindered and enhanced suc cessful task completion with painful cold respectively, supporting a role for TRPV1-expre ssing afferents in cold nociception. CCI is a common model of chronic neuropathic pain. CCI-t reated rats exhibited impaired success with 10, 37 degrees, and rough stimulation, with differing duration and tempor al patterns. Although aversive behaviors were obs erved, success was not significan tly affected with 48 degree stimulation in these animals. Changes in TR PV1 expression and intensity of TRPM8 expression were also observed following CCI supporting current evidence that these channels contribute to thermal allodynia accompanying neuropathic injury. These findings provide a foundation for clinically relevant means of evaluating pain mechanisms and analgesia. 13


CHAPTER 1 INTRODUCTION Clinical Significance Pain normally serves as a protective function by alerting us to the presence of an external, noxious stimulus or by bringing some imbalance in our internal state to our attention (i.e. aches experienced during infection). Howe ver, disease and injury can l ead to changes in the nervous system that produce and maintain pathological, chronic pain. Chronic pain can be highly distressing and excruciating, with either no external cause, or it can be triggered by normally pleasant sensation (e.g. the gentle touch of a love d one). Treatment to mitigate these effects can be expensive and associated with severe side effects that may be as intolerable as the pain. Over time, some medications may lose their effectiv eness. These aspects of chronic pain have motivated researchers to understand how these features develop so that we may improve treatment techniques. In order to better address the needs of the pa tient, basic research must use models and methods of pain measurement that more accurately reflect the problems we are trying to overcome. Chronic pain in the orofacial region is e xperienced by 20 to 25% of the United States, including patients suffering from a broad spectrum of disorders such as trigeminal neuralgia, temoporomandibular joint (TMJ) disorders, and headaches (e.g., migraine), as well as pain associated with trauma to the face and mouth. De spite the clinical preval ence of orofacial pain, basic research has primarily used the rodent hind paw and sciatic nerve as the standard model for neuropathic pain, with few studies focusing on pain involving the face or trig eminal nerve. It is assumed that the processing of cutaneous sensations is equivalent in the sciatic and trigeminal pathways, but differences in re ceptor expression and function ha ve been noted (Kayser et al., 2002; Kobayashi et al., 2005) that could have a measurable effect on the development and 14


treatment of chronic pain states in these areas. Also, c onditions such as chronic headache and trigeminal neuralgia are unique to the trigeminal system. To provide the best assessment and treatment of chronic pa in within the orofacial region, basic science mu st include a specific assessment of trigeminal sensory processing. Additionally, assessment of pain in experimental animals ha s relied heavily on reflexive withdrawal of the hindlimbs from either a hot or punctate mechanic al stimulus. The withdrawal reflex represents the most immediate and basi c response to a damaging stimulus, often preceding conscious perception of pain. This may not be the most appropriate method for evaluating cold sensitivity in general, or neuropathic pain sp ecifically. Whereas heat can be immediately damaging with brief exposure, prolonged exposure to cold is required before injury occurs, thus the drive for reflexive withdrawal from localized cold stimulation is likely not as pronounced as for damaging heat. With respect to neuropathic pai n, the capacity of this chronic pain to prevent a patient from enjoying and functioning in daily life is what drives the patient to seek treatment. Thus, in the laboratory, assays that measure the capacity of pain to hinder normal function or rewarding task would more accurately reflect the patient experience. Operant assays can accomplish this goal. Benefits of Incorporating Operant Assessmen ts of Pain into Scientific Research Behavioral evaluation of pain has relied pr imarily on withdrawal thresholds (Hargreaves et al., 1988; Imamura et al., 1997), which are spina lly-mediated or unlearned behaviors such as isolated grooming (Deseure and Adriaensen, 2002; Deseure and Adriaensen, 2004), which are brainstem-mediated (Mauderli et al., 2000). While such measures have been important for laying the foundations of our understandin g of pain, they can be evoked in decerebrate animals (Woolf, 1984) and thus do not directly eval uate higher order cortical pro cessing of pain. It has become increasingly evident that chronic, maladaptive pain states are not simply the result of changes at 15


the peripheral and spinal level of nociception, but are maintained by changes at the cortical level (de Leeuw et al., 2005; Seifert and Maihofner, 20 08). Therefore, the evaluation of pain and analgesia in experimental animal s should incorporate methods that directly evaluate cortical processing by requiring that the animal make a decision about its environment. Operant pain assays rely on an animals natural avoidance of unpleasant or noxious stimuli. Unlike many reflex-driven tests, the anim al is allowed to move freely in the testing box. This eliminates stress associated with restraint (M auderli et al., 2000), or anticipation of painful stimulus that the animal can see approaching, wh ich is particularly probl ematic for testing the face. Restraint stress has been shown to enha nce nociceptive responses (Gameiro et al., 2005; Gameiro et al., 2006) and increase operant escape from painful h eat (King et al., 2003; King et al., 2007). Stress can therefore enha nce pain even if that is not the intent of the experimenter. In addition to fully evaluating all levels of nociceptive processing and reducing potential stress, operant methods also remove some of the bias that can aris e in the execution and evaluation of reflexive withdrawal or unlearned behaviors as responses to pain. The influence of experimenter bias in basic sc ientific research has been co mmented on (Eisenach and Lindner, 2004) and demonstrated through stat istical analysis of studies conducted between laboratories (Crabbe et al., 1999) and of expe rienced individuals within a single laboratory (Mogil et al., 2006). Even when careful effort is made to blind the experimenter to treatment group, reflex and unlearned responses require that the assessi ng scientist make a judgment regarding what responses should be considered nociceptive. Such judgments vary depending on experience and personal interpretation. Operant methods for assessing pain remove the experimenter from the application of the pain stimulus and often involve a permanent reco rd of the behavioral response, which can be re-examined as desired. 16


Due to these limitations, we have developed an operant assay for evaluating facial pain. This method assesses the animals ability and willingness to place its face in contact with a stimulus to obtain reward. The more pain the an imal experiences, the less successful it should be at performing this task, which has been valida ted (Neubert et al., 2005 ). This behavioral assessment provides a means of examining the role of various molecular mediators in thermal processing in vivo as well as changes that occur following neuropathic injury. Role of Transient Receptor Potential Channels in Thermal Processing Vanilloid Transient Receptor Potential Channels Feel the Heat In the last decade, several members of th e transient receptor potential (TRP) receptor family have been identified as molecular mediator s of thermal stimuli (see Table 1-1), which has considerably enhanced our understanding of peri pheral sensory processing. The most thoroughly studied of these, and the first to be characteri zed was TRP vanilloid 1, a critical molecule for encoding noxious heat (Caterina et al., 1997; Ca terina et al., 1999; Cate rina et al., 2000). TRPV1 also mediates hyperalges ia following inflammation (Caterina et al., 2000; Neubert et al., 2008; Wexel, 2008), incisional pain (Pogatzki-Zahn et al., 2005), and contributes to neuropathic pain (Walker et al., 2003; Levine and Alessandri-Haber, 2007). Additional members of the vanilloid family, TRPV2 (Caterin a et al., 1999), TRPV3 (Xu et al., 2002), and TRPV4 (Guler et al., 2002), are also important for mediating respon ses to painful heat and warmth respectively. Discrepancies noted between the in vitro activation range of TRPV1 (>42C) and the impairment observed in TRPV1 knock-out mice (>49C) (Caterina et al., 2000; Davis et al., 2000) led to further speculation a nd investigation regarding the encoding of heat sensation and pain. It has been suggested that a small numbe r of TRPV1 negative nociceptors may be needed to produce a withdrawal response (Julius and Ba sbaum, 2001) and recently demonstrated that TRPV1 provides the primary heat signaling via lami na V, but only part of the input to lamina I 17


(Eckert et al., 2006). Additional investigation indicates that a subset of heat responsive nociceptors rely on TRPV1/2 independent mech anisms (Woodbury et al., 2004). The encoding of heat therefore involves a num ber of molecular mediators, in cluding the four vanilloid TRP channels and potentially others. The expression patterns of these molecules, as well as the convergent inputs they provide to different le vels of the dorsal horn encode the nuanced experience of warmth and painful heat. Transient Receptor Potential Channels Make Cold Hurt The molecular mediators of cold have been a hot topic in pain a nd sensory processing for the last six years, fueled by the discovery and ch aracterization of two putat ive cold receptors in the TRP channel super family: TRP melastatin 8 (TRPM8) and TRP ankyrin 1 (TRPA1). TRPM8 is activated by cooling compounds, such as menthol, and by temperatures at and below 25C (McKemy et al., 2002; Peier et al., 2002). TRPM8 was never disputed a molecular mediator for innocuous cold perception. However, the channel is expressed in cells with both nociceptive and non-nociceptive characteristics (X ing et al., 2006; Dhaka et al., 2007; Takashima et al., 2007; Dhaka et al., 2008). Initial studies led some to sugge st that TRPM8 served primarily in innocuous cold perception and that some other molecule or molecules encoded painful cold stimulation (McKemy, 2005), but subsequent work has cemented a role for TRPM8 in cold pain (Xing et al., 2006; Colburn et al ., 2007; Dhaka et al., 2007; Xing et al., 2007; Dhaka et al., 2008). Another receptor characterized a year late r appeared to match the expectations for a molecular transducer of painful cold. TRPA1 (formerly ANKTM1), was shown to be activated by cooling, with a lower threshol d in the nociceptive range (<18 C) (Story et al., 2003), in menthol insensitive (TRPM8 negative) neurons (Bandell et al., 2004). TRPA1 is expressed exclusively in cells with nociceptive characteris tics that also express TRPV1 (Kobayashi et al., 18


2005). It is also activated by a number of punge nt compounds that produ ced burning or pricking sensations, such as icilin, mu stard oil, and cinnamon aldehyde (Bandell et al., 2004; GarciaAnoveros and Nagata, 2007). Others fa iled to replicate act ivation of TRPA1 in vitro by cold (Jordt et al., 2004; Nagata et al., 2005). This discrepancy has been attributed to the observation that TRPA1 activity is dependent on intracellular calcium (Doerner et al., 2007) and that cooling in heterologous expression systems can increase local calcium and coincidentally activate the TRPA1 (Zurborg et al., 2007). However, direct cooling activation of TR PA1 was demonstrated at physiological resting membra ne potential using a calcium free, inside-out patch clamp (Sawada et al., 2007), providing further in vitro evidence that TRPA1 is activated directly by painful cold. Encoding Thermal Sensations is a Group Effort Studies of knock-out mice conclusively dem onstrated that the l ack of TRPM8 profoundly impaired irritation induced by acetone, withdrawal from intensely cold stimuli, discrimination between cool stimuli, cold avoidance, and ici lin-induced shaking (Colburn et al., 2007; Dhaka et al., 2007), indicating that TRPM8 is required for both cold perception and nociception. In contrast, impairments in cold responses were not observed in TRPA1 knock-out mice by one laboratory (Bautista et al., 2006), and only to a small degree in a modest sample of females knock-out mice by another laboratory (Kwan et al., 2006). However, the findings in TRPA1 knock-out mice do not necessarily rule out a role for this receptor in encoding certain aspects of painful co ld. TRPM8 and other co ld activated molecules likely maintain the normal cold perception in TRPA1 knock-out mice. There is evidence to suggest that TRPA1 acts as a mechanoreceptor (Kwan et al., 2006; Cahusac and Noyce, 2007; Kindt et al., 2007). It is possible that TRPA1 mediates the pr icking or tingling sensations that accompany dramatic cooling, whether by direct activation from cooling or by coincidence 19


detection of another channel that mediate activation via increase d local calcium concentration. The burning aspects of cold pain are likely medi ated by fibers that co-expression of TRPM8 and TRPV1 (Okazawa et al., 2004; Kobayashi et al ., 2005; Xing et al., 2006; Dhaka et al., 2007). There are also other cold nociceptors not expr essing either TRPA1 or TRPM8 (Babes et al., 2002; Madrid et al., 2006; Munns et al., 2007) that contribu te to the encoding of cold pain. Thus, like heat, there are a myriad of molecular play ers with differing patter ns of expression and connectivity that are responsible to for the various nuances of cold perception. Using Operant Testing to E valuate the Mechanisms of Normal and Neuropathic Pain Operant behaviors provide an additional and clinically relevant means of evaluation thermal and mechanical sensitivity, particularly co ld sensitivity. This be havioral testing method in combination with pharmacological manipulat ion of TRP channels has provided us with additional insights regarding the processing of cold stimuli. Finally, we demonstrated that neuropathic pain can be measured using this me thod, and that it can provide insights regarding the role of peripheral, cognitiv e, and affective factors in pain-related decision making. The results reported here support recent work related to the mechanisms of cold nociception and sensation. These findings also lay a foundation for future applications of operant pain assessment to study the mechanisms and treatment of pathological pain. 20


Table 1-1. Summary of activa tion range, agonists, and antag onists for thermally activated Transient Receptor Pote ntial (TRP) channels. TRP channel Activating stimuli Agonists Antagonists TRPV1 (VR1) 42-52C, acidity Capscaicin, RTX, ethanol (Levine and Alessandri-Haber, 2007), camphor (Xu et al., 2005), 2-APB (Colton and Zhu, 2007), anandamide (Ross, 2003) I-RTX (Jhaveri et al., 2005), capsazepine, BCTC (Behrendt et al., 2004), AMG0347 (Steiner et al., 2007) TRPV2 (VLR1) >52C 2-APB (Colton and Zhu, 2007), cannabidiol (Qin et al., 2008) Gadolinium Lanthanum (Leffler et al., 2007) TRPV3 >33C 2-APB (Colton and Zhu, 2007), camphor (Moqrich et al., 2005), carvacrol, eugenol, thymol (Levine and Alessandri-Haber, 2007) TRPV4 27 42C, hypotonicity, acidity, mechanical 4*PDD (Strotmann et al., 2003) TRPM8 (CMR1) <25C Menthol, icilin (2 M) (McKemy et al., 2002) Capsazepine, BCTC (Behrendt et al., 2004) TRPA1 (ANKTM1) <18C (debated), mechanical Bradykinin, mustard oil, icilin (50 M), cinnamaldehyde, menthol (<100 M) (Garcia-Anoveros and Nagata, 2007) Menthol (>100 M) (Karashima et al., 2007) BCTC = 4-(3-chloro-pyridin-2-yl)-piperazine-1-ca rboxylic acid (4-tert-butyl-phenyl)-amide, 2APB = 2-aminoethoxydiphenyl borate, 4*PDD = 4,*-phorbol 12,13-didecanoate 21


CHAPTER 2 CHARACTERIZATION OF OPERANT RE SPONSES TO THERMAL AND MECHANICAL FACIAL STIMULATION WI THOUT PAIN INDUCTION Operant assessments of cutaneous sensation a nd pain have been employed in recent years with respect to hindpaw stimulation. These a ssessments present thermal stimuli individually (Mauderli et al., 2000; Vierck et al., 2004; Vierck et al., 2005; Vierck et al., 2008), in pairs presented simultaneously (Lee et al., 2005; Vierck et al., 2005; Jabakhanji et al., 2006; Walczak and Beaulieu, 2006; Vierck et al., 2008), or as a thermal gradient (Lee et al., 2005). However, none have reported an operant method for evaluating facial pain, regardless of stimulus type. Also, no operant methods currently exist to evaluate mechanical sensitivity in any part of the body. In this chapter, the operant method for evaluating facial pain is described, including modifications made to evaluate mechanical sensit ivity. This method evalua tes a rodents ability to obtain a milk reward while in contact with a stimulus. The rodent may be presented with one stimulus, or may choose between two stimuli to ob tain the reward. The stimulus could be any temperature within a range from -4 to 52C or one of two textured mechanical stimuli. We used the single stimulus task to characterize operant re sponses of male rats to individual thermal and mechanical stimuli in the absence of pain manipulation. We also used the preference task to assess affective aspects of pain processing and de termine what effect previous experience might have on preference. These findings provide the basis for evaluating the effects of TRP channel modulation (Chapter 3) and neur opathic injury (Chapter 4). Operant Methods for Evaluating Facial Pain Animals Male hairless Sprague-Dawley rats (Charles Ri ver, Raleigh, NC, 5-8 weeks old) were used for all experiments described in this and the following chapters, unless otherwise specified. Three to five rats were housed in large cages with enrichment (as described in (Rossi and 22


Neubert, 2008). A standard 12hour light/dark cycle was maintained and rats were allowed access to food and water ad libitum when not bei ng tested. Weights were recorded every week to monitor general health. Animal testing pr ocedures and general handling complied with the ethical guidelines and standards established by the Institutional Animal Care & Use Committee at the University of Florida (Institut e of Laboratory Animal Resources, 1996). Single Stimulus Task Facial testing was conducted us ing a reward-conflict paradigm as described previously (Neubert et al., 2005). Briefly, rats were trained to dri nk sweetened condensed milk while making facial contact with a si ngle thermode and lick-tube ( Object 2-1 ). During the training period (approximately 2 weeks) baseline intake was recorded and box distance from the sipper tube was gradually increased. Ra ts were considered ready for stimulus testing once they were able to drink an average of 10 grams of milk with their faces in correct contact with the thermodes, set at 37 C. Following training, rats were either tested at a range of thermal stimuli to establish a stimulus response pattern, or they were tested at a particular stimulus to establish a baseline of behavior before the induction of a pain state, drug ad ministration, or a combination of the two. To maximize thermal stimulus contact, the facial testing region for each animal was depilitated under light isofluoran e anesthesia (inhalat ion, 2.5 %) once a week or as needed. Rats were also fasted overnight for 13-15 hrs to motivate performance, but no more than three times per week. All stimulus testing was conducted within an ambient temperature of 20 to 26C. For every 20 or 30-minute testing session, milk intake was measured and raw data recorded using Windaq software and hardware. Th is raw data was transformed into a numerical form using a custom-written subroutine in La bview (courtesy of Dr. Charles Widmer). The numerical data was used to calculate the number of licks and the number of stimulus contacts, 23


from which the success ratio is derived. The su ccess ratio is the number of licks (successful attempts) divided by the number of stimulus contacts (total attempts). Stimulus Preference Task The thermal preference of the rats was record ed as previously desc ribed (Rossi et al., 2006). Rats were trained in the single task condit ion, as described above, and initially placed in the thermal preference apparatus with both therm odes set at 37C to allow them to become accustomed to this new task. A second such session was recorded to ensure rats did not demonstrate a side preference. Rats were able to move freely from one side of the compartment to the other and explore both thermodes at will. Unless otherwise sp ecified, the start side was not controlled by the experimenter, but was recorded as part of the offline data analysis. When tested repeatedly with a combination of stimuli, the hot a nd cold (or soft and rou gh) sides of the testing chamber were alternated to prevent learned aversion or pref erence for one side of the box. Following data acquisition, raw data files for each side of the thermal preference box were examined together to determine the startin g side and the number of switches made between sides. The raw data was transformed to a numerical form and licks, stimulus contacts, duration, and the success ratios for each side of the preferen ce apparatus were determined as in the single stimulus task. Additionally, beca use the rats have a third opt ion of abstaining from task completion to avoid either stimulus, the time spent unstimulated was determined by subtracting the total stimulated time from the total testi ng time (1200s). Licks, stimulus contacts, and duration could be expressed either as raw data or as a percentage of the total. Assessment of Mechanical Sensitivity We adapted our existing thermal testing a pparatus by placing a Velcro-covered plastic sleeve over the thermodes to serve as a rough or soft mechanical stimulus (Figure 2-1). The number of licks made in contact with the mechan ical stimulus were recorded and calculated in 24


the same manner as the thermal testing analysis. However, contact with the mechanical stimulus cannot be directly measured because the sleeve blocks the completion of the electrical circuit needed to register facial contac t. Therefore, stimulus contacts were determined indirectly by compressing the raw licking data and counting th e number of licking bouts (Figure 2-2). Slight differences were occasionally noted between th e appearance of the compressed windaq file and the visual output in Labview (F igure 2-2), so for the sake of consistency all bouts were counted from the compressed windaq file. Our ratio for assessing successful task completion was then calculated by dividing licks (successful attempts) by licking bouts (approximate total contacts or attempts). For comparison of these mechanical s timuli with smooth metal, this analysis method was also applied to lickin g data obtained with 37C smooth metal stimulus. Mechanical preference was also evaluated by covering the thermodes on each side of the preference box with rough and soft stimuli. For each stimulus, outcomes were calculated as described above. Side switching was also determined in off-line analysis, but may underestimate absolute switching because mechanical stimulus contacts without licks could not be recorded. Unstimulated duration was also calculated by subtracting the sum of the licking durations on both stimuli from the total testing time (1200s). In some sessions the start side was controlled by the experimenter. In these inst ances the rat would be placed in the box on one side with the middle barrier in place. The barrier was removed on ce the rat made contact with the stimulus and after this point the rat was allowed to explore the preference apparatus freely. Statistical Analysis All statistical analyses described herein were performed using SPSS (SPSS, inc. v.14 or 16). One-way analysis of vari ance (ANOVA) was used to compare the effects of individually presented thermal and mechanical stimuli on operant behavioral outcome measures. Because the individual thermal data was represented by a large sample size consisting of multiple cohorts 25


tested repeatedly at different times, data was submitted to a box plot analysis, and outliers were removed on an outcome by outcome basis, where an outlier is defined as any datum greater or less than one and a half times the interquartile range. Of the 1024 data points, 2, 9, and 8% were identified as outliers and removed from licks, stimulus contacts and success ratio respectively. For both thermal and mechanical preference data, paired t-tests were used to determine the difference between the percentage of licks on eac h stimulus and repeated measures ANOVA was used to determine significant differences in the pe rcentage of time spent on either stimulus or unstimulated. One-way ANOVAs were used to co mpare licks and time spent on either stimulus or off the thermode across stimulus pairs. For thermal preference data only, success ratios for the single stimulus condition versus pairing with another stimulus were also evaluated using one way ANOVA. To evaluate the effect of previ ous experience on time with either stimulus or unstimulated, repeated measures ANOVA was used. Post-hoc comparisons were made using Tukeys test for all one way ANOVAs, and the least squared differences test (LSD) for repeated measures ANOVAs. Statistical signifi cant was set to p<0.05 for all analyses. Operant Behavioral Profile under Normal Conditions Operant Responses to Thermal Stimuli Individual stimuli Rats were tested with a range of therma l stimuli from noxious cold (-4C) to noxious heat (52C). There was a significant e ffect of temperature on licks (F9, 1006 = 50.666), stimulus contacts (F9, 933 = 73.864), and success ratios (F9, 955 = 104.301, p<0.001 for all outcomes; Figure 2-3). Cold and cool stimuli produce a modest reduction in licks and success ratios, relative to neutral (37C) or warm (42C) stimulation. Cold stimuli did not significantly increase stimulus contacts. In contrast, hot stimuli produce a sharp decline in licks and success ratio as temperature increases, which is accompanied by an increase in stimulus contacts at the most 26


noxious hot temperatures (48 and 52C). These data indicate that successful task completion is hindered more by noxious heat than noxious cold With noxious heat, rats must make frequent contacts with the stimulus in order to obtain the milk reward, while this strategy is not necessary with cold stimulation. However, the fact that ther e is a reduction in successful task completion in the presence of cold stimulation i ndicates that cold is more aversive than neutral (37C) or warm (42C) stimuli. Thermal preference Although individuals may exhibit a side preference when expos ed to a pair of neutral temperatures, this preference is not consistent across testing sessions and when all rats outcomes are averaged for each stimulus no side bias is obser ved, as previously reported (Rossi et al 2006). With few exceptions, individual rats exhibit a te mperature preference each time they are exposed to hot and cold pair of stimuli (Table 2-1). We te sted a range of cold stimuli paired either with 45 or 48C. Most individuals preferred 45C regard less of the cold stimulus paired with it. In contrast, when non-noxious cold stimuli (10, 18, and 24C) are paired with 48C, most rats prefer the cold stimuli. When 48C is paired with -4C, noxious cold, most rats prefer 48C. The rats individual preferences are predicti ve of the mean group licking preference and time distribution for each stimulus combination. For all pairs including 45C as the hot stimulus, 45C is strongly preferred, as indicated by percen tage of testing time a nd percentage of total licks (Figure 2-4 A, B). There wa s no significant difference in the percentage of time or licks spent on the 45C stimulus or the cold stimulus when compared across stimulus pairs. However, rats did spend slightly more time at 24C than at 18 or 10C and as a consequence unstimulated time is significantly smaller when 24C is the cold option. As indicated by individual preferences, non-noxious cold stimuli (10, 18, 24C) are preferred when paired with noxious 48C. In cont rast, when 48C is paired with a noxious cold 27


stimulus (-4C), 48C is preferred (Figure 2-4 C, D). For all pairs, there was no significant difference in the percentage of time spent on the hot side (10-15%), regardless of whether or not the cold stimulus was noxious (Figure 2-4 C). In contrast, percentage of time and licks on the cold stimulus, are significantly greater for non-n oxious temperatures than -4C (Figure 2-4 C, D). Conversely, the percentage of licks on 48C is significantly lower when 48C is paired with non-noxious cold stimuli (Figure 2-4 D). As was the case for 45C, the percentage of unstimulated time is modulated by the intensity of the cold stimulus, with the lowest unstimlated time spent with the least intense stimulus (24C) and the greatest unstimulated time spent with the coldest stimulus (-4C) (Figure 2-4 D). Taken together, these data indicate that the percentage of time on the hot stimulus is consistent across stimulus pairs. In the case of 48C, this is even true when the preference switches from cold to hot (10, 18, 24 C versus -4C). These data al so indicate that the percentage of time spent on any stimulus also decreases with increasing intensity. As a consequence, the percentage of time spent unstimulated is the greatest when both stimuli are very intense (-4 and 48C) and the lowest when both stimuli are the least intense (24 and 45C). Modulation of successful task completion for ea ch stimulus in the thermal preference task Success at 45C was not significantly changed by being paired with any of the cold stimuli, although it was slightly in creased (Figure 2-5A). In contrast, success at all of the cold stimuli paired with 45C was decreased relative to when the stimuli are presented alone (Figure 2-5B). Interestingly, the success ratios for 10, 18, and 24C are significantly different from each other in the presence of 45C, but not when thes e cold stimuli are presented individually, which we have demonstrated previously regarding individual cold stimuli (Rossi et al., 2006). In contrast, success at 48C is modulated by cold in an intensity dependent fashion. Success at 48C is significantly decreased when paired with -4 or 10C, as compared when 48C 28


is presented alone, and no diffe rent when paired with 18 or 24C (Figure 2-5C). Conversely, success at -4 and 10C are lower with 48C than when presented individu ally (Figure 2-5D). Success at 18 and 24C are greater with 48C than alone, but only significantly so for 18C (Figure 2-5D). As with 45C, an effect of cold stimulus intensity is more apparent when paired with 48C than when presented alone. Taken together, these data indicate that pa iring stimuli can modul ate successful task completion at either stimulus as compared to wh en that stimulus is presented alone. While the effects observed at 45C can be explained by the strong preference for 45C, the effects at 48C suggest that there is an intera ction between the two stimuli inde pendent of preference that can modulate successful task completion at each stimulus. It is possible that stimuli 10C and below cross sensitize with noxious heat, resulting in less success than when stimuli are presented individually. Using the thermal preference task to condition aversion or preference In addition to characterizing thermal prefer ence, time distribution, and the effect of stimulus interactions on successful task comp letion, we also sought to determine if prior experience in the thermal testing apparatus coul d affect thermode pref erence the following day. Rats (n =10) were tested first at 37 and 37C, exhibiting no side preference and spending about 60% of the testing time unstimulated (Figure 2-6A ). They were then tested at 42 and 10C, exhibiting a strong preference for the left, 42C thermode and spending about 45% of testing time unstimulated. This preference was not the result of a bias in the starting side (Table 2-2). Twenty four hours later, the ra ts were tested at 10 and 10C. Rather than exhibiting no group preference, as would be expected with two equa l stimuli, rats spent significantly more time on the left thermode than the righ t (Figure 2-6A). The time spent on the right thermode with 10 and 10C was also significantly lower than the time spent on the right thermode at 37 and 37C, 29


while there were no significant differences be tween time on the left thermode, or unstimulated time for either same stimulus pair. The rats also exhibited a bias for star ting on the left thermode (Table 2-2). Taken together, these findings s uggest that the left-thermode preference (right thermode avoidance) of the previous day was sufficient to produce a conditioned preference when one should not have occurred. We were also able to use previous experi ence to obscure subsequent preference. Rats were first tested at 52 and 18C, which induced a preference for 18C. Again, a bias in start side was not observed (Table 2-3). We generally obs erve that simply switc hing the hot and cold stimuli the following day does not have a signifi cant effect on thermal preference. However, changing one of the stimuli by a few degrees c ould have a significant impact on behavior. Therefore, the next day the hot and cold stimuli were switched, the cold stimulus was maintained at 18C, while the hot stimulus was decreased to 48C. Rather than exhibiting a preference for 18C as indicated above, the pr eference was abolished. After be ing exposed to noxious heat on the left thermode, fewer rats started on the left side. The next day the hot and cold stimuli were switched and 18C was paired with 48C. Rather than exhibiting a preference for 18C as indicated above, the preference was abolished. Rats also spent significantly less time on the left thermode when it was set to 18C as compared both to the 18C right side thermode of the previous day, and the 18C left side thermode two days later (Figure 2-6B). Unstimulated time was also significantly greater when preference fo r 18C was abolished than when it was present. Fewer rats also started on the left side (Table 2-3). As repor ted above, there was no significant difference in time spent on the 48C stimulus when preference was abolished versus present. There was also no significant difference in time spent on 18C when it was paired with 52C as compared to when it was paired with 48C and pr eference was intact. Take n together, these data 30


indicate that previous experi ence can condition avoidance of one thermode that leads to avoidance of that side and thus an apparent lack of preference. Operant Responses to Mechanical Stimuli Individual stimuli Rats were tested with soft and rough mechani cal stimulation, as well as a neutral thermal stimulation (37C) to compare the responses to the mechanical stimuli with that of a smooth metal thermode. There were no effects of mechanical stimulation on licks (F 2,97 = 2.109, p = 0.127) or bouts (an indirect meas ure of stimulus contacts; F 2,97 = 2.453, p = 0.091; Figure 2-7 A,B). However, there was a significant effect of mechanical stimulation on licks per bout (an indicator of successful task completion; F 2,97 = 8.275 p<0.001; Fig.2-7 C). The greatest success occurred with the soft stimulus and no signifi cant difference in success was observed between smooth metal and the rough stimulus (Figure 2-7 C). Taken together, these data indicate that rats find the soft mechanical stimulus the least aversi ve or most pleasant, even when compared to smooth metal. Mechanical preference Preference for mechanical stimulation in nave rats was dependent on the starting stimulus. The majority of rats starting volunt arily on the soft stimulus exhibited a strong preference for this stimulus. In contrast, when rats began with the rough stimulus, no particular preference was favored by the group (Table 2-4). When a soft preference wa s exhibited, rats that began on the rough stimulus spent more licks and time on the rough stimulus as compared to those that started on the soft s timulus, while licks and time on the soft stimulus were not affected by start side (Figure 2-8) As a consequence, total licks we re slightly higher and unstimulated duration was lower when rats started with the rough stimulus than with the soft stimulus (Figure 2-8). These findings indicate that starting side can influence mechanical preference. When soft 31


is the first stimulus encountered there is little drive to sample the rough stimulus. When the rough stimulus is the first encountered, it may be mildly aversive, motivating exploration of the other stimulus. However, it is not so aversive th at the soft preference is consistently maintained across multiple testing sessions. The decreased unstimulated durati on and the increase in licks observed when rough is the starti ng stimulus may suggest that th e rats sample both sides more frequently to arrive at the soft preference. In support of this idea, rats starting with soft stimulation had an average of one successful swit ch, while when rough stimulation is first they average three successful switches. Discussion of Normal Operant Responses In this chapter, we demonstrated that opera nt task completion is strongly inhibited with increasing heat intensity, but only somewhat hi ndered by increasing cold intensity. However, when thermal stimuli of equivalent intensity ar e paired, the aversiveness of cold stimuli is apparent. Noxious cold and hot stimuli are also able mutually hinder successful task completion, suggesting that these stimuli activate a comm on nociceptor population and pain pathway. A novel, operant methodwas also used to assess me chanical sensitivity. Like operant responses to cold stimuli, successful task completion is not strongly hindered by rough stimulation in normal rats. Unlike the strong preferences exhibited for thermal stimulus pairs, mechanical preference testing revealed only a modest preference for soft stimulation, which depended on the starting stimulus. Single Stimulus Response We have previously shown that increasingl y intense hot stimuli produce a steep decline in successful task completion when stimuli are presented individua lly (Neubert et al., 2005). In contrast, increasingly intense cold stimuli only produce a slight declin e in successful task completion, where significance was only observed at -4C (Rossi et al., 2006). The thermal 32


stimulus response shown here reflects these findings and provides a summary of normal responses taken from multiple cohorts of male hairless Sprague Dawley rats spanning a three year period. These findings are similar to another operant assay that evaluates pain based on how long rats spend on an illuminated (i.e. normally aver sive) thermoneutral platform in order to escape thermal stimulation of the paws (Mauderli et al ., 2000; Vierck et al., 2004) These authors also observed that escape duration increases steeply with increasing h eat, but the slope of responses with increasing cold is nearly fl at (Mauderli et al., 2000; Vierck et al., 2004). These findings are in contrast with a reflexive w ithdrawal study, which reported steep decline in withdrawal latency from 5 to -5C stimulation delivered by a peltie r-cooled floor when a 100s cut off was applied to define nociception (Allchorne et al., 2005). In contrast to th ese findings a more recent study, which applied a safety cut off of 180s, found that nave rats responded with latencies >180s for all cold stimuli tested (20-0C) (Tanimoto-Mori et al., 2008). The result s of the latter study are in agreement with operant findings described here and by Vierck, Mauderli, and colleagues (Mauderli et al., 2000; Vierck et al., 2004). Discrepancies between the two withdrawal studies could be rela ted to the different cut off times. Additionally, they could be related to differences in testing order, which was not specified by either author, although they did indicate that different stimuli were tested on different days. We have previously demonstrated that the testing of hot stimuli in the operant facial assay is not order dependent (Neubert et al., 2005), however initial experiment s indicated that success with painful cold (2C) was artificially low if this was the first stimulus the rats were exposed to after training. Thus it is possible that responses to cold stimulation coul d be more susceptible to order effects than heat responses. 33


With respect to the translational relevance of cold pain assessment in experimental animals, it may be especially important to incl ude paradigms that assess responses to multiple cold exposures rather than a si ngle stimulating event. Operant paradigms follow the former stimulation pattern and reflex assays the latter. Descriptions of painful sensations produced by cold stimulation are more variable than those de scribing painful heat, which is likely related to differences in the populations of cold receptors being activated. Human subjects report that repeated cold stimulation of glabrous skin pr oduces a deep radiating ache that increases in proportion to stimulus intensity (Mauderli et al., 2003), whereas contact w ithin the noxious range can produce a cold burning sensation (Morin and Bushnell, 1998) likely the consequence of deep c fibers that respond to both noxious cold and heat (Simone and Kajander, 1996). The deep radiating ache could be a composite of activity of non-nociceptive cold receptors in superficial skin and deep nociceptors near vessels (Mauderli et al ., 2003). Behaviorally distinguishing different populations of cold nociceptors, part icularly the a delta population, as can be done with laser stimulation for heat rela ted responses (Tzabazis et al., 2005; Tran et al., 2008) could provide an important dimension to understanding the development of cold allodynia characteristic of neuropathic pain. Thermal Preference and Unstimulated Time Our findings with respect to single stimuli impl y that while cold stimuli may not be as immediately painful as heat, they may still be av ersive. To more direct ly address the influence of affect on thermal processing, thermal preference between disparate hot and cold stimuli was assessed. The thermal preference findings indicate that where intensity is equivalent, warm or hot stimuli are preferred. By pa iring cold stimuli with 45C, a hot stimulus of equivalent intensity, the aversive nature of these stimuli are revealed by the strong heat preference. This has also been shown in a hindpaw assay, where mice were able to move freely between a neutral 34


(31C) floor and a cooled floor (Walczak and Beaulieu, 2006). The rats preference for heat when both stimuli are equal in intensity has also been demonstrated in humans. Cold pressor pain is rated more unplea sant than contact heat, although both have the potential to be painful (Rainville et al., 1992). In a recent study a 41C s timulus was rated more pleasant than a 12C stimulus, although neither stimulus was indicated as painful and rated with similar intensity by the subjects (Rolls et al., 2008). In contrast to most two-choice preference a ssays that stimulate the paws, this facial preference assay is actually a th ree choice task that provides the option of forgoing stimulation entirely to avoid contact with pa inful or aversive stimuli. This measurement is analogous to behavior produced by a single-stimulus shuttle box assay, where increasingly noxious stimulation to the hindpaws results in increasi ng time spent on a thermoneutral escape platform (Vierck et al., 2004). In general, rats spent 40% or more of testing time unstimulated, but we observed that the percentage of unstimulated time could be modulated by the aversiveness of the stimuli presented. Thus, unstimulated time and thermal preference likely reflect the affective and cognitive components of pain-related decision making in experimental animals. Evaluation of unstimulated time in addition to thermal preference is important because a treatment could potentially have no effect on thermal preferen ce, but could increase or decrease unstimulated time via equivalent proor anti-nociceptive action on both stimuli respectively. Interaction Between Stimuli In a multi-stimulus assay, it is important to consider the potential for each stimulus to recruit cell populations that can enhance or bloc k transduction of the other stimulus, presumably enhancing or blocking pain. In th is assay we can assess how successf ul rats are at obtaining milk reward in contact with either s timulus and compare this with success for each stimulus when it is presented alone. We observed that success ratios could be modulated by pairing hot and cold 35


stimuli. When non-noxious cold and 45C were paired, an insignificant increase in success at 45C was observed, as well as robust decrease in success at cold. These changes in success are likely due primarily to the strong 45C preference exhibited for these stimulus pairs, although the slight elevation of success at 45 C when paired suggests that the cold stimuli might weakly block heat pain at this intensity. In contrast, the effect of pairing on succe ss at noxious heat (48C) depended on the intensity of the cold stimulus, but not necessarily thermal prefer ence. In the presence of low intensity cold (18 or 24C), successful task completion was not changed for 48C, but was increased for 18 and 24C, following a pattern predicted by thermal preference for the cold stimuli. However, in the presence of high inte nsity cold (10 and -4C ), success at 48C was reduced, despite the fact that 48 C is preferred when paired w ith -4C. The opposite was also true; success was reduced for 10 and -4C in the presence of 48C. These findings suggest that intense cold activates substrates that can facilitate heat pai n, and vice verse. This cold-induced facilitation of heat pain could be due to the act ivation of peripheral nociceptors that respond to both cold and heat, or the convergent, facillator y input of distinct cold and heat-responsive nociceptors onto sec ond order neurons in the dorsal horn of the trigeminal nucleus caudalis. The latter pos sibility is more likely, as TRPM8 and TRPV1, molecular mediators of cold and heat respec tively, are minimally co-expressed in nave trigeminal ganglia (Kobayashi et al., 2005). TRPA1 has also been proposed to respond to nociceptive cold stimulation, but this is debated. TRPA1 a nd TRPV1 are highly co-expressed and agonists of this channel have been shown to weakly enhance cold pain (Albin et al., 2008). Although TRPM8 and TRPV1 co-expre ssion is reported to be very low in nave trigeminal ganglia (Kobayashi et al., 2005), this does not nece ssarily mean that such a population could not 36


have a significant influence on th ermal processing in the context of alternating or mixed hot and cold stimulation. It is also possible that th e cold and hot nociceptors have convergent inputs within the trigeminal spinal nucleus. Lingual application of the TRPM8 agonist menthol has been shown to cross-sensitize subjects to th e irritation produced capsaicin, a TRPV1 agonist (Cliff and Green, 1996). Additionally, local field potentials in the trigeminal nucleus caudalis respond to mixed thermal stimuli and agonists of TRPM8, TRPA1, and TRPV1 (Zanotto et al., 2007; Zanotto et al., 2008). Assessing changes in success in our thermal preference task could provide a means of studying the interact ion between cold and hot nociceptors in vivo Conditioned Response in the Thermal Preference Assay Thus far, we have established that this thermal preference assay can assess intensity related properties of pain by measurement of stimulus duration and interactions between cold and hot nociceptors by examining success changes. This assay also can also assess affect related properties of pain by the thermal preference and time spent unstimulated. Another common paradigm used to assess pain affect is conditi oned place aversion (CPA), where an inflammatory pain state becomes associated with one comp artment in a two or three chambered box and avoided by the animal in future trials (Vaccarino et al., 1992; Sufka, 1994). We found that we could use the stimulus pairs to condition a pref erence where one should not exist, or abolish a preference that should exist, in a manner simila r to conditioned place aversion. However, while CPA generally necessitates separate sensory te sting using reflex withdrawal to evaluate intensity-related propertie s of pain (Colpaert et al., 2006; va n der Kam et al., 2008), the facial thermal preference assay has the po tential to assess both intensity and affect aspects of pain simultaneously. 37


Operant Assessment of Mechanical Sensitivity The first operant assessment of sensitivity to tactile stimuli is described here. While no operant assessment has been devised to assess mechanical sensitivity in the hindpaw, a recent study has assessed withdrawal latencies from a text ured cork floor, which is comparable to the task defined here (Tanimoto-Mo ri et al., 2008). Limitations of this assessment include the inability to directly measures c ontacts with the mechanical stimulus and the lack of variety in the mechanical stimuli we are able to provide currently. The approximation of stimulus contact by counting bouts could introduce expe rimenter bias and, as menti oned, it cannot assess unrewarded contacts unless some other method is included. Photobeams between the stimulus bars could be used to more directly measure stimulus contact s and duration. Additiona lly, the current set of stimuli likely only produces innocuous, mildly av ersive sensations, which is supported by our preference finding and the influence of starting stimulus on preferen ce. This assay could benefit from inclusion of additional stimuli, with a consistent gradation from innocuous to painful sensation. We are exploring th is possibility using varying gr its of sandpaper to broaden the assessment potential of this assay. Despite these issues, a clear difference in su ccessful task completion was observed for the three individual stimuli evaluated. Surprisi ngly, operant responses to smooth metal and rough velcro did not significantly differ, while clearly success was greatest for the soft stimulus. This indicates that the soft stimulus is more pleasant than either rough velcro or smooth metal, even when the metal is maintained at an innocuous temperature (see single thermal stimulus response). Mechanical preference testing also supports the idea that soft is more pleasant than rough, but the fact that preference could be influenced by starting stimulus also indicates that rough is only mildly aversive. 38


Conclusion In this chapter, normal operant responses to thermal stimuli reveal differences between cold and hot stimulation, which are further em phasized in the thermal preference task. The thermal preference assay can assess both inte nsity and affective aspects of facial pain simultaneously. Additionally we describe a means of assessing mechanical allodynia in an operant manner. These methods of assessment provi de a clinically relevant means for examining mechanisms of pain because they can assess both the direct response to the stimulus and the avoidance of the stimulus. This is relevant to the patient experience because an individual may feel the need to forgo rewarding experiences to avoid pain or pain ma y begin to outweigh the reward. Thus, this chapter lays the foundation for evaluating the ro le of TRP channels in pain processing. 39


Figure 2-1. Modification made to the operant testing apparatus to assess mechanical sensitivity. A) Mechanical stimulus cuffs were made from plastic tubing (Fisher, wall thickness 1/16 mm) 11.4cm long, cut open on one side, with a 6.4cm length of rough or soft velcro (see magnified view for texture) a ttached to the region accessible to the rat from the inside of the box. B) View from the inside of the box, with mechanical stimulus cuffs on the thermode. The animal must brush its face against this stimulus in order to access the lick tube just outs ide the opening. Addi tionally, the movements made by the animal to lick and adjust head position provide br ushing or scratching contacts while the animal is in contact w ith the soft or rough s timulus respectively. C) Side exterior view of the reward access. 40


Figure 2-2. Example of licking bout determination based on compre ssed outputs of raw data in Windaq or Labview. A) Licking bouts for Channel 5 are indicated by astricies, and the compression rate is circled (22 seconds/d ivision; where a divi sion is the width of one red box). There are twelve bouts of licki ng visible in this out put. B) The same output viewed in Labview, which indicates el even bouts rather than twelve. Please note, the original images were captured and copied into Corel draw, which has reduced the image quality. This has made it difficult to view two of the brief bouts that were visible originally by viewing the raw data directly. 41


Figure 2-3. Effect of single thermal stimuli on opera nt task completion. A) Licks, B) stimulus contacts, C) and licks/contact (success ratio) were determin ed when rats were tested with thermal stimuli ranging from very cold (-4C) to very hot (52C). Dark blue = cold, light blue = cool, tan = warm, ora nge = moderate heat, red = noxious heat. Table indicates the number of rats for each stimulus and outcome measure. indicates a significant difference from 37 a nd 42C, + indicates significant difference from all stimuli cooler than the temperatur e indicated, and ++ indicates significant difference from all other stimuli. Signifi cance is set at p <0.05, and are determined by one-way ANOVA and post-hoc Tukeys te st. All data are mean SEM. 42


Figure 2-4. Distribution of time spent on the cold, hot, or off the thermode and percentage of licks spent at either thermode when cold stimuli were paired with 45 or 48C. A) More time is spent on the 45C stimulus when paired with a 10, 18, or 24C cold stimulus. Significantly less time is spent unstimulated when the cold stimulus is 24C. B) The majority of licks also occu r in contact with the 45C stimulus. C) More time is spent on the non-noxious cold stimulus (10, 18, 24C) when paired with 48C. In contrast, more time is spent at 48C when the cold stimulus is noxious (4C). The lowest percentage of time of the thermodes occurs when the cold stimulus is 24C and the most when it is -4C. D) More licks occur at innocuous cold than at 48C, and the opposite is true when -4 a nd 48C are paired. indicates p<0.05 across different pairs, determined by ANOVA with Tu keys test post-hoc. All data are mean SEM. 43


Figure 2-5 Successful task completion when stimuli are paired or presented alone A) Success at 45C is not significantly increased by pa iring with non-noxious cold stimuli as compared to when it is presented alone (n = 29). B) Success at the cold stimuli (10, 18, 24C) is significantly decr eased when paired with 45C. (10C n = 41, 18C n = 12, 24C n = 14). C) Success at 48C is significantly reduced when paired with -4 and 10C, but not significantly different when paired with 18 or 24C as compared to when it is presented alone (n = 28). D) Su ccess is significantly decreased at -4 and 10C when paired with 48C, significantly increased at 18C, and not significantly increased at 24C. indici ates p<0.05 for success alone ve rsus paired, determined by ANOVA and Tukeys test post-hoc (except -4 C, which was determined by paired ttest). All data are mean SEM. 44


Figure 2-6. Preference can be c onditioned or abolished by stimulus exposure 24 hours prior. A) Two equivalent stimuli normally do not pr oduce a preference, as indicated by 37 and 37C shown on the top. However, when preceded by a preference inducing combination (42 and 10C), more time is spent on the left thermode (previously preferred) than the right (previously avoided). The time spent on the right 10C thermode is also significantly different fr om the right thermode set at 37C. B) Previous experience can also hinder the pres entation of a stimulus driven preference. In this case, a strong prefer ence for the right th ermode (and avoidance of the left thermode) was induced with a combination of 52 and 18C. The following day, rats were tested at a combination of 18 and 48 C and did not show a preference. The time spent on the left side 18C thermode which was previously avoided, was significantly lower than the time spent at 18C the previous day or two days later when the 18 and 48C stimulus combination was repeated. indicates p < 0.05, with repeated measures ANOVA and post-hoc LSD. 45


Figure 2-7. Effect of smooth, soft, or rough mechanical stimulation on op erant task completion. A) Licks, B) bouts of licking (an indirect m easure of stimulus contacts), C) and licks per bout (a success measure) were determined when rats were tested with a smooth metal stimulus (37C, n = 30), a soft stimulus (n = 20), and a rough stimulus (n = 48). There were no significant effects of stimul us type on licks or bouts, but licks/bout were significantly greater with soft stimulation than either smooth or rough. indicates p <0.05, as determined by one-way ANOVA and post-hoc Tukeys test. All data are mean SEM. 46


Figure 2-8 Effect of starting side on licks and duration spent with the soft and rough stimuli, or on total licks and unstimulated time when so ft is preferred. A) Licks with soft stimulation were not different, while licks with rough stimulation were greater when rough was the starting stimulus. Difference in total licks between the two starting conditions did not reach statistical significance. B) Time spent in contact with soft stimulation was also not effected by star ting stimulus, while time spent on the rough stimulus was greater when this was the starting stimulus and unstimulated time was significantly reduced. indica tes significant difference between soft or rough as the starting stimulus (determined by unpaired t-test). p <0.05 All data are mean SEM. 47


Table 2-1. Number of rats exhibiting a cold, hot, or no preference at the stimulus combinations tested (bold numbers indicate the group preference). Temperature Pair (C) Cold Prefere nce Hot Preference No Preference Total n -4 & 48 1 5 17 10 & 48 37 15 2 54 18 & 48 8 109 24 & 48 7 209 10 & 45 2 12 01 4 18 & 45 0 9 09 24 & 45 2 10 0 12 37 & 37 11 14 2 27 (For 37&37C, the count in the cold preference column reflects number of rats with a left thermode preference). Table 2-2. Influence of previous experience on frequency of star t side and preference with 10 and 10C stimuli 24 hours later. Preference Induced First Day Left 42C Right 10C n rats started on this side 4 6 n rats preferred this side 10 0 Second Day Left 10C Right 10C* n rats started on this side 8 2 n rats preferred this side 8 2 Total n 10 Please note that the two rats who started on the right exhibited a left-side preference and the two rats that preferred the right started on the left. Table 2-3. Influence of previous experience on frequency of start side and thermal preference 24 hours with 18 and 48C stimuli 24 hours later. Preference Abolished First Day Left 52C Right 18C n rats started on this side 5 4 n rats preferred this side 1 8 Second Day Left 18C Right 48C n rats started on this side 3 6 n rats preferred this side 4 5 Total n 9 48


Table 2-4. Effect of start side on the number of rats pr eferring soft, rough, or neither mechanical stimulus. Starting Stimulus Soft Preference Rough Preference No Preference Total Rats Soft 25 3 2 30 rough 14 16 10 40 Note: rough start includes both experimenter controlled and voluntary occurrences; no differences were noted in pref erence pattern between involunta ry or voluntary rough start. Object 2-1. Video clip of rats perf orming the single stimulus operant task. 49


CHAPTER 3 PHARMACOLOGICAL MANIPULATION OF OP ERANT BEHAVIOR USING TRANSIENT RECEPTOR POTENTIAL CHANNEL AGONISTS Findings within the last two decades have uncovered six transient receptor potential channels that are involved in the encoding of thermal stimu li. Among these, TRPV1, TRPM8, and TRPA1 play an important role in normal and pathological pain perception. Doubt regarding the role of TRPM8 in cold nociception has been recently laid to rest w ith the profound lack of cold-mediated responses exhibited by two inde pendently developed strains of TRPM8 knock-out mice (Colburn et al., 2007; Dhaka et al., 2007). However, conflicti ng results still fail to resolve what role if any TRPA1 may play in cold nociception. Also, both TRPM8 and TRPA1 exhibit overlapping expression patterns with TRPV1 (Kobay ashi et al., 2005; Xing et al., 2006). Thus molecular receptors for cold and heat may inter act with each other with respect to different stimuli by their expression within a single prim ary afferent neuron. Th ese expression patterns define subpopulations of nociand thermoceptors that may exhibit overlapping connectivity with second order neurons. Thus, co-expression of TR P channels and connectivity of TRP-expressing primary afferents likely contribute to the complex and nuanced experience of thermal, chemical, and painful stimuli. In this chapter, we examine the effects of various TRP channel agonists on operant responses to cold stimuli, thermal preference, and unlearned cold-mediated behaviors. We examine the capacity of the different TRP cha nnel agonists to modulate operant responses to cold stimuli. We used the TRPM8 agonist me nthol, the TRPM8/TRPA1 a gonist icilin, and the TRPV1 agonists capsaicin and re siniferatoxin (RTX). RTX is an ultrapotent TRPV1 agonist, which binds to the channel and causes a massive influx of cations, includ ing calcium. This results in apoptotic cell death wh en RTX is applied peripherally or excitotoxic degradation of the primary afferent terminal when applied centrally (Clapham, 1996; Wexel, 2008). Thus, this 50


agonist can be used to specifica lly ablate afferents that expre ss TRPV1, coincidentally removing channels that are co-expressed w ith TRPV1. We demonstrate in th is chapter that cold pain can be manipulated both by the cooli ng agonists menthol and icilin, as well as the burning agonist capsaicin. We also demonstrate that selective ablation of TRPV 1-expressing profiles using the ultrapotent agonist RTX can sign ificantly impair pain specific responses, while leaving cold avoidance intact. Methods Animals Most of the animals were housed as descri bed in chapter 2, including six female hairless rats that were used to examine the effects of menthol on cold sens itivity (Rossi et al., 2006). The effect of capsaicin on operant responses to 10C and the effects of peripheral RTX on operant responses to cold were examined with haired male Sprague Dawley rats (Jackson Laboratories, 250-300g) conventionally housed in pairs. Sprague Dawley rats with normal fur were shaved and depilitated under anesthesia (isoflurane, 2.5% with oxygen) two to three time per week. Preparation and Administration of TRP Channel Agonists Menthol (10%) was dissolved in a vehicl e of 1.6% ethanol in Tween80 and phosphate buffered saline (PBS) and administered subcutan eously (s.c., 150 l) in each cheek. Icilin was dissolved in dimethyl-sulfoxide (D MSO) to a volume of 100l for in traperitoneal (i.p.) injections or 10l for intracisternal (i.c.m.) injections. A commercially available capsaicin cream was used for topical application of capsaicin (0.035%, Capzasin P; Chattem, INC; Chattanooga, TN). Capsaicin was applied bilaterally to the cheeks, left on for 5 minutes, removed with water, and the skin was dried with a cotton ball. Resi niferitoxin (RTX, 250 ng, LC Laboratories, Woburn, MA) was dissolved in a vehicle of 0.25% Tween 80 in PBS, 0.05% ascorbic acid to a volume of 50 l for infraorbital nerve applications, or a volume of 10 l for intr acisternal injections. 51


Topical capsaicin application, s. c. injections and i.p. injections were performed on awake, restrained animals. Operant behavioral testing was conducted immediately following capsaicin removal, 15 minutes following menthol or vehicle administ ration, 45 minutes following icilin or vehicle administration, and twenty four hours to two we eks following RTX or vehicle administration. The effects of topical capsaicin, menthol, and icilin are not permanent; therefore these drugs could be administered more than once in individual animals. However, because capsaicin can desensitize heat responses with re peated application, a period of at least one week was allowed to pass before administering capsaicin a second time. A cross-over design was used to administer menthol and vehicle. To avoid sensitization or desensitization from these compounds, as well as sensitivity from multiple injections, a recovery period of two days minimum passed between treatments (one week between i.c. m. injections). The effects of RTX are long lasting; therefore comparisons between RTX and vehicle treatment are inter-individual. Infraorbital and Intracisternal Injections All injections were performed under anesth esia (isoflurane, 2.5-5% with oxygen by inhalation). Perineural administ ration of RTX to the infraorbit al nerve was performed using a 25-gauge, 1.5 inch needle. The intersection of the zygomatic arch and the dorsal skull was located by palpation just medial to the eye on the top of the head. The needle was inserted at this point and carefully pressed along the bone of the skull until resistance was felt from the bone at the bottom of the arch, where th e nerve passes through the infraorbital foramen. The fluid was injected and the needle was carefully removed. Intracisternal injection of RTX or icilin was performed by dorsally flexing the rats head at a 45 degree angle, inserting a need le near the base of the occip ital bone, and stepping the needle 52


down until the atlanto-occipital membrane was breeched. A small amount of cerebrospinal fluid (CSF) was aspirated to confirm needle placement before drug administration. Additional Behavioral Tests Evaluation of wet dog shaking induced by icilin Dose selection and timing of thermal test ing was based on quan tification of wet dog shakes (WDS). Sprague Dawley rats (n = 4-5 ea ch dose, indicated in Ta ble 3-1) were injected with icilin i.p. or i.c.m. and observed in holdi ng cages for up to two hours following injection. WDS were counted in thirty minute intervals and the rate of WDS frequency was calculated for each rat by dividing the counts by the appropriate observation time, and averaged across rats. The following doses were used: 0.0025, 0.025, 0.25, a nd 2.5 mg of icilin (in DMSO). DMSO treated rats did not receive icili n. The other rats received icilin treatments, with at least 2 washout days between doses. Only the doses used fo r thermal preference testing are shown in Table 3-1. Capsaicin eye-wipe test Rats were taken into the observation room and allowed to acclimate for five minutes. Rats were removed from their c ohort one at a time and placed in an observation cage, partially covered by a clear sheet of Plexig las. Rats were gently restra ined by the experimenter, 20l of capsaicin solution (0.01%, dissolved in ethanol and diluted with PBS), was carefully delivered to one eye with a micropipetor, and the rat was im mediately released. Wipes were counted by the experimenter; an eye wipe was considered e ither a pass of the forepaw caudo-rostrally across the treated eye, or scratching of the rostral corner of the eye by the hindpaw. Responses typically lasted from 30 seconds to one minute 30 seconds. The second eye was not treated until signs of discomfort in the first eye were gone. Specifically, after the ra t no longer squinted or blinked the treated eye, and began engaging in ex ploratory investigation of the observation cage. 53


Statistical Analysis Effects of menthol and vehicle on operant responses were evaluated using a one way ANOVA. The effect of icilin dose and delivery on the rate of wet dog shaking, as well as the effect of icilin dose on thermal preference, wa s determined with a Kruskall Wallace test and post-hoc Whitney Mann tests. The effects of capsaicin and RTX on operant response to cold stimuli were determined using unpaired t-tests. Thermal preference in either RTX or vehicle treated rats was determined using repeated m easures ANOVA or paired t-test. Comparisons between RTX and vehicle treated rats for each s timulus in the preference task were made with unpaired t-tests. Significan ce was p <0.05 for all analyses. Results Effect of the TRPM8 Agonist Menthol on Operant Responses to Cold Male and female rats (n = 6 each sex) were tested at 24, 10, and -4C fifteen minutes following injection of either menthol or vehicle. Each individual rats outcome measures were compared to the baseline average for its sex to produce a percent increase or decrease from baseline for the menthol and vehicle treatment gr oups. There were no sign ificant effects of sex, so the data from all rats was pooled. At 10 C, there was a significant increase in stimulus contacts (F2,35 = 8.582) and a significant decrease in success ratios (F2,35 =5.476) for menthol treatment relative to vehicle (Figure 3-1). Ther e was no significant change in licks for either treatment (data not shown). These changes ar e indicative of allodynia following menthol treatment. At 24 and -4C there were no significant effects following menthol treatment (Figure 3-1). This data was previously reported in Molecular Pain (Rossi et al., 2006). Effect of the TRPM8/TRPA1 Agonis t Icilin on Wet Dog Shaking Wet Dog Shaking increased in a dose dependent manner (Table 3-2). The bodily location and intensity of WDS differed depending on deliv ery method. When administered i.p., the 54


classic head-to-tail WDS was produced, sometimes so intense at the high dose (0.25mg) that the rat was briefly unbalanced. When administered i. c.m., the shakes tended to be more focused in the upper body, beginning in the head but usually ending in the upper abdominal region. There were doseand administration-related differen ces in the rate of WDS across the two hour observation period (Table 3-2). The high dose (0 .25mg) of icilin produced a regular rate of WDS that persisted throughout th e two hour observation period. The rate was significantly higher for i.p. administration (two to three WDS/minute) than for i.c.m. (one to two WDS/minute), particularly in the middle hour of the observation period (Table 3-2). In contrast, the low dose icilin transiently elevated WDS in the first 30 minutes for i.c.m. administration (roughly one WDS every two minut es) and the second 30 minutes for i.p. administration (roughly one WDS every three minutes). In the remaining time, while WDS occurred with low dose icilin, they did not occur with signi ficantly greater frequency than DMSO. DMSO did not illicit regular WDS; in the entire two hour tes ting period rats exhibited an average of three WDS for both administration routes. Of the doses tested, 0.025mg was the highest dose that elevated WDS, but did not produce a persistent and regular rate of WDS. WDS remained elevated between 30 and 60 minut es for both high and low doses, so thermal preference testing was performed within this interval. Effect of the Icilin on Thermal Preference At baseline, rats strongly prefer 10C over 48C, licking more than 90% on that side and spending about 40% on the total te st time on the 10C thermode (F igure 3-2). Following i.c.m. administration of DMSO, although the baseline preference for 10C was not altered there was a significant increase in the licking and time spent on the 48C thermode. This indicates that DMSO itself may block some of the heat pain experience with 48C stim ulation, decreasing the rats natural avoidance of this temperature. Following i.c.m. administration of a low dose of 55


icilin (0.025mg), the strong preference for 10 C exhibited at baseline and following DMSO administration was abolished. Roughly 65% of to tal licks occurred with 48C stimulation, in contrast to about 5% at baseline or 20% with DMSO. Rats al so spent about 20% of the total test time at 48C. Intraperitoneal administration of DMSO and low dose icilin had no significant effects on thermal preference with facial stimulation (data not shown). Taken together, these data indicate that a low dose of icilin strongly decreased prefer ence for 10C in the face when administered i.c.m. Following high dose icilin (0.25mg), a preference for 10C is maintained (about 60% of licks), but was not significantly different from DM SO or baseline. These data suggest that the two doses of icil in have different effects that ma y interact with the heat-blocking effect of DMSO. Effect of the TRPV1 Agonist Capsai cin on Operant Responses to Cold To examine the effect of TRPV1 activation on operant behaviors in the presence of cool or cold stimulation, a capsaicin cream (0.035%) wa s applied bilaterally to the face, left on for five minutes, and wiped away with water. Rats were tested immediately following removal of capsaicin at either 10 or -4C (Figure 3-3). When tested with 10C stimulation, licks where not significantly changed, while stimulus contacts we re significantly decreased and success ratios were significantly increased with capsaicin than without. In contrast, wh en tested with -4C stimulation, licks and success rati os were significantly decreased, while stimulus contacts were not significantly different. Thes e data indicate that the moderate cold can alleviate the pain associated with capsaicin activation of TRPV1, wh ile noxious cold enhances the discomfort of capsaicin. Effect of TRPV1 Lesion with Resinif eratoxin on Operant Responses to Cold It has been previously established that RTX e liminates sensitivity to noxious heat and can alleviate inflammatory and neuropathic pain. Howe ver, it had not been established what effect 56


the removal of TRPV1 could have on cold sensitiv ity. We evaluated the effect of peripheral or central application of RTX on operant responses to cold stimuli, th ermal preference, and unlearned behaviors induced by menthol and icilin. Peripheral versus Central RTX and Cold Sensitivity Initial experiments with RTX targeted th e peripheral end of the trigeminal nerve (infraorbital branch). This treatment method significantl y reduces the capsaicin eye wipe response by a third and leads to a substantia l loss of TRPV1 imm unoreactive cells in the trigeminal ganglia as compared to vehicle treatme nt (Figure 3-4). There were also short term changes with respect to cold stimuli (2 and 10 C). One day following perineural RTX injection, stimulus contacts were dramatically reduced and su ccess ratios were increased in the presence of 10C stimulation relative to vehicle treated rats (Figure 3-5 B, C). However, by thirteen and fifteen days post-injection, res ponses to 10 and 2C respectively returned to normal in the RTXtreated rats (Figure 35). Th is could indicate that RTX treat ment may lead to an immediate inflammatory response underlying increased cold seeking behavior (see responses to capsaicin with 10C stimulation above) or that some functional recovery occu rs following peripheral RTX. Either of these possibilities could involve the small percentage of TRPV1 positive neurons spared by peripheral injection. Therefore, we sought to use a method that would more thoroughly and permanently block TRPV1 f unction within the trigeminal system. Effects of TRPV1 lesion by RTX on Cold Pain and Avoidance The central application of RTX via i.c.m. injection targets the brainstem and cervical spinal cord. Following successful injection of RTX, capsaicin eye wipe response was entirely negative and inflammatory hyperalgesia is bl ocked (Wexel, 2008). Three female hairless Sprague Dawley rats treated centrally with RTX remained capsaicin eye wipe negative for up to a year, suggesting that no functional recovery of TRPV1 activity occurs. Unlike peripheral 57


application, TRPV1-positive cells re main intact in the ganglia, but TRPV1-positive fibers are absent in the spinal trigeminal nucleus (Wexel 2008). Following injection, rats were allowed to recover for two weeks and tested in the single stimulus operant task with -4C. While there were no significant effects on licks RTX treatment significantly d ecreased stimulus contacts and increased success ratios, indi cating insensitivity to co ld pain (Figure 3-6). A group of RTXand vehicle-treated rats was also assessed for thermal preference (-4 or 48C) post-injection. Recall from Chapter 2 that untreated rats pr efer 48C stimulation with this combination of stimuli. While RTX-treated rats are insensitive to both -4 and 48C when these stimuli are presented individually, licking and te sting duration indicated RTX-treated rats still prefer 48C, although this preferen ce does not reach statistical signi ficance as it does for vehicletreated rats (Figure 3-7 A, B). RTX-treated rats also had signif icantly fewer stimulus contacts and success ratios than vehicle-treated rats fo r the 48C stimulus and for responses at both stimuli combined (Figure 3-7 C, D). This suppor ts previous findings that RTX treatment renders animals insensitive to noxious heat. Thus, while cold sensitivity is highly impaired follow RTX treatment, some cold perception mu st still exist that contributes to a cold aversion. Recall from Chapter 2 that where both stimuli are non-painful, heat is preferred. Therefore, RTX lesion of TRPV1-expressing fibers likely rend ers the stimulus pair non-painful, but leaves the capacity for discrimination between warm and cold intact. Effects of TRPV1 lesion by RTX on Thermoge nic and Nocifensive Responses induced by TRPM8 Agonists We have also assessed the effects of RTX treatment on innate behaviors induced or enhanced by agonists of TRPM8. We have obser ved that RTX-treated rats remain sensitive to icilin (0.25mg i.c.m.). Icilin is a TRPM8/TR PA1 agonist that induces wet dog shakes, which are thought to be the rodent analog of shivering a nd dependent on TRPM8 activity in the periphery 58


(Colburn et al., 2007; Dhaka et al ., 2007). We have also observed that RTX application to the sciatic nerve can eliminate the ability of menthol to enhance nocifensive responses to acetone cooling of the hindpaw (general observations). Taken together, these data provide behavior al evidence to support the idea that cold nociception represents a distin ct subset of cells which express TRPV1. This nociceptive population is likely responsible for transmitting th ermal pain regardless of modality (hot vs. cold), as well as agonist-rela ted irritation, demonstrated by no cifensive behaviors. However, there are also cold receptive neurons not expres sing TRPV1 that encode the cold quality of thermal stimulation. This population likely also play s an important role in the regulation of basal body temperature and thermogenesis, as well as a ffective judgments regarding thermal stimuli. Discussion of TRP Channel Mani pulation and Sensory Processing Effects of Menthol on Cold Sensitivity It was hypothesized that 10% me nthol would have effects on a ll three cold stimuli because they are within the activation range for TRPM 8, yet effects were only observed with 10C stimulation. The lack of menthol induced allody nia with 24C stimulatio n likely has to do with sub-populations that may not be sufficiently ac tivated by this level of stimulation. TRPM8 mRNA has been identified in both small and medium diameter rat DRG neurons in vivo which are presumed to correspond to the cell bodies of Cand A fibers, respectively (Kobayashi et al., 2005). This has also been demonstrated in mice expressing green fluorencence protein under the control of the TRPM8 promoter (Dhaka et al., 2008). In the periphery, 24C may not be of sufficient magnitude to activate TRPM8-containing C-fibers but could activate cold responsive A fibers that block C-fiber activity and preven t the induction of menthol -induced allodynia (Liu et al., 1998). Within the spinal dorsal horn, cool-sensitive lami na I spinothalamic (STT) cells 59


have a specific sensitivity to temperatures between 34 and 15C. The sensitivity of polymodal nociceptive HPC (for heat, pinch, cold) cells to noxious cold be gins at about 24C, and their response to cold accelerates at te mperatures below 15C. It has been suggested that an increase in HPC activity beyond that of cool-sensitive cell s signals the sensation of burning pain (Craig, 2003). Even in the presence of menthol it s eems 24C does not provide input of sufficient magnitude to increase HPC STT cell activity beyond cool-sensitive STT cells in lamina I and thus could explain why menthol did not enhance sensitivity to 24C. The lack of menthol-induced hyperalgesia with -4C stimulation is likely due to the concentration used. Menthol is known to produce a burning sensation in humans at a concentration of 40% and also increases sensitivity to cold, presumably by its action on C-fibers (Wasner et al., 2004). It is po ssible that 10% menthol, while su fficient to induce allodynia at 10C, was not sufficiently concentrated to induce hyperalgesia at -4C. It is important to note that it also did not induce insensitivity to -4 C, reducing the possibility that TRPM8-expressing cells were desensitized to further cold stimulation. Effect of Icilin on WDS We first characterized the WDS produced by different doses of icilin when administered either i.p. or i.c.m. A difference in temporal profile for i.p. versus i.c.m. administration was observed. Following i.c.m. administration of icili n, WDS elevation was evident within the first 30-minutes and was either maintained (high dose) or declined (low dose) over the following 90 minutes. Following i.p. administration, WDS eleva tion was evident in the first 30-minutes, but increased over the second 30-minut e interval, rather than remain ing steady or declining. This difference is likely related to the different local environment and di stance of the icilin agonist. Intraperitoneal inje ction delivers the agonist acts to the cutaneous ends of primary afferent neurons, which are diffusely distributed in hete rogeneous tissue. Once the agonist is in the 60


vicinity of a primary afferent, it must activate a sufficient number of channels to depolarize the cell. In contrast, i.c.m. injection delivers the ic ilin directly to the syna pses between the primary afferent neurons and the second order neurons, increasing the chance that enough channels will interact with the agoni st to alter the neurons membrane potential. Effect of Icilin on Thermal Preference Based on the quantification of icilin-induced WDS, we chose two doses of icilin to manipulate thermal preference: a low dose that elevated WDS somewhat, but did not produce a persistent rate of shaking, and a high dose that produced a persistent rate of one to three WDS/minute. Thermal preference for 10C was abolished when rats were administered low dose (0.025mg) icilin. This effect of low dose icilin is likely the consequence of increased sensitivity to cold due to icilins action on TRPM8 channels and perhaps TRPA1 channels as well. Icilin potentiates the act ivation of TRPM8 by cold stimuli (Chuang et al., 2004) and also activates TRPA1, but with less pot ency (Doerner et al ., 2007).. The capacity for TRPA1 to be directly activated by cold stimul ation is still debated (Story et al., 2003; Jordt et al., 2004; Bautista et al., 2006; Kwan et al., 2006), althoug h it clearly plays some role in nociceptive processing. There is evidence that TRPA1 may be involved in pro cessing of mechanical stimuli (Kwan et al., 2006; Cahusac and Noyce, 2007; Ki ndt et al., 2007), and may therefore mediate pricking sensations induced by icilin. Even if it is not directly ac tivated by cold stimulation, TR PA1 may still be indirectly involved in cold nociception. It has recently been shown that calcium alone can activate the TRPA1 channel (Doerner et al., 2007), raising the possi bility that this channel may act as a coincidence detector and modulator of nociception. There is evidence to suggest that there are other, unidentified cold receptors (Babes et al ., 2002; Munns et al., 2007). If these are coexpressed with TRPA1, it might be possible that calcium influx via these receptors could further 61


activate TRPA1. Icilin could act on TRPA1 in th ese cells to increase their response in the presence of cold stimuli. Alternatively, becau se of its high degree of coexpression with TRPV1, activation of TRPA1 combined with activity in coldresponsive ne urons could be responsible for the burning sensation sometimes reported by prolonged contact with cold stimuli. Additionally, TRPA1 has also been shown to f acilitate excitatory synaptic transmission in the substantia gelatinosa (Kosugi et al., 2007). Thus, appl ication of icilin could potentially enhance transmission of cold input from peripheral afferents by its actions on TRPA1 at the first synapse. In contrast, the high dose of icilin did not significantly alter thermal preference relative to DMSO. This difference may be due to the escap e option available to rats with the facial preference task, as well as poten tial interaction of DMSO with this dose of icilin. DMSO is commonly used as a solvent for application of agonists in electrophys iological and calcium imaging studies. It has also been used in vivo to sensitize A delta fibers (Wilson et al., 1999; Tzabazis et al., 2005) and has also been shown to block c fiber conduction (Evans et al., 1993). The mechanism for DMSO induced a delta sensitiz ation has never been cl arified, but the c fiber block may be a consequence of sensitized a delta fibers inducing long term depression in c fibers (Liu et al., 1998). In this study, DMSO also increased licking a nd time spent with 48C, but did not have a significant effect on 10C repsonses indicating a possible analgesic effect for heat pain. The inhibition of secondary pain conducted by c fibers may allow rats to recover from 48C stimulation more readily. This would enable th em to be somewhat more successful on the hot thermode. When DMSO is combined with a low dos e of icilin that produces cold sensitivity, the combined effects of these compounds is complementary, and rats are even more successful at 48C. Evidence suggests that TRPM8 is expresse d in both A delta and c fibers, whereas TRPA1 62


is primarily expressed in c fibers. At low dose icilin, the activity of A delta fibers containing TRPM8 is likely enhanced by icilins actions on TRPM8. When DMSO is combined with high dose of icilin, icilins capacity to sensitize TRPA1-expressing c fibers may be blocked by DMSO. Conversely, any sensitiz ing effect that DMSO may ha ve on TRPM8-expressing A delta fibers could potentially be inhib ited by desensitization of TRPM8 that may occur in the presence of high dose icilin. Further experimentati on is needed to resolve these issues. A Role for TRPV1 in Cold Pain The hyperalgesic responses to noxious cold induced by TRPV1 activity, and insensitivity to noxious cold following ablation of TRPV1 sup port the idea that TRPV 1-expressing afferents can contribute to the perception of painful cold. The alleviation of capsaicin irritation by 10C is supported by previous electrophysiological findings (Babes et al., 2002) and by menthol in oral irritation studies (Green and McAuliffe, 2000). The ability of RTX to eliminate menthol irritation further supports previous findings that hypothesized that menthol -induced irritation was dependent on capsaicin-sensitive nociceptors (Cliff and Green, 1996). There are two possible explanations for the effects of TRPV1 activation and lesion on co ld mediated behaviors. The first possibility is that TRPV1 is expressed in cold nociceptors, activated by cooling below 10C. The second is that TRPV1 expressing primary afferents provide convergent inputs with coldresponsive primary afferents onto the same second order neurons in the dorsal horn. These two possibilities are not mutually exclusive and likel y both contribute to the effects of TRPV1 on cold perception. With respect to the first explanation, TRPM 8 and TRPA1 are co-expressed to some degree with TRPV1, as previously mentioned. Under normal conditions, the perc entage of nociceptive cells expressing both TRPV1 and TRPM8 has been reported to be low (Xing et al., 2006; Dhaka et al., 2008), which has led some to suggestion that their contributi on to normal cold pain may be 63


minimal (Dhaka et al. 2008). However, sma ll population size may not be adequate reason to dismiss this population in normal cold nociceptio n. TRPV1 knock-out mice were found to have intact withdrawal responses to painful heat within the activation ra nge of TRPV1 (42-49C) (Caterina et al., 2000). Later studi es have attributed the mainte nance of these responses to a small population of heat nociceptors that remain intact and provide input to lamina I (Eckert et al., 2006). In a similar manner, removal of the small TRPM8/TRPV1 co-expressing population may be sufficient to block cold nociception. Further behavioral testi ng of with TRPM8 and TRPA1 knockout mice, and treatment with RTX, could help determine if the TRPV1/TRPM8 plays a substantial role in normal cold pain perception. Another candidate population of cold nociceptors is ce lls that express TRPV1 and TRPA1, which represents nearly 100% of the TRPA1 population in vivo (Kobayashi et al., 2005). The role of TRPA1 in cold nocicepti on however has been highly debated. Early in vitro work demonstrated that TRPA1 expressed in Ch inese hamster ovarian cells was responsive to cold <18C, and that cells in mouse DRGs exhi bited properties of TRPA1-expression (Story et al., 2003). Subsequently, this work has been both supported (Bandell et al., 2004; Macpherson et al., 2006; Sawada et al., 2007) and contested (Jordt et al., 2004; Nagata et al., 2005; Zurborg et al., 2007). Behavioral evalua tion of TRPA1 knock-out mice developed in two different laboratories also yielded conflicting results. Kwan and colleagues demonstrated that female knockout mice exhibited a robust decr ease in cold irritati on (paw lifts with cold and duration of post-acetone responses), while males exhibited only slight, insignificant decr ease in cold irritation (Kwan et al., 2006). In contrast, Bautista and colleagues did not reveal any effect of TRPA1 knock-out on cold irritati on (acetone test), withdrawal th resholds, or shivering, although they did not make an effort to discriminate betw een sexes (Bautista et al., 2006), even within the 64


nociceptive range. However, even if TRPA1 is not directly activ ated by painful cold, it is known to be activated by intracellular calcium (Zurbor g et al., 2007) and may serve as a coincidence detector for another unidentified cold receptor (Babes et al., 2006; Munns et al., 2007) that increases intracellular calcium. The second explanation for the effects of TRPV1 on cold pain involves the convergent input of cold and hot nocicepto rs onto the same second order ne uron in the dorsal horn of the spinal trigeminal nucleus. Cold-responsive units in the superficial trigemin al nucleus caudalis all responded to noxious heat and the majority of th ese also responded to me nthol, with substantial overlap in menthol, cinnamaldehyde, and capsaicin responsivity (Zanotto et al., 2007), indicating convergent inputs of TRPM8, TRPA1, and TRPV 1 expression afferents onto second order neurons. Furthermore, menthol and cinnamaldehyde are able to cross de sensitize the responses of these units (Zanotto et al., 2008). These findings support the idea that cells expressing TRPM8 or TRPV1 only, and TRPV1/TRPA1 provide convergent input onto individual second order neurons within the trimeinal nucleus caudalis. Conclusions We demonstrate here that menthol and ic ilin can modulate cold-mediated behaviors, although these effects did not always follow expect ations. Additionally, we demonstrated a role for TRPV1-expressing afferents in cold pain in a non-pathological stat e, although it is still unclear which molecular mediator(s) for cold sensitivity are coupled with TRPV1. Cold nociceptors are likely heteroge neous, which certain subpopulations and different proportions of activity encoding different aspects of cold sensation. Furthe r manipulation of TRP channel activity in pathological states will allow us to fu rther clarify the role of TRP channels in chronic pain. 65


Figure 3-1 Effect of menthol (10% s.c.) on operant responses to co ld stimuli. A) Facial contacts are significantly incr eased by menthol treatment with 10C stimulation, but not -4 or 24C stimulation. B) Success ratios are si gnificantly reduced by menthol treatment with 10C stimulation, but not -4 or 24C stimulation.* indicates p <0.05 (one way ANOVA). Data are mean SEM. 66


Figure 3-2. Effect of two doses of icilin a nd DMSO on thermal preference with facial stimulation at 10 and 48C. The number of lic ks, A) and of the distribution of testing time, B) are shown for baseline (n = 12) and following DMSO (n = 6), 0.025mg (n = 8), or 0.25mg icilin (i.c.m., n = 6). Data ar e mean SEM. A) There was a significant effect of treatment on licking at 10C (K ruskalWallis test, chi squared = 16.55, p = 0.001) and 48C (chi squared = 20.93, p < 0.001). B) There was a significant effect of treatment on time at 10C (chi squared =16.59, p = 0.001) and 48C (chi squared =21.24, p < 0.001), but not on time off the thermodes (chi squared = 6.50, p = 0.09). Post-hoc comparisons were made using th e Mann-Whitney U test. indicates a significant effect of low dose icilin treatment as compared to all other treatments (baseline, DMSO, and high dose icilin). + indicates a significant difference between DMSO and baseline only. For all statistical tests p < 0.05. 67


Figure 3-3 Effect of the TRPV1 agonist capsaicin on operant responses to moderate (10C) and noxious cold (-4C). A) Capsaicin (topical, 0.035-0.075%) significantly reduced licks with -4C stimulation, but not 10C. B) Capsaicin significantly reduced stimulus contacts with 10C, but no -4C stimulati on. C) Capsaicin significantly reduced the success ratio (licks/contact) with -4C stim ulation and significantly increased it with 10C stimulation. indicates significan t difference between capsaicin and no treatment, p <0.05 (t-test). Data are mean SEM. 68


Figure 34 Effect of periphera l resiniferatoxin (RTX) treatment on TRPV1 expression on in the trigeminal ganglia and TRPV1 function. A&B) Sections of trigeminal ganglia from an RTX (A) and vehicle (B) treated rats, labeled with a rabbit anti-TRPV1 antibody and counterstained with hematoxylin ( 200X). White astricies indicate TRPV1 positive cells (brown). C) Quantification of positive cells remaining in RTX versus vehicle treated ganglia. D) RTX significan tly decreased wipes in response to intraocular capsaicin (0.1% in PBS), used to evaluate function of TRPV1. indicates p <0.05 (t-tests). Data are mean SEM. (Images courtesy of Alan Jenkins). 69


Figure 3-5 Effect of peripheral TRPV1 lesion by resiniferatoxin (RTX) on operant response to 10, 2C stimulation. A) RTX had no effect on licks. B) RTX significantly reduced stimulus contacts with 10C only one da y post-injection. C) RTX significantly increased success ratios (licks/contact) up to thirteen days postinjection. indicates significant difference between RTX (n = 3) a nd vehicle (n = 4), p <0.05 (t-test). Data are mean SEM. 70


Figure 3-6 Effect of central TRPV1 lesion by resi niferatoxin (RTX) on operant response to -4C stimulation. A) Licks were not significan tly different between RTXor vehicletreated rats. B) RTX-treatment significan tly reduced stimulus contacts. C) RTXtreatment significantly increased success ra tios (licks/stimulus c ontact). indicates significant difference, p <0.05 (t-test). Data are mean SEM. 71


Figure 3-7 Effect of central TRPV1 lesion by resi niferatoxin (RTX) on thermal preference for -4 and 48C stimulation. A) Licks occurred more with 48C stimulation for both treatment groups and total licks were sli ghtly greater with RTX. B) Duration of stimulation also indicates 48C preference. C) Stimulus contacts were significantly reduced with RTX treatment relative to vehi cle for 48C stimulati on. D) Success ratios (licks/stimulus contact) for the combined stimuli were significantly greater with RTX-treatment than with vehicle. indica tes significant difference between stimuli or time distribution (paired t-test, re peated measures ANOVA for duration), + indicates significant difference between tr eatments (unpaired t-tests), p <0.05. Data are mean SEM. 72


Table 3-2. Rate of Wet Dog Shaking (WDS per minute, mean SEM) calculated from observed counts over thirty-minute intervals fo llowing two doses of icilin or DMSO administered intraperitoneally (i.p.) or intracisternally (i.c.m.). Rate of WDS per Interval Route of Delivery Dose Rats (n) 0-30 30-60 60-90 90-120 Total Observation 0.25 mg 4 3.6* 3.4.3* 2.8.3* 1.5.5* 2.7.2* 0.025 mg 4 0.1.05 0.3.04* 0.1.03 0 0.1.01* i.p. DMSO 5 0.08.02 0.01.01 0 0 0.02.01 0.25 mg 4 2.2.3* 1.9.3* 1.2.3* 0.6.1* 1.5.2* 0.025 mg 5 0.6.1* 0.2.06 0.1.02 0.1.02 0.2.06* i.c.m. DMSO 4 0.03 0.03 0 0 0.04.004 Total 26 Chi squared values (Kruskal-Wallis test, df = 5, p 0.001) 22.03 23.27 20.43 20.37 22.99 Significant difference between ic ilin dose and DMSO vehicle va lue of same delivery type. indicates significant difference between delivery t ype at the same dose. Significance is defined by p<0.05 (Mann-Whitney U test). 73


CHAPTER 4 EFFECTS OF NEUROPATHIC PAIN ON OP ERANT RESPONSES TO THERMAL AND MECHANICAL STIMULATION Neuropathic Pain Despite the clinical prevalence of orofacial pain, basic research relating to trigeminal nociception is relatively limited as compared to other somatosensory systems in the body. The chronic constriction injury (CCI) is one model that has been used to mimic neuropathic pain states observed clinically, origin ally applied to the sciatic nerv e (Bennett and Xie, 1988). CCI was first adapted for the trigeminal nerve by Vo s and colleagues (Vos et al., 1994), and has been used by a handful of other groups to examine peri pheral and behavioral ch anges associated with this injury (Imamura et al., 1997; Benoliel et al., 2001; Deseure and Adriaensen, 2004; Chichorro et al., 2006), as well as the eff ect of novel analgesics (Benoist et al., 1999; Chichorro et al., 2006; Lim et al., 2007; Ling et al., 2008). However, behavioral assessm ents in CCI-treated animals primarily rely on reflexive withdrawal and a ssessments of grooming behavior. Only three studies have used an operant assessment of pa in following sciatic CCI (Vierck et al., 2005; Jabakhanji et al., 2006; Walczak and Beaulieu, 2006). In this chapter, the e ffect of bilateral CCI on operant behavioral measures is assessed. Once we established the pattern of behavior s observed following CCI, we determined the effect of treatment with gabapent in and pregabalin on those behaviors. Gabapentin was initially used as an antiepileptic, then la ter found to be effective as an an algesic and was approved for the treatment of neuropathic pain (lab eled Neurotin) (Wheeler, 2002). De spite the fact that it is an analog of the inhibitory mo lecule gamma-aminobutyric acid (GABA), it does not act on GABA receptors (Jensen et al., 2002). Pregabalin is a derivative of gabape ntin, also approved for treating neuropathic pain (labeled Lyrica) (Selak, 2001). Both gabapentin and pregabalin bind and block the activity of alpha 2 delta subunit of voltage gated calcium channels at the first 74


synapse in the dorsal horn (Gee et al., 1996; Fiel d et al., 2006). Following neuropathic injury, ectopic discharge of injured neurons contributes to the upregulati on of the alpha 2 delta 1 subunit (Boroujerdi et al., 2008), which c ontributes to pain maintenance by increasing the excitability of neural activity within the dorsal ho rn (Li et al., 2006). In the cu rrent study, we demonstrate that both gabapentin and pregabalin improve success in the operant task wi th cold stimulation. Methods Induction of Neuropathic Pain, Monitori ng Recovery, and Behavioral Testing Rats were trained and tested with single th ermal and mechanical stimuli as described in Chapter 2. Following training and ba seline behavioral testing, rats received eith er a bilateral chronic constriction injury (CCI) or sham operati on of the infraorbital portion of the maxillary trigeminal nerve. These surgeries were perfor med intraorally, as previo usly described (Imamura et al., 1997), with some modification. We chose this route ra ther than the more common external surgical site so that the incisi on site would not be within the s timulated portion of the skin. Briefly, rats were deeply anesthetized with a ketamine/xylazine cocktail (2:1, 8mg/kg ketamine, 4mg/kg xylazine, 1.2ml/kg cocktail). For lo cal anesthesia, xylocaine (2%, 1:100,000 epinephrine) was applied to reduce operative se nsation and for hemostasis. An incision was made in the buccal vestibule, beginning from th e hard palate and extending approximately 1 cm rostrally, roughly parallel with the lip. The nerve was exposed from surrounding tissues by blunt dissection and gently elevated with a hooked instrume nt so that two ligatures (5-0 vicryl sutures) could be tied securely around the nerve. This pr ocedure was the same for sham surgeries, except that sutures were passed twice unde r the nerve, but not tied. This process was repeated for the second side. All incisions were closed with 2-4 ligatures (5-0 vicryl suture). Rats weights, general behavior, facial sw elling and scratching were monitored for one week after the surgeries to track the progress of recovery and note signs of excessive grooming 75


to the innervated area. A subjective swelling seve rity score (none, mild, moderate, severe; Table 4-2) was used to grade the amount of swelling in the orofacial region affected by surgical treatment. The location and extent of scratchi ng within the innervated area was also noted. Behavioral testing began after the one week r ecovery period. Four sets of surgeries were conducted, for a total of 22 rats assigned to each treatment submitted for single stimulus testing. Following surgical treatment and r ecovery, rats were tested with one of the following weekly schedules: (1) three times with 10C and once wi th 37C; (2) once with rough mechanical, twice with 10C, once with 37C, and one with 48C; or (3) once each with rough mechanical, 10, 37, and 48C per week. One surgical group (n = 5 per treatment) was submitted for stimulus preference testing before and after surgical treat ment. Fasting occurred three times per week on the nights prior to 10 and 48C stimulation, but not prior to 37 C, or mechanical simulation. Evaluation of Innate and Aversive Behaviors during Operant Testing We established a behavioral scoring system to assess a dditional innate and aversive behaviors occurring during the first five minutes of operant testing. Five innate behaviors (facial grooming with forepaws, facial grooming with hindpaws, forepa w shaking, head shaking, and wet dog shaking) and three aversive or aggressive behaviors (head tilting, wiping or pushing at the stimulus, and biting at the stimulus or sippe r tube) were each given one point for a maximum possible combined score of 8 for the five minut e period. We chose to assess the first five minutes of testing rather than the whole testing period because untreated rats attend to task completion immediately and typicall y do not engage in grooming or innate behaviors until they completed at least one bout of drinking. This and all behavioral tes ting described above was assessed by a treatment-blinded investigator. 76


Evaluating the Effect of Surgical Treatm ent, Novelty, and Pregabalin on Thermal Preference One group of rats (experienced) was trained and submitted for thermal preference testing both before and after surg ical treatment (see above), renderi ng them experienced with this task before and after pain induc tion. Another group of rats (in experienced) was not exposed to the thermal preference task until after surgical treatment and four weeks of single stimulus testing, primarily with 10 and 37C stimulation. By this time, CCI-treated rats in this group no longer exhibited allodynia, nor di d they suffer from heat hypera lgesia. At five weeks postsurgery, this group was first exposed to the ne utral (37/37C) place pref erence task for two sessions, then evaluated for thermal preference with 10 and 48C stimul ation twice with two days separating each exposure (hot and cold sides were switched). Preferences with each surgical group were consistent and therefore the two days were poole d. At six weeks postsurgery, they were evaluated for thermal prefer ence again following treatment with pregabalin (45 minutes prior, 10mg/kg i.p.). Both the expe rienced and inexperienced groups consisted of CCI-, sham-, and un-treated rats. Histology Tissue preparation Rats were euthanized by isoflurane asphyxiation and decapitation at several postoperative time points (2, 5, 6, 7, and 12 weeks). The infraorbital trigeminal nerves and trigeminal ganglia were dissected free and pos t-fixed in 10% neutral buffered formalin for 48 hours before being transferred to 70% ethanol. Tissues were embedded in paraffin blocks and 10 m sections were cut with a microtome and m ounted on slides. At least two non-adjacent sections, separated by 50 m, were taken from each pair of nerves and ganglia. Tissues were deparafinized and rehydrated in a grad ed series of alcohol washes into PBS. Nerves were stained 77


with Gills hematoxylin and eosin Y (Fischer, Kalamazoo, MI) to visualize tissue for evaluation of inflammation as described below. Immunohi stochemistry was performed on sections of trigeminal ganglia sections to visualiza tion TRPV1 and TRPM8 proteins and quantify immunoreactivity. Immunohistochemistry Target retrieval was used for immunostaining of TRPV1 onl y; it interferes with TRPM8 immunostaining. Sections were incubated in 100 ml target retrieval so lution (Dako Cytomation North America, Carpinteria, CA) for 15min in boi ling water, then incubated in an oven overnight at 60C overnight. On the second day, sections were cooled and washed with 0.1% Tween20 in PBS. They were then blocked with 10% nor mal goat serum (NGS) for 30 minutes at room temperature, and incubated w ith rabbit anti-VR1 an tibody (1: 500, Affinity Bioreagents; Golden, CO) in 5% NGS in 0.1% Tween20 in PBS (the anti body diluent) overnight at 4C. This process was the same for immunostaining of TRPM8 with the following changes: target retrieval was not performed and the chicken anti-TRPM8 antibod y (1:500, Aves, Tigard, OR) was incubated for two days at 4C. After incubation in the primary antibody, sect ions were washed and incubated with the appropriate secondary antibody in the antibody diluent at 1:1,000 (for TRPV1; goat anti-rabbit IgG AlexaFluor 488 and for TRPM8; goat anti-chick en IgY Alexa Fluor; Invitrogen). Slides were then washed in PBS and coverslipped with Vectashield Mounting Medium (Vector Laboratories, Burliugame, CA). Immunostained sections of trigeminal ganglia were viewed with a Micro-Leica DMLBZ microscope at 200x magnif ication, and images were captured using the Q Imaging Micropublisher 5.0 camera and Q Capture Pro software. The area of the trigeminal ganglion innervated by the infrao rbital nerve was surveyed an d cell counts were made using three representative images from each ganglion from each individual (n = 6 per rat). While no 78


difference was observed in the percentage of TR PM8 positive cells, a difference in the intensity of TRPM8 staining between treatment groups was noted qualitatively (contrast conditions were the same for all groups). Thus additional images (n = 6 per rat) consisting of TRPM8 positive cells were taken and the intensity of staining was analyzed using ImageJ 1.4g (National Institutes of Health). Assessment of nerve inflammation and injury Assessment of nerve inflammation and injury was made by a blinded investigator. Two longitudinal, non-adjacent sections of the left and right nerves from each individual were assessed in the approximate surgical area. A scoring system was used to assess inflammation and injury to the nerve (Figure 4-1), and the presence of suture and foreign body reaction (giant cells) was noted. Each of the four sections per rat received an inflammation score and an injury score. These were then added to produce a co mposite score for each rat (maximum of 24), used for statistical analysis. If sutu re and foreign body reaction was noted in any of the four sections, then the rat was considered to have suture. Statistical Analysis For analysis of operant data following surgical treatment, outliers were first removed from all data sets using a box plot analysis (SPSS, v.16), on an outcome by outcome basis. Outliers are defined as any value beyond one and a half times the interquartile range. General linear models were used to assess the effects of surgical treatment, time and their interaction on operant outcome measures and behavior scores fo r all stimuli tested, the effects and interaction of drug and surgical treatment on operant respon ses to 10C, as well as scores assessing inflammation, percentage of TRPV 1 expression, and TRPM8 intensitie s. One wayAnalysis of Variance (ANOVA) was used to assess the effect of time within each surgical group, followed by post-hoc Tukeys test. Unpaired t-tests were used to make comparisons between behavioral 79


outcome measures of CCIand sham-treated rats at each post-operative time point. One way ANOVA was also used to determin e significant effects of experi ence and pregabalin treatment on thermal preference following surgical treatmen t. Statistical significant was set at p<0.05 for all analyses. Results General Observation of Immediate Po st-Surgical Recovery and Behavior All rats weights were recorded before su rgery and monitored daily for one week after surgery to track recovery. CCI-treated rats lost an average of 3% of their body weight at 24hrs post-op, but recovered to their preoperative weights or greater by the seventh post-operative day. Most sham-treated rats lost less than 3% or maintained their pre-operative weight at 24hrs postsurgery. Those shams that did lose more than 3% of their pre-operative we ight gained it back by the fourth post-operative day. Although CCI-treat ed rats had a trend fo r lower body weights than their sham-treated counterparts, no significan t differences were observed between the two groups. All operated rats exhibited moderate to severe swelling in the surgical region for the first two post-operative days, which was more severe a nd longer lasting in CCI-t reated rats than in their sham-treated counterparts. It took six days for swelling to disappear in half of the CCIoperated rats (Table 4-1). In contrast, it took only three days for swelling to disappear in half of Sham-operated rats (Table 4-1). Swelling was co mpletely gone in all CCI-operated rats by seven to nine post-operative days and in all sham-operated rats by six post-operative days. All surgically treated rats were observed engaging in isolated facial grooming on recovering from anesthesia. In particular, some CCI-treated rats would sw at at the rostral portion of their faces with their fore paws, in a motion reminiscent of normal face washing, but without contact to the face and often inters persed with forepaw shakes. 80


Some CCI-treated rats were also observed jerking or jumping as if in response to little or no apparent facial contact with their environment. These be haviors might be indicative of spontaneous pain, but jerking a nd jumping were not observed consistently after the first two postoperative days. Scratching was noted in the surgical area of 15 CCI-operated rats, and in 9 sham-operated rats (n=19/treatment). This self-i njurious behavior was mo re persistent in the CCI-operated rats. However, these injuries were superficial and healed by the end of the one week recovery period. Effect of Surgical Treatment on Operant Re sponses to Cold, Neutral, and Hot Facial Stimulation Operant responses to 10, 37, and 48C stimulation were assessed pre and post-operatively and the effects of treatment, time, and their in teraction were evaluated using a general linear models (Figure 4-2 inset table). For both 10 and 37C stimulation, general linear models revealed only a significant effect of time on licks, and within group ANOVAs indicated that these effects were only present in the sham-treated rats, not thei r CCI-treated counterparts. The increases within the sham group at two, three (1 0C), and four weeks (37C) post-surgery are likely the consequence of normal appetitive incr ease that occurs with time (Figure 4-2A, diamonds and squares). For 48C stimulation, genera l linear models revealed a significant effect of treatment and time, and a significant interactio n between the two on licks (Fig. 2 inset table). However, no within group differences from base line were revealed, nor were any between group differences at any of the given post-opera tive time points (Figure 4-2A, circles). For 10 and 37C stimulation, general linear mode ls reveal significant effects of treatment and time on stimulus contacts, as well as significant interactions between the two (Figure 4-2, inset Table). Within group differences revealed that the contacts of CCI-treated rats were significantly greater than baseline from one to three week s post-surgery, with between group 81


differences also apparent in th is time frame (and at four week s for 37C stimulation). There were no within group differences for sham-treated animals (Figure 4-2B, diamonds and squares). In contrast, for 48C stimulation a general linear model revealed only a si gnificant effect of time on stimulus contacts, but was not supported by wi thin group analysis (Fig ure 4-2, inset Table). The only significant difference in stimulus cont acts with 48C stimul ation was between the treatment groups at four week post-surgery (Figur e 4-2B, circles). Take n together, these finding indicate that following CCI-treatment, rats contact the 10 or 37C stimulus more frequently in order to obtain the milk reward. This effect is not observed in sham-treated rats, nor is it the case for 48C stimulation. For 10C stimulation, a general linear model rev ealed a significant eff ect of treatment and time, as well as a significant interaction on success ratios (Figure 4-2, inset Table). CCI-treated rats had lower success ratios at one and two weeks post surgery, and between group differences were apparent at all post-opera tive time points (Figure 4-2C, di amonds). For 37C stimulation, a general linear model revealed a significant effect of treatment on success ratios, no significant effect of time, but a significant interaction between the two (Figur e 4-2, inset table). CCI-treated rats had lower success ratios for all post-operative time points, with between group differences apparent as well (Figure 4-2C, squares). For 48C stimulation, a genera l linear model revealed the same pattern of significance as with 37C st imulation. However, there were no difference from baseline within e ither treatment group and only between group differences observed at two and three weeks post-surgery (Figure 4-2C, circles) Taken together, these findings indicate that CCI-treated rats are less successf ul at obtaining the milk reward in the presence of 10 and 37C stimulation, but task completion with 48C stimula tion is not rendered much more difficult with injury than without. 82


In summary, for 10 and 37C stimulation, the lack of an injury e ffect on licks coupled with a post-injury increase in stimulus contacts (i.e. reward attempts) indicates that the CCItreated rats must try harder postoperatively, to obtain the same amount of reward they could preoperatively. As a result, thei r post operative success ratios ar e reduced. These effects are specific to CCI treatment and are not observed with sham injury. In contrast, for 48C stimulation, changes in licks, stimulus contac ts and success ratios were not observed following injury. Therefore, CCI treatment was able to induce allodynia to cold and neutral stimulation, with maximal cold allodynia obs erved two weeks post-injury a nd lasting allodynia to neutral stimulation following injury. It was not able to induce heat hyperalgesia. Effects of Surgical Treatment on Op erant Response to a Rough Stimulus The findings with respect to neutral stimula tion could be the conse quence of mechanical, rather than thermal allodynia. Therefore, surgi cally treated rats were al so tested preand postoperatively with a rough stimulus to assess mech anical sensitivity. No preoperative differences in operant responses were observed between future shamor CCI-treated rats A general linear model revealed significant effects for treatment, time, and significant inte raction between the two (Figure 4-3, inset Table). There we re no significant effects or inte raction for licks, while there were significant effects and interaction for bouts and licks per bout (Figure 4-3, inset Table). Within-subjects analysis revealed no significant effect of time on any outcome measure in the sham-treated rats. In contrast, CCI-treated rats exhibited significantly elevated bouts at one week post-surgery as compared to baseline, and significantly reduced licks per bout at one week post-surgery as compared to all other time-point s (Figure 4-3 B,C). Be tween-group differences were also apparent for these time points. Taken together, these findings indicate that neuropathic injury produced a transient allodynia to the brush of a rough stimulus. 83


Effect of Surgical Treatment on Innate and Aversive Behaviors in the Presence of Thermal and Mechanical Stimulation In addition to the operant outcome measur es, general behavior in the testing box was observed for each session. Early in post-operative testing, it be came apparent that many of the CCI-operated rats were exhibiti ng behaviors likely indicative of stimulus-specific pain. In particular, many CCI-operated rats adopted an up ward head-tilt, not observed in their shamoperated counterparts during this time period (Figur e 4-4, Objects 4-1 and 42). Rats exhibiting this strategy would initially plac e their faces with the dermatom e innervated by ION against the thermode as they had been trained, but as co ntact with the 10C thermode continued, the rat would tilt his head upward, shifting contact to more caudle portion of the face not innervated by the constricted portion of the nerv e. Early in testing some indi viduals exhibiting this head tilt would jerk back on first contact. This head tilt was also accompanied by paw wiping, pushing, or in some cases biting at the thermode. A behavioral scoring system was used to assess the occurrence of innate and aversive behaviors occurring in the presen ce of the thermal and mechanical stimuli during the first five minutes of operant testing. In order to establish a normal pattern of scores for the stimuli tested, five nave rats were scored along with surgically treated rats. There we re no significant effects of time on the innate, aversive, or combined scores of the nave rats, thus the data were pooled for each stimulus for comparison across stimuli (T able 4-2). All scores were significantly greater in nave rats with 10C stimulation than with 37 or 48C. For the innate component this is likely due to the greater frequency of head and wet-dog shaking that occurs with 10C stimulation (Table 4-2). Head tilting did not occur in nave rats, but thermode wiping was observed, particularly with 10C and rough stimula tion (Table 4-2). Biti ng did not occur with thermal stimulation, but was observed with the ro ugh stimulus. Effort was made to distinguish 84


between exploratory nibbling and av ersive biting of the velcro, but false identification of a bite as aversive may still account for these responses in nave rats. We also assessed changes in the behavior scores following surgical treatment. General linear models revealed a signifi cant effect of treatment on the innate score with 10 and 48C stimulation, but not 37C or rough mechanical stim ulation (Figure 4-5 inse t table). Post-hoc analysis revealed that CCI-treated rats exhi bited greater scores than their sham-treated counterparts at one and three weeks post-surgery with 10C s timulation, and at one and two weeks with 48C stimulation (Figure 4-5 A). Gene ral linear models also indicated there were no significant effects of time with any of the stimuli, although with in-subjects analysis revealed that CCI-treated animals also had significantly greater innate scores with 10C stimulation at one and three weeks relative to pre-surg ical scores (Figure 4-5 A). There were no significant withinsubjects effects for sham-treated rats with any stimulus. There was also a significant interaction between treatment and time with 10C stimulatio n only. Taken together, these findings indicate that CCI-treatment can intermittently increase the occurrence of innate behaviors with cold stimulation, but does not substant ially alter innate responses w ith 37, 48C, or rough mechanical stimulation. For the aversive component of the behavioral score, general linear models revealed a significant effect of both treatment and time, as well as a significant interaction for all stimuli (Figure 4-5 inset table). Post-hoc analysis revealed that CCI-treat ed rats had significantly greater aversive scores than their sham-treated counter parts for all post-operative time points and all thermal stimuli, and from one to three weeks for rough stimulation (Figure 4-5 B). Withinsubjects analysis revealed that aversive scores were significantly greater in CCI-treated rats postoperatively and not changed in sham-treated anim als. Taken together, these findings indicate 85


that aversive behaviors are incr eased following injury for all of the stimuli tested. Closer examination of the aversive behaviors indicates that this increase in score is due primarily to the induction of head tilting behavior following surgical treatment and to a lesser extent an increased frequency in thermode wiping, regardless of stim uli (Table 4-3). Head tilting occurs more frequently and persistently in CCI-treated rats than their sham counterparts. Thermode wiping occurs in both groups post-operatively, although this behavior is generally more frequent in the CCI-treated animals. Biting never occurred with 37C stimulation and rarely with 10 or 48C stimulation. In conclusion, nerve injury was able to produce a persistent occurrence of aversive behaviors with all stimuli tested, and only transi ently effected innate behavior exhibited with cold stimulation. These findings are in agreem ent with the decreased success observed in the operant task with 10 and 37C stim ulation. However, the robust induction of aversive behaviors with 48C stimulation would appear to contrast with the lack of effect on success in the operant task. This may suggest that the changes within peripheral and subcorti cal structures following injury do not lead to a substant ial change in operant response to a stimulus that is already noxious. Effect of Surgical Treatment on Nerve Inflammation and Injury Nerves from CCIand sham-treated rats were assessed for degree of inflammation and nerve injury. The knots of the vicryl ligatures were visibly present only around nerves taken two weeks post-surgery (Figure 4-6 A) ; at all later time points suture material was identified microscopically (Figure 4-6 B). Of the two sham -treated rats with sutu re present (Figure 4-6 inset table), one scored unusually high and a ppeared as an outlier when the scores were submitted to box plot analysis. A re-examinati on of this individuals behavior revealed a consistent occurrence of head tilting and re duction in success ratios with all stimuli post86


operatively. This individual was therefore excluded from the sham treatment group for subsequent analysis. For nerves taken at 27 weeks post-surgery, a general linear model was used to assess effects of and interaction betw een treatment and time on the composite score. There was a significant effect of treatment (F1 = 15.217, p = 0.001), but no significant effect of time (F3 = 0.272, p = 0.845) and no significant interaction (F3 = 0.404, p =0.752). Thus, all nerves collected at 2-7 weeks have been pooled to illustrate the difference between each group (Figure 4-6 inset table). Inju ry, inflammation, and composite sc ores were all significantly greater in constriction injured ne rves taken within 2-7 weeks post-surgery as compared to timematched sham treated nerves or nerves taken at 12 weeks post-surger y of either surgical treatment (Figure 4-6 inset table). As a negative control, nerves from two naive individuals were included and consistently assigned scores of zer o for all assessments and declared healthy by the blinded investigator. Effect of Surgical Treatment on TRPV1 a nd TRPM8 expression at Two Weeks PostTreatment Trigeminal ganglia from CCIand sham-treat ed rats were taken at two and seven weeks post-surgery and sections were either staine d with antibodies against TRPV1 or TRPM8. Assessment of immunoreactivity was re stricted to the area innervated by the infraorbital (treated) region of the trigeminal nerve. An increased percentage of TRPV1 positive cells were noted in ganglia taken at two weeks post-CCI, as compared to all other treatment groups, even ganglia taken seven weeks post-CCI (Figure 4-7). Wh ile no changes in the percentage of TRPM8 positive cells was noted, the intensity of TRPM 8 staining was greatest in ganglia taken two weeks post-CCI compared to all other treatment conditions (Figure 4-8) Ganglia from CCItreated rats also exhibited more intense TRPM8 staining than time-matched sham tissue at seven weeks post-surgery. Taken together these findings indicate that TRPV1 expression is detectable 87


in more cells following CCI, but this expres sion wanes between two and seven weeks postsurgery. TRPM8 expression may also be up-regu lated in cells that normally express the channelor the protein may be more concentrated on the plasma membrane following CCI. Peak intensity of TRPM8 staining also occurs at two weeks post surg ery, and while it wanes by seven weeks, is still significantly greater than sham at this time point. Effect of Pregabalin and Gabapentin Treatme nt on Operant Responses to Cold Facial Stimulation in Surgically-T reated and Untreated Rats A subset of rats was treated with pregab alin (10 mg/kg), gaba pentin (30mg/kg), or vehicle (water) and tested with 10C stimulation at two week postsurgery, when cold allodynia is maximal. To control for drug effects in the abse nce of surgical treatment, nave rats were also included for each drug treatment. A general linear model revealed significant effects of surgical treatment on stimulus contacts and success ratios, but not licks, as was reported above. There was also a significant effect of drug treatment on all three out come measures and a significant interaction between surgical and drug treatment for stimulus c ontacts and success ratios (Figure 4-9, inset table). There was a significant effect of gabapentin on licks in surg ically treated and nave rats relative to no drug treatment, and a significant e ffect of pregabalin on sham-treated and nave rats relative to no drug treatment (Figure 4-9 A). Within group effects of drug treatment on stimulus contacts were near the significance cut off for CCIand sham-treated rats (p = 0.05 and 0.047 respectively), and were not significant fo r nave rats (Figure 4-9 B). Between-group effects on stimulus contacts were not observed for any drug treatment. Within group effects revealed a significant effect of gabapentin on su ccess ratios for both surgically treated and nave rats, and a significant effect of pregabalin on suc cess ratios for surgically treated rats only. There were no significant between group effect s on stimulus contacts (Figure 4-9 C). 88


Taken together, these data suggest that bot h gabapentin and pregabalin can enhance successful task completion in surgically treate d rats, and gabapentin can also enhance task completion in nave rats. The elevation in licks with drug treatment suggests that some of these effects may be appetitive in nature, particularly w ith respect to the sham-treated and nave rats. However the trend for decreased stimulus contac ts in CCI-treated rats with drug treatment suggests that the significant incr ease in success ratios obs erved in this group is the result of drugrelated analgesia. The significant pregabalin e ffect observed in sham-treated rats may also indicate that although co ld allodynia is not pres ent in this group, the less-severe injury is sufficient to cause changes that render them sens itive to the effects of the drug. In conclusion, we demonstrate that gabapentin and pregabalin are able to ameliorate the cold allodynia experienced by CCI-treated rats and that effects in sham-treated and nave rats indicate affective or appetitive effects of these drugs on operant behaviors wi th a 10C stimulus. Effect of Surgical Treatmen t on Mechanical Preference When the effect of time on m echanical preference was evalua ted for each surgical group, no significant effect of time was observed for licks or testing duration for either stimulus. Thus, all post-operative time points were pooled to compare the eff ects of treatment on licks and duration for each starting stimulus condition. As with nave rats in chapter 2, preference was influenced by the starting stimulus in surgically treated rats. For both CCI and sham-treated rats, soft stimulation was greatly preferred when this was the starting stimulus (Table 4-4). However, when rough stimulation occurred first, soft preference occurred more consistently among CCItreated rats post-operatively, than among sham-treated rats (Table 44, parenthetical values). Of the three groups, CCI-treated rats exhibited a soft stimulus preference the most frequently (Table 4-4). 89


When rough stimulation occurs first, CCI-t reated rats exhibite d a pronounced preference for soft stimulation not observed in sham or nav e rats. Interestingly, CCI-treated rats also had significantly greater total licks and significantly lower unstimulated time relative to sham-treated and nave rats (Figure 4-10 A,B). When soft st imulation occurs first, all three surgical groups exhibit a strong preference for the soft stimulus. In this case, CCI-treated rats also exhibit significantly greater total licks than nave rats (Figure 4-10 C). They al so spent significantly more time on the soft stimulus and significantly le ss time unstimulated as compared to nave rats (Figure 4-10 D). These findings indicate that CCI-treated rats may have a more pronounced aversion to the rough stimulus post-operatively than do sham or nave rats. The difference in total licks and unstimulated time likely indicates th at the increased drive in CCI rats to avoid rough stimulation leads to increased exploration and more reward intake than the other two groups not strongly motivated away from the rough s timulus. It is also possible that the chronic pain state experience by CCI-treated rats may enhance sensitivity to reward. Effect of Surgical treatment on Thermal Pref erence and the Influence of Task Novelty and Drug Treatment To evaluate the effect of surgical treatm ent on thermal preference (10 and 48C) when cold allodynia is maximal (two weeks post-surger y), one group of rats was evaluated both preand post-operatively, along with nave controls. It was expe cted that the pronounced cold allodynia and lack of heat hypera lgesia exhibited by CCI-treated ra ts in the single stimulus task might mean that CCI-treated ra ts would avoid 10C and prefer ence to 48C at two weeks postsurgery. However, this was not the case in this experienced group of rats. CCI-treated rats still exhibited a preference for 10C stimulation, also demonstrated by sham-treated and nave rats (Figure 4-11). Also contrary to expectation was the decrease in unstimulated time exhibited by CCI-treated rats as compared to both other treatment groups. Th e pattern of stimulus contacts 90


and success ratios, as well as the presence of aversive behaviors indicate that CCI-treated rats were suffering from pain (Figure 4-12). The lack of a substantial influe nce of CCI-treatment on preference suggests either that 10C in a neuropathic state is still more desirable than the noxious 48C stimulus, or that experienced CCI-treated rats are able to m odulate their behavior to get the most reward with a to lerable degree of pain. A second, inexperienced group underwent novel th ermal preference testing at five weeks post-surgery, when pain was no longer evident in the CCI-treated rats. In contrast to the experienced CCI-treated rats, inexperienced CCI-treated rats did not exhibit a thermal preference, while their sham-treat ed counterparts exhibited a sli ght preference and nave rats exhibited no difference in preference from their experienced counterparts (Figure 4-11). This suggests that although CCI-treated rats may no longer exhibit signs of pain in the single operant task, preference can be influenced by previous pain. When inexperienced rats were treated with pregabalin, cold preference was enhanced in both CCI-treated and nave rats, but not affected in their sham-treated counterparts (Figure 4-11). This would suggest that preg abalin may be able to modulate affective pain processing in CCI-treated and nave rats w ith respect to cold stimulation. The lack of effect in sham-treated rats could be due to a subtle deficit in sensory discrimination or alteration in affect towa rds cold and heat that may accompany surgical treatment. Discussion of Neuropathic Pain and Operant Behavior We show here that rats receiving CCI of the trigeminal nerves develop pronounced sensitivity to 10C stimulation that is maximal at two weeks post-surgery. They also exhibit a robust, long term allodynia to 37C stimulati on, which was somewhat unexpected and did not follow the pattern of behavior exhibited toward the rough mechanical stimulus. This sensitivity is indicated by both decreased succes s with the operant task and pers istent occurrence of aversive behaviors in the presence of these stimuli. In contrast, CCI-treatment did not hinder successful 91


task completion in the operan t task with 48C stimulation, even though aversive behaviors persisted. Sham-treated rats did not demonstr ate lasting changes in operant performance and aversive behaviors were observed transiently in so me individuals. We also show here that treatment with pregabalin and gabapentin can a lleviate cold allodynia in neuropathic animals, and also produce effects in sham-treated and na ve rats suggestive of affect-modulation. While the poor success in the single stimulus task was transient for rough stimul ation, CCI-treated rats still exhibited a pronounced prefer ence for soft stimulation not shown in sham or nave rats. Finally, experience with the thermal preference task led to a nearly normal pattern of preference in CCI-treated rats, despite si gns that they were experien cing allodynia. In contrast, inexperienced CCI-treated rats who had previously been in pain exhibited a lack of thermal preference, despite the fact that their sham a nd nave counterparts did exhibit a preference. Pregabalin was able to reset nor mal preference in CCI-treated rats. Effects of CCI on Behavioral Response s to 10, 37C and Mechanical Stimulation According to the current body of literature, CCI to either the trigeminal or the sciatic nerve is typified by allodynia and/or hyperalgesia to all stimulus modalities, as established by measuring reflexive withdrawal or assessment of grooming behavior evoked by stimulation applied by the experimentor. For cold stimula tion, this could involved a cooling compound such as acetone (Chichorro et al., 2006; Walczak and Beaulieu, 2006; Lim et al., 2007), or a cold floor (Bennett and Xie, 1988; Allchorne et al., 2005). Cold allodynia has also been demonstrated following both unilateral and bilateral CCI usin g operant methods based on time spent escaping cold stimulation (Vierck et al., 2005; Jabakhanji et al., 2006; Walczak and Beaulieu, 2006). Particularly, the findings of this study are in agreement with that of Vierck and collegues, who reported both post-operative increa ses in lick-guard responses and escape duration with cold stimulation following a bilateral sciatic CCI. 92


However, our findings with respect to 37 C stimulation were somewhat surprising. Neuropathic pain studies typically focus on co ld allodyna, heat hyperalgesia, and mechanical allodynia (assessed typically with Von Frey filaments); most do not even test responses to thermal stimuli at body temperature, or dont repo rt it if they do. At least one other group reported reduced latency to withdraw from 38C stim ulus in the streptozotocin model of diabetic neuropathy (Beyreuther et al., 2006) and in a model of chemotherapeutic pain (Beyreuther et al., 2007). The fact that these animals and our injured animals experi ence alterations in mechanical sensitivity may still indicate that this effect is primarily a consequence of mechanical allodynia. Dynamic mechanical allodynia is a common feature of neuropat hic pain (Bowsher, 2005; Lang et al., 2006; Samuelsson et al., 2007) and also a consequence of anim al injury models (Field et al., 1999), usually assessed by brushing a soft bris tled brush or cotton swab across the affected skin. The lack of a prolonged decrease in su ccess with rough stimulation suggests that the profound sensitivity to 37C is not simply m echanical allodynia, but thermal allodynia. Additionally, mechanical sensit ivity does not explain the different behavior profiles exhibited with 10 or 48C stimulation. It has been noted in normal human subjects that dynamic mechanical stimulation provided by a thermal pr obe can reduce pain sensations elicited by moderate cooling (termed innocuous cold noc iception or ICN) and by menthol during ICN (Green and Pope, 2003; Green and Schoen, 2005; 2007) but it remains to be seen if such a phenomenon occurs in individuals suffering from ne uropathic pain. The capacity for a delta or c fiber activity to mutually suppress evoked cortical potentials to the other fi ber type (Tran et al., 2008) may contribute to this phenomenon. If this phenomenon of ICN is intact or only transiently disrupted following neuropathic injury, it could contribute to an a pparent recovery of normal cold responses over time. 93


Another explanation for the las ting allodynia apparent with 37C is that CCI-treated rats are hyper-reactive to the operant testing chambe r because of repeated testing and first postoperative exposure to exquisitely painful stimuli. However, if that were the case we would likely expect that this exaggerated re sponse would be apparent with the same time course for other stimuli and it is not. Also, CCI-treated rats c ontinue to make attempts to obtain reward, even without fasting (as is the case with 37C and rough stimulus te sting), so we may rule out a decrease in the hedonic value of sweetened condensed milk in the presence of pain. Effects of CCI on Hot-Mediated Behaviors Withdrawal from radiant heat is typically enhanced following unilateral sciatic CCI (Bennett and Xie, 1988), which has also been demonstrated following unilateral CCI of the trigeminal nerve (Imamura et al., 1997). In c ontrast, in a study comp aring unlearned behaviors and operant behaviors in respons e to heat, bilateral CCI of th e sciatic nerves did not produce increased lick-guard to or escape from nociceptive heat (up to 47C) (Vierck et al., 2005). The operant responses to 48C reporte d are in agreement with Vierck and collegues; there was no significant effect of bilateral CCI-treatment on success ratios post-opera tively (Vierck et al., 2005). In contrast to Vierck and collegues, we did observe a persistent post-operative elevation of unlearned aversive behaviors in the CCI-treated rats and a transient elevation in these behaviors within sham-treated rats post-operatively (Vie rck et al., 2005). It is possible that the aversive behaviors reflect an exquisite sensitivity to the smooth metal thermode, rather than a response to the heat of the stimulus. However, enhancement of withdrawal to radiant heat, which lacks a mech anical component, has been observed following trigeminal CCI (Imamura et al., 1997; Liang et al., 2007). Both the aversive behaviors exhibited here and reflexive withdrawal directly measur e segmental processing of painful stimulation, whereas the operant assessment reflects a direct measure of cortical integration and assessment 94


of signals from the periphery. Thus the differences observed between unlearned and operant behaviors here may indicate that there is an enha ncement of heat nocicepti on at the level of the brainstem, but this enhancement does not resu lt in cortical changes necessary to impair successful performance of the operant task with respect to painful heat. The different effect of neuropathic injury on cold versus heat rela ted operant behaviors may be related to differential gating of these st imuli by the thalamus. Le sioning various thalamic nuclei can induce transient analgesia with differing efficacies to different stimulus modalities in a neuropathic state (Saad et al., 2006) and patien ts with specific thalamic lesions have been shown to exhibit impairments in cold sensitivit y, but normal heat perception (Kim et al., 2007). The lack of effect on heat-mediated operant beha viors could also be related to the convergent processing of the affective, cognitive, and intensit y aspects of painful stimulation. For both cold and heat testing the rats are fasted to motivate pe rformance. In the case of heat stimulation the neuropathic injury does not appear to conflict sufficiently with the drive to satisfy hunger, while cold stimulation does. Perhaps if the rats were not fasted prior to heat testing, we might observe a robust difference between sham-treated and CCI-treat ed rats in the operant task. We have also shown that noxious heat is preferred over noxious cold in intact rats (Rossi et al., 2006). Effects of CCI on Cold-mediated Behaviors and TRPM8 Expression The peak in cold sensitivity at two weeks post surgery exhibited by CCI-treated rats is accompanied by an increased intensity, alt hough not increased cell counts of TRPM8 immunoreactivity in the trigeminal ganglia. TRPM 8 has a primary role in the perception of both innocuous and painful cold stimuli (McKemy et al., 2002; Peier et al., 2002), however, results are mixed with respect to the role of this cha nnel in the development of cold sensitivity in a neuropathic state. Some have shown that TR PM8 mRNA expression is not changed at one or two weeks (Obata et al., 2005; Kats ura et al., 2006) following spinal nerve ligation in rats, nor is 95


protein expression at one week (K atsura et al., 2006). In cont rast, others have shown that TRPM8 mRNA (Frederick et al., 20 07) and protein expression are in creased in DRGs and in the dorsal horn of the spinal cord ipsilateral to scia tic CCI at two weeks post-injury (Proudfoot et al., 2006). The increased intensity of TRPM8 immunoreactivity we observed following CCI could be reflective of increased TRPM8 expression observe d by others (Proudfoot et al., 2006; Frederick et al., 2007). However, the lack of change in total number of TRPM8 positive cells would seem in contrast with the findi ngs of Proudfoot and colleagues (Proudfoot et al., 2006). This discrepancy may be due to differences in anti body sensitivity and/or inclusion criteria for positive TRPM8 immunoreactivity. They repor t that TRPM8 immunoreactivity occurred primarily in c fibers (indicated by small, peri pherin positive cells) of na ve DRGs (Proudfoot et al., 2006). In contrast we observe that TRPM8 is expressed in wi de range of cell sizes in nave TG, likely indicating a mix of positive c and a delta fi bers. It has also been noted that there is a greater percentage of TG cells expressing TR PM8 mRNA than in the DRG (Kobayashi et al 2005), which could also underlie slig ht differences in our findings. An alternate explaination for the increased intensity of staining is that TR PM8 channels already pr esent on cells become clustered together on the surf ace of the cell following injury, rather than being diffusely distributed within th e plasma membrane. Cold Sensitivity and TRP Expression In addition to increased aversi ve behaviors to nociceptive heat we observe an increase in TRPV1 immunoreactivity in CCI-treat ed rats at two weeks post-surg ery. Most reports indicate an increased expression of either mRNA, protein, or both in DRG, TG, or spinal cord dorsal horn ipsilateral to the model injury (Ma et al., 2005; Wilson-Gerwing et al., 2005; Biggs et al., 2007). In particular a recent study indicated that TRPV1 is upregulated in uninjured neurons within the 96


ipsilateral TG following mental nerve transection, appearing more frequently in medium sized cells than in the absence of injury (Kim et al., 2008). Anot her study, however, did not observe any change in mRNA expression following un ilateral CCI using a novel multiplex ribonuclease protection assay to assess expression of multiple TRP channels in a single sample (Frederick et al., 2007). TRPV1 has been long implicated in the deve lopment of both enhanced heat nociception and mechanical allodynia following neuropathic injury. Several st udies have shown alleviation of neuropathic pain symptoms following treatment with TRPV1 antagonists like BCTC (Pomonis et al., 2003) and capsazepine (Walker et al., 2003). However, such compounds are now known to also act on TRPM8 and may act on ot her TRP channels. More specific silencing of TRPV1 with antis ense RNA has been shown to alleviate neuropathic pain, including cold pain (Christoph et al., 2006; Christoph et al., 2007). Another study al so implicated TRPV1 in the development of cold allodynia following sciati c CCI; these authors observed an increase in neurons responding to both capsa icin and menthol (i.e. contai ning both TRPV1 and TRPM8) in ipsilateral DRGs (Xing et al., 2006; Xing et al ., 2007). They also observed that these cells exhibited enhanced responses to menthol and cold stimulation, re lative to what is observed in nave DRG cells (Xing et al., 2007). Unfortunately, due to limitati ons in our staining procedures, we are unable to accurately assess co-label ing of TRPM8 and TRPV1 directly, but the appearance of TRPV1 immunoreactivity in larger cells that often exhibit TRPM8 reactivity may indicate an increase in co-labeling following CCI. The increased expression of TRPV1 and enhanced intensity of TRPM8 immunoreactivity correspond to the maximal enhancement of cold sensitivity at two week post-surg ery, suggesting that both channels may play a role in peripheral sensitization to cold stimuli. 97


Effect of Drug Treatment on Cold Allodynia We also demonstrate that pregabalin and gabapentin have a significant effect on operant behaviors (including succe ss) not only for CCI-treated rats, but in sham-treated and nave rats as well. In the CCI-treated rats, stimulus cont acts following gabapentin and pregabalin were reduced by about half and a quarter respectively relative to both baseline and vehicle, although these do not reach statistical signifi cance due to the greater variability in drug treated animals. In contrast, stimulus contacts were not changed in nave rats. This suggests that these compounds, which are currently FDA approved for the treatmen t of neuropathic pain, do provide pain relief to injured animals. This also supports previous findings regarding the an ti-allodynic potential of pregabalin (Ling et al., 2008; Tanimoto-Mori et al., 2008) and gabapentin (Ling et al., 2007) with respect to cold stimulation, evaluated using standard reflex te sting in models of neuropathic pain and using a cold preferen ce task (Walczak and Beaulieu, 2006). There were significant effects of licks for both drug treatments in surgically treated and nave animals, which suggests either that these drugs may have appetitive effects not previously noted or this may be a consequence of affect mo dulation. In studies that more closely resemble our paradigm, a lever press is paired with food reward and additional foot shock punishment in some trials, pregabalin (10 30mg/kg) signi ficantly increase leve r press during painful punishment (Field et al., 2001; Evenden et al., 2006). Both pregabalin and gabapentin (100mk/kg) treatment also decreased escape behavior in a fear conditioning assay (Nicolas et al., 2007). A recent animal model of post-traumatic stress disorder showed short term anxiolytic effects of pregabalin, in higher doses than pres ented here (100 & 300mg/ kg) (Zohar et al., 2008). Gabapentin reduced both anxietylike behaviors and increased mech anical withdrawal thresholds in rats with sciatic CCI, alt hough unlike the current study, they did not observe anxiety-related effects in sham-treated animals (Roeska et al ., 2008). This difference could be related to 98


differences in severity of the two sham procedures or differences in the stressfulness of the open arm test versus our facial stimulation task. In dental patients, pregabalin (600 mg oral) has also been shown to reduce anxiety within a 3-4 hour period following administration (Nutt et al., 2008). Thus our findings support animal models reporting anti-allodynic effects, as well as anxiolytic effects in both rodent s and humans. The influence of pregabalin on affective pain judgment is also supported by our finding that it can re-establish nor mal preference in CCItreated rats with minimal experi ence in a thermal preference task. Modulation of Preference After Injury and Affective Aspects of Pain Despite the fact that apparent sensitivity to a single rough stimulus is tr ansient (although impairments in success could be underestimated), preference for the soft stimulus was apparent in CCI-treated rats when star ting on the rough stimulus, sugges ting that the rough stimulus was avoided. This was not exhibited in nave or sham treated rats tested concu rrently. It should also be noted that aversive behaviors were exhibited in these CCI trea ted rats and were observed with both stimuli. This indicates th at neuropathic injury can cause changes not only in absolute sensitivity to stimuli, but can influence d ecision making processes related to stimuli. With respect to preference for 10C versus 48C, we found that pr evious experience may have led CCI-treated rats maintain preference fo r 10C, despite the fact that their increased contacts, decreased success, and presence of av ersive behaviors indica tes that they clearly experience pain with this stimulus at two weeks post-surgery. As observed in the single stimulus condition, there was no reduction in success for th e 48C stimulus, but as before, aversive behaviors were noted. In contrast, inexperienced CCI-treated rats, who had previously experienced pain with 10C, did not exhibit a preference, despite the fact that their sham and nave counterparts were able to do so within the same limited number of trials. These rats were repeatedly exposed to 99

PAGE 100

10C stimulation for the four weeks prior to ther mal preference testing, a nd the CCI-treated rats in this group exhibited deficits in operant success and aversive behaviors indicative of pronounced pain as reported above. By four weeks their resp onses were normal and aversive behaviors were no longer consistently observed. This should be confir med by testing preference inexperienced animals two weeks postsurgery when pain is maximal. As discussed in chapter 2, inflammatory pain m odels can be used to condition aversion to a compartment (Vaccarino et al., 1992; Colpaert et al., 2006; van der Kam et al., 2008). These findings may indicate that previ ous experience of cold allodynia following neuropathic injury could condition aversion to 10C in our preference assay. Conditioned changes in pain sensitivity have also been demonstrated in humans using intrinsic re inforcement of pain by coupling ratings with changes in stimulus intensity (Hlzl et al., 2005; Becker et al., 2008). Additionally, the role of memory in the maintenan ce of chronic pain has been recently raised by a report of two individuals that experienced a cessation of chr onic pain accompanying amnesia (Choi et al., 2007). While further follow up is needed to determine the preference and behavior of inexperienced rats tested at 2 weeks post-surg ery, the findings reported here suggest that the stimulus novelty and previous experience of neuropathic pain may be used to study the relationship between memory and pa in in experimental animals. Conclusion We show here that operant responses to cold stimulation are in ag reement with previous studies using reflex and unlearne d behavioral assessment of neur opathic allodynia, while operant responses to heat are not signifi cantly altered by neuropath ic injury. We also show for the first time that both operant and unlear ned behaviors to neutral stim ulation are altered following neuropathic injury in a manner su ggestive of lasting a llodynia. Cold allodynia was reversible with pregabalin and gabapentin. Mechanical preference and thermal preference changes were 100

PAGE 101

also noted, reflecting the influence of experience on operant behaviors in a neuropathic state. These behavioral changes are also accompanied by changes in both TRPV1 and TRPM8 expression at two weeks post-surgery, which may underlie cold allodynia observed at this time point. These findings indicate that operant assessment of facial pain can provide insights not indicated by reflex based methods potentially indicative of cha nges in cortical processing of pain. This may lead to a better understanding of the relationship between peripheral sensitization and central processing that lead to enhan ced, and sometime intractable, chronic pain. 101

PAGE 102

Figure 4-1. Explanation of injury and inflammation scores used to as sess the health of surgically treated infraorbital nerves. A-B are images of an untreated nerve, with scores of 0 for 102

PAGE 103

both injury and inflammation. C-J are exam ples of varying degrees of injury and inflammation apparent in chr onic constriction injured nerves as indicated in the table below. A, C, E, I (left panels) are lower magnification to illustra te nerve injury, as indicated by the infiltration of chronic infl ammatory cells, lymphocytes. B,D, F-G, J (right panels) are higher magnification to show the degree of inflammation. G depicts giant cells/foreign body reaction and inflammatory cells. Injury and inflammation scores for the four sections viewed were added together, with a maximum possible composite score of 24/animal. All nerves were stained with hematoxylin and eosin. 103

PAGE 104

Figure 4-2. Effect of surgical treatment on operant responses with 10, 37, and 48C stimulation. Diamonds = 10C, squares = 37C, circles = 48C. A) Licks were not significantly affected by CCI treatment (filled symbols) for any stimuli. Sham-treated rats (open symbols) exhibited time related increases in licks for 10 and 37C stimulation, as indicated by lines. Between group differen ces occurred with 10C. B) Stimulus contacts were significantly affected by CCI-treatment with 10 and 37C, but not 48C stimulation. The only effect observed for sh am-treated rats was a significant increase at four weeks with 48C stimulation. Between group differences occurred for the majority of post operative time with 10 a nd 37C stimulation, and only at four weeks with 48C stimulation. C) Success ratios (l icks/contact) were significantly affected by CCI treatment with 10 and 37C, not 48 C stimulation. There was no effect of sham treatment. Between group differenc es occurred for all post-operative time points with both 10 and 37C s timulation and at two and three weeks with 48C. indicates a significant betw een group difference for the time point and stimulus indicated (unpaired t-test). ** indicates both a significant between group difference at the indicated time point, and a within group difference for the CCI-treated rats relative to baseline (ANOVA). Significance is p <0.05. All data is mean SEM. Table indicates statistical outcome of general linear models evaluating time and treatment effects and interactions for the three stimuli shown. 104

PAGE 105

Figure 4-3. Effect of surgical tr eatment on operant responses with rough mechanical stimulation. A) Licks were not significantly effect by treatment or time. B) Bouts were significantly increased by CCI-treatment at one week post-surgery, and significant between group effects were observed at one and two weeks. C) Success ratios, approximated by licks/bout, were significantly reduced by CCI-treatment at one week post-surgery. Inset table shows the stat istical results of a general linear model evaluating the effects and interaction of tr eatment and time. indicates significant between group different or significant difference from baseline. + indicates significant difference from all other time poi nts. Significance is p < 0.05 and all data are mean SEM. 105

PAGE 106

Figure 4-4. CCI-treated rats exhi bit aversive behaviors towards the stimulus not observed in Sham-treated rats. (A) An example of a CCI-treated tilting his heat upward, shifting contact away from the mystacial pad aff ected by the constriction injury. (B) This sham-treated rat, which underwent surgery on the same day as the rat on the left and was also tested within the same session. Na ve animals have the same posture as the sham-treated animals ( not shown). 106

PAGE 107

Figure 4-5. Effect of surgical tr eatment on innate and aversive be havior scores with thermal and mechanical stimuluation. Table indicates general linear model statistics for each stimulus. A) Innate behaviors were only transiently elevated following CCItreatment with cold stimulation. B) In cont rast, aversive behaviors were persistently elevated for all thermal stimuli following CC I-treatment and from one to three weeks for mechanical stimulation. ** indicates bo th between and within group differences from baseline (one way ANOVA with post-hoc Tukeys test). indicates a between group difference only (unpaired t-test). P< 0.05 and all data are mean SEM. 107

PAGE 108

Figure 4-6 Presence of suture and quantification of inflammation in surgically treated infraorbital trigeminal nerves. A) Example of grossly visible suture (viole t fibers) around nerve removed two weeks post-surgery. B) Exampl e of microscopically visible suture (striate whirls of transparent material with infiltrating lymphocytes) within nerve removed at five weeks post surgery (hemat oxylin and eosin). Table indicates number of animals with suture pres ent in the nerves and inflammation, injury, and composite scores (mean SEM). *indicates p< 0.05 for CCI-treated rats at 2-7 weeks postsurgery relative to all ot her groups for all scores. 108

PAGE 109

Figure 4-7. TRPV1 immunoreactivity in trigeminal ganglia from nave and surgically treated rats. A-E are exemplars of TRPV1 imm unoreactivity in different treatment groups, as indicated, at two (left) or seven (right ) weeks post-surgery. F is graph quantifying the number of TRPV1 positive cells in the trigeminal ganglia for the three treatment groups. + indicates significant difference fr om all other groups, where p< 0.05. Data are mean SEM. 109

PAGE 110

Figure 4-8. TRPM8 immunoreactivity in trigeminal ganglia from nave and surgically treated rats. A-E are exemplars of TRPM8 immunoreactivity in different treatment groups, as indicated, at two (left) or seven (right ) weeks post-surgery. F is graph quantifying the intensity of TRPM8 staining for th e three treatment groups. + indicates significant difference from all other groups, indicates significant difference between the two surgical treatments at seven weeks post-surgery, where p< 0.05. Data are mean SEM. 110

PAGE 111

Figure 4-9. Effect of drug treatment on operant beha vior in surgically treated and nave rats with 10C stimulation at two weeks post-surgery. Inset table shows statistics for general linear model evaluating the effect of surgical treatment, the effect of drug treatment, and their interaction. DF = degrees of freedom. A) Licks we re significan tly increased in all three surgical treatment groups with gabapentin (30mg/kg, i.p.), and in shamtreated and nave rats with pregabalin treatment (10mg/kg, i.p.). There was no effect of vehicle (water, i.p.) on licks in any group. B) The effect of drugs on stimulus contacts did not reach significance for any group. C) Success ratios (licks/stimulus contact) were significantly in creased in all three surgical groups with gabapentin, and in surgically-treated rats with pregabalin. indica tes significant between group effects, + indicates as significant within-g roup increase relative to no drug treatment (determined by one-way ANOVA). Significance is p < 0.05 and all data are mean SEM. 111

PAGE 112

Figure 4-10. Effect of surgical treatment and starting stimulus on mechanical preference, total licks, and time spent unstimulated. When rough is the starting stimulus (A & B), CCI-treated rats lick (A) and spend signifi cantly more time (B) on the soft stimulus than either sham or nave rats. CCI-treated rats also have significantly greater total licks and lower unstimulated time than the other groups. When soft is the starting stimulus (C & D), CCI-treated rats have more total licks than nave rats (C) and spend more time on the soft stimulus and less time uns timulated than the nave rats (D). + indicates significant difference compared to all other treatments and indicates significant difference from nave rats only (as determined by one way ANOVA and Tukeys test post-hoc). P<0.05. All data are mean SEM. 112

PAGE 113

Figure 4-11. Effect of surgical treatment and experience on thermal preference and unstimulated time. A) CCI treated rats that are experience, but in pain still exhibit a cold preference, while inexperience rats have no preference, which is restored to normal by pregabalin treatment. B) Sham-treated rats exhibit similar preferences despite different levels of experience or drug treatm ent. C) Both experience and inexperience nave rats exhibit a pronounced cold pref erence, which is enhanced by pregabalin treatment. + indicates significant differen ce compared to all other treatments and indicates significant difference from the lowest value (as determined by one way ANOVA and Tukeys test post-hoc). P<0.05. All data are mean SEM. 113

PAGE 114

Figure 4-12. Effect of surgical treatment and experience on success for each stimulus in the thermal preference taks. A) Experience a nd inexperience CCI tre ated rats exhibit reduced success ratios on the cold stimulus as compared to all Sham-treated (B) and nave (C) rats. This is ameliorated by pregabalin treatment. + indicates p<0.05 significant difference compared to all other treatments (one way ANOVA and Tukeys test post-hoc). A ll data are mean SEM. 114

PAGE 115

Table 4-1. Seven day post-opera tive recovery progress for su rgical groups as indicated by number of rats with swe lling in the surgical area ra nging from severe to none. Swelling severity 1 day 2 days 3 days 4 days 5 days 6 days 7 days CCI severe 9 34200 0 N = 19 moderate 9 12 9 8 7 2 0 mild 1 45766 6 none 0 0 1 2 6 11 13 SHAM severe 8 00000 0 N = 19 moderate 8 10 2 2 0 0 0 mild 3 67640 0 none 0 3 10 11 15 19 19 Table 4-2. Percentage of occurr ence of behaviors and behavioral scores (mean SEM) with 10C, 37C, 48C, and rough stimulation for na ve rats (n = 5, across multiple sessions). Stimuli FGF FGH FS HS WDS HT Wipe Bite Total n 10C 97 91 34 66 46 0 26 0 35 37C 92 80 32 32 8 0 0 0 25 48C 92 100 28 32 12 0 4 0 25 Rough 90 35 35 25 10 0 20 5 20 Innate score Aversive score Combined score 10C 3.3 0.1* 0.3 0.1* 3.6 0.1* 37C 2.4 0.2 0 2.4 0.2 48C 2.6 0.1 0 2.7 0.1 Rough 2.0.3 0.3.1 2.2.3 p<0.05. FGF = facial grooming with forepaws FGH = facial grooming with hindpaws, FS = forepaw shaking, HS = head shaking, WDS = wet dog shaking, HT=head tilting, wipe = wiping at the stimulus, bite = biting the stimulus 115

PAGE 116

Table 4-3. Percentage of CCIand Sham-treated rats exhibiting head tilting (HT) or thermode wiping (wipe) with thermal and mechanical stimulation, or biting with mechanical stimulation preand post-operatively. Weeks Post-Surgery Stimulus Temp. (C)/Mechanical Surgical Treatment Aversive Behavior Baseline 1 2 3 4 10C CCI HT 0 77 100 75 55 37C CCI HT 0 67 80 90 80 48C CCI HT 0 87 80 70 60 rough CCI HT 7 80 73 90 60 10C Sham HT 0 19 9 5 5 37C sham HT 0 3 3 0 0 48C sham HT 0 23 8 33 0 rough Sham HT 0 0 0 0 0 10C CCI wipe 0 58 55 65 65 37C CCI wipe 7 47 53 60 30 48C CCI wipe 20 47 80 60 70 rough CCI wipe 0 38 58 56 0 10C Sham wipe 31 31 64 40 25 37C sham wipe 0 8 8 11 0 48C sham wipe 8 38 50 0 22 rough Sham wipe 8 15 17 22 0 rough CCI bite 0 0 7 20 40 rough Sham bite 15 31 8 0 25 10C n 16 26 11 20 20 (both 37&48C) n 13-15 13-15 12-15 9-10 9-10 rough n 13-15 13-15 12-15 9-10 4-5 For samples sizes, lower number indicates number of rats in sham group and higher in the CCI group. Table 4-4. Percentage of surgica lly treated rats exhibiting soft, rough, or no stimulus preference post-operatively (n = 30 per treatment). Stimulus preference Treatment group Soft Rough None CCI 73 (13/20) 17 10 Sham 53 (8/22) 40 6 Numbers in parentheses indicate the fraction of surgically-treated rats with a soft preference when starting with rough stimulation. Object 4-1. Video clip of aversive head tilting behavior exhibited by a CCI-treated rat at postoperative day 10, with 10C stimulation. Object 4-2. Video clip of normal operant behavi or exhibited by a sham-treated rat at postoperative day 10, with 10C stimulation. 116

PAGE 117

CHAPTER 5 FUTURE DIRECTION In the current work, an operant method for ev aluating facial sensitivity to thermal and mechanical stimuli in rodents has been described. This method was used to evaluate the effect of TRP channel agonists on thermal perception a nd finding support a role for a TRPV1-expressing population in cold pain, as well as menthol irritation, but not in the maintenance of a putative thermogenic behavior, wet dog shaking. The loss of TRPV1-expressing population also does not affect discrimination of cold from warmth. Th is work also characterized both unlearned and operant behaviors towards mechanical and a ra nge of thermal stimuli following neuropathic injury. Alleviation of cold allodynia by pregabalin and gabape ntin was demonstrated and the influences of affect, experience, and memory on neuropathic pain behaviors were discussed. While this work contributes to and supports th e current knowledge regarding the role of TRP channels in thermal processing, normal, and pathol ogical pain, it does not re present an end point. Scientific inquiry is like a small child; the answer to a question often lead s to another series of questions. In this final chapter, future directions of this work are discussed. Adaptation of the Assay to Evaluate Pain in Mice The importance of incorporating behavioral ev aluations of pain that require a decision making process has been emphasized in this work Operant and thermal preference assessments, with the exception of TRPM8 (Dhaka et al ., 2007) and TRPV4 (Lee et al., 2005) knock-out mice, are lacking with respect to the behavioral characterization of TRP channel knock-out mice. To address this gap in the literatu re, the assays described in this work have been scaled for use in mice, making all of the assessments in the curren t study possible for various strains of wild type and genetically modified mice. The lab is curr ently in the process of characterizing the full range of thermally-mediated operant behaviors in TRPM8 and TRPV1 knock-out mice, and may 117

PAGE 118

do so with TRPA1 knock-out mice in the future. Thermal preference in various strains of knockout mice will also be assessed, particularly with respect to the discrimination between hot and cold stimuli. There is potent ial for co-expression among TRP cha nnel populations, as previously mentioned, and since these channe ls can be modulated by local voltage changes, as well as intracellular signaling cascades, interactions amo ng channels, as well as their intracellular milieu, under different stimulating conditions must be considered (Belmonte and Viana, 2008). The next step in TRP channel resear ch should be aimed at evaluating the additive effects of these channels in encoding normal a nd pathological thermal perception. Thermal preference assessments may be one way to accomplish this. We showed in the current work that although pharmacological knock -out of TRPV1 eliminates cold pain, it does not affect cold avoidance. In TRPM8 knockout mice, with profound impairment in cold sensitivity, but no heat related impairments, what pr ofile of behavior would we observe if cold is paired with moderate heat or wi th noxious heat? Colburn and co lleagues demonstrated that these mice prefer 5 over 45C and do not distinguish be tween 15 and 25C in contrast to wild type mice (Colburn et al., 2007), suggesting that the aversive qual ity of very cold pain is impaired in the absence of TRPM8. We would lik e to determine if this is true for facial stimulation as well. Likely it is, but the contribution of TRPM8 to co ld perception and pain in the face and head may be slightly different in the trigeminal nervous syst em than in the sciatic, as differences have been noted in the expression of TRPM8 in the TG versus DRGs (Kobaya shi et al., 2005). An additional method to evaluate the additive effects of multiple channels is to selectively remove them via genetic knock-out and pharmacologi cal means or RNA silencing methods. For example, treatment of TRPV1 knock-out mice w ith a TRPM8 antagonist, such as capsazepine 118

PAGE 119

(also an antagonist for TRPV1, but not an issue in this case), could also address the contribution of TRPM8 versus other cold-activated channels to cold perception and pain. We can also use knock-out mice to assess the c ontribution of different TRP channels to the development and modality specificity of neur opathic pain. TRPV1 knock-out mice exhibit differences from wild type in various models of chronic pain (Blcskei et al., 2005), but CCI was not assessed, nor was cold sensitivity. If th e abnormal expression of TRPV1 with TRPM8 or other cold sensing channels underlies cold allodynia following neuropathic injury, then cold allodynia should not be observed in TRPV1 knock-out mice with a neuropathic injury. CCI to the sciatic nerve in TRPM8 knock-out mice faile d to produce increased responsiveness to the acetone spray test, which could not be explai ned by a lack of mechanical hypersensitivity (Colburn et al., 2007). However, re sponses to direct cold stimula tion were not assessed. There is evidence for other cold activated receptors (B abes et al., 2006; Munns et al., 2007), but do any of these contribute to the cold allodynia that characterizes neuropathic pain? Thermal Preference, Conditioned Aversion, and Drug Treatment In Chapter 2, we presented findings to indica te that the thermal preference task could be used to condition an aversion or preference for one thermode, sugges ting that this assay could be used to evaluate the effect of analgesic a nd anxiolytic compounds on pain-related decision making. To address this issue, I propose a series of experiments (T able 5-1). We would need to establish how many times and with what inter-testing period condi tioned aversion/preference can be reliably reproduced in the same group of animal s. If multiple testi ng in a single group of animals is possible, then the drug treatments pr oposed in Table 5-1 could use a within-individual design, as long as sufficient recovery period was a llowed between treatments If multiple testing in a single group is not possible, then the experi ments in Table 5-1 would need to be replicated for each drug treatment applied to unique groups of animals. If this work is to be conducted with 119

PAGE 120

mice, the work of chapter 2 would need to be repeated in mice to establish similarities or differences that may exist between mice and rats. The addition of video tr acking will also allow us to measure movement and general activity in the preference box, which can be used to assess sedation or loss of motor control or bala nce that might accompany drug treatment. Stimulus Novelty and Neuropathic Injury Chapter 4 presented data suggesting that experience and novelty can effect how CCItreated rats perform on the therma l preference task. We have al so generally observed that CCItreated rats also are less successf ul than their sham or nave counterparts when exposed to a novel stimulus that is not part of their normal routine. This could reflect injury-related alterations in cortical regions associated with affective procession of painful stimuli. A controlled study is needed to confirm that CCI-tre ated rats exhibit reduced performance in the operant task with respect to novel stimuli. This could be done by training rats without collecting baseline data, surgically treating as subset of rats with either CCI or sham, then testing surgically treated and untreated rats at two weeks post surg ery with any stimulus other than the training stimulus. This process could be re peated in subsequent groups for different stimuli, or at a later time post surgery. The latter condition could help address the possibili ty that behavioral manifestations of neuropathic pain subside as a consequence of habituation to the pain evoked by testing conditions. If CC I-treated rats consistently respond p oorly to novelty, this may allow for studies evaluating the effect of electrolytic or chemical lesion of cortical regions, such as the anterior cinglulate cortex, that could be involved in affective judgments about painful stimuli. General Conclusion This work characterizes an operant method for evaluating facial pa in in experimental animals that is relevant to the human condition. In combination with histological, functional imaging, and other molecular techniques, this be havioral evaluation method can provide insights 120

PAGE 121

regarding both peripheral and centr al pain processing. These insight s will lead to more effective pain treatment, or perhaps prevention, that coul d profoundly enhance the quali ty of life for future generations. Table 5-1. Experimental schedule for evaluati ng the effect of differe nt classes of drugs on conditioned aversion in the thermal preference task. Left Thermode Right Thermode Day 1: conditioning aversion 52C 18C Day 2: measuring aversion 18C 18C (OR) Day 1: conditioning aversion 45C -4C* Day 2: measuring aversion 45C 45C Where indicates the experimenter cont rolled starting side the avoided side. On Day 2 drug treatment would be applied prior to testing. Treatment Classification Potential Drug(s) Expected Outcome Untreated N/A Aversion to side Vehicle (injection control) PBS No change in aversion Anxiogenic Yohimbe Enhanced aversion Anxiolytic/non-analgesic Valium, Diazepam Aversion abolished, success no different Anxiolytic/analgesic Morphine, pregabalin, Cymbalta Aversion abolished, success increased Analgesic only Lidocaine, NSAID, RTX Aversion intact, success increased At end, confirm 50/50 preference for Day 2 state in the absence of conditioning or treatment. NSAID = non-steroi dal anti-inflammatory drug 121

PAGE 122

APPENDIX DOSE DETERMINATION OF PREGABALIN USED FOR TREATMENT OF CHRONIC CONSTRICTION INJURY The literature typically reports an effective analgesic dose for pregabalin between 30-100 mg/kg in animals (Field et al., 1999; Hong-Ju et al., 2004; Beyreuther et al., 2006; Ling et al., 2008). However, these assessments are made using reflex based assays. Pregabalin also has the potential for anxiolytic properties that could contribute to analgesi a (Field et al., 2001; Nicolas et al., 2007; Nutt et al., 2008; Zohar et al., 2008). We sought to determine if using an operant method of pain evaluation revealed a similar or different effectiv e analgesic dose, which we used in later studies (see Chapter 4) to treat animals following neuropathic injury. Evaluation of General Activity by Rearing In order to evaluate changes in general ac tivity induced by pregabalin, rearing (vertical locomotion) data was collected as previously de scribed (Neubert et al., 2007; Rossi and Neubert, 2008). Briefly, rats were placed in a cylindrical chamber and the number and duration of rears were recorded for a 15-minute period. This info rmation was used to ca lculate the duration per rearing event reported here. Rearing data was recorded for four consecutive days to allow rats to become accustomed to the chamber prior to pregabalin administration. Administration and Dose Determination of Pregabalin In order to determine which dose to use fo llowing neuropathic inju ry, rats (n = 10, not part of the surgically treated animals described in Chapter 4) were injected intraperitoneally (i.p.) with different doses of pregabalin (PG) or PB S vehicle and tested in the rearing chamber 30 minutes later. After rearing assessment, capsaicin cream (0.035%, Capzasin P; Chattem, INC; Chattanooga, TN) was applied to the faces of gently restrained rats, left on for 5 minutes, and removed with water. Immediately after capsaici n removal, rats were placed in the operant thermal chambers and their responses to a 45C stim ulus were recorded. The doses of pregabalin 122

PAGE 123

used were: 1, 5, 10, and 100 mg/kg (n = 5 rats per dose). Because these experiments were performed on multiple groups of rats on different da ys, data was converted to a percentage of the rats baseline values and subseque nt statistical analysis was perf ormed on the percentage values. Pregabalin Dose Determination Based on Re aring and Alleviation of Capsaicin-Induced Heat Hyperalgesia We also wished to examine how neuropathic behavior revealed by the operant task could be altered by commonly used analgesic treatments such as pregabalin and gabapentin. Before administering pregabalin in neur opathic animals, we assed the effects of several doses of pregabalin on general activity and heat hyperalgesia in nave rats to determine the lowest effective dose. The rearing assay was used to a ssess the effects of pregabalin on general activity. The highest dose examined, 100 mg/kg (i.p.), had significant effects on r earing behavior (Fig. 3A). At this dose, rats were hyper active in the re aring chamber, with ataxia and difficulty balancing on their hindpaws. This resulted in many short rearing events as compared to baseline. None of the other doses tested had significant effects on rearing behavior. Following assessment of general behavior, we evaluated the ability of these doses to alleviate capsaicin-induced heat hyperalgesia. We found that th e two highest doses tested, 10 and 100 mg/kg, significantly increased licks and the success ratio as compared with PBS and lower doses of pregabalin, but ha d no significant effect on stimulu s contacts (Fig. 3 B-D). The number of licks and the success ratio produced in the presence of 10 and 100mg/kg with capsaicin was not significantly different from those outcomes produced by 45C alone. In other words, 10 and 100mg/kg were able to strongly alleviate capsaicin-m ediated hyperalgesia. However, rats that received 100mg/kg pregabalin were initially ataxic, then became increasingly lethargic as testing progressed. In contrast, rats that received 10 mg/kg did not show signs of 123

PAGE 124

sedation. Based on these findings, we chose 10mg/kg of pregabalin for subsequent treatment of CCIand sham-operated rats. 124

PAGE 125

Figure A-1. Determination of lowe st analgesic dose of pregaba lin without significant effects on general activity, measured by rearing behavior Rats (n = 5 each dose) were injected with pregabalin or PBS i.p. and their rearing behavior was recorded 30 minutes later. A. The duration per rearing event (s/rear) was calculated for each dose and is expressed as a percentage of the baseline value to normalize across multiple testing sessions. B-D. Immediately following rearing, rats were tested at 45.5C with topical capsaicin (0.035%). The percentage cha nge from baseline licks (B), stimulus contacts (C), and the success ratio (lic ks/contact, D) are shown for each dose. Baseline for stimulus testing refers to beha vior recorded in the presence of the 45.5C stimulus without capsaicin treatment. Da ta are mean SEM. Main effect of pregabalin treatment for each outcome, as determined by ANOVA: (A) F5,34 = 6.755, (B) F5,43 = 14.283, (C) F5,43 = 0.665, (D) F5,43 = 15.215. ** indicates p<0.05, as determined by post-hoc Tukeys test. 125

PAGE 126

LIST OF REFERENCES Albin KC, Carstens MI, Carstens E. Modulation of oral heat and co ld pain by irritant chemicals. Chemical Senses 2008;33:3-15. Allchorne AJ, Broom DC, J. WC. De tection of cold pain, cold all odynia and cold hy peralgesia in freely behaving rats. Mol Pain 2005; 1:14 Dec. 2005 . Babes A, Amuzescu B, Krause U, Scholz A, Flon ta M-L, Reid G. Cooling inhibits capsaicininduced currents in cultured rat dorsal root ganglion neurones. Neurosci Lett 2002;317:131-134. Babes A, Zorzon D, Reid G. A novel type of cold-sensitive neuron in rat dor sal root ganglia with rapid adaptation to cooling stimu li. Euro J Neurosci 2006;24:691-698. Bandell M, Story GM, Hwang SW, Viswanath V, Ei d SR, Petrus MJ, Earley TJ, Patapoutian A. Noxious cold ion channel TRPA1 is activ ated by pungent compounds and bradykinin. Neuron 2004;41:849-857. Bautista DM, Jordt S-E, Nikai T, Tsuruda PR, Read AJ, Poblete J, Yamoah EN, Basbaum AI, Julius D. TRPA1 mediates the inflammatory actions of environmental irritants and proalgesic agents. Cell 2006;124:1269-1282. Becker S, Kleinbhl D, Klossika I, Hlzl R. Op erant conditioning of enhanced pain sensitivity by heat-pain titration. Pain 2008;4 Sep. 2008 [In Press] Behrendt HJ, Germann T, Gillen C, Hatt H, Jo stock R. Characterization of the mouse coldmenthol receptor TRPM8 and vanilloid rece ptor type-1 VR1 using a fluorometric imaging plate reader (FLIPR) assay. Br J Pharmacol 2004;141:737-745. Belmonte C, Viana F. Molecular and cellular lim its to somatosensory specificity. Mol Pain 2008;4:18 April 2008 . Bennett GJ, Xie YK. A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 1988;33:87-107. Benoist JM, Gautron M, Guilbaud G. Experimental model of tr igeminal pain in the rat by constriction of one infraorbital nerve: change s in neuronal activities in the somatosensory cortices corresponding to the infraorb ital nerve. Exper Brain Res 1999;126:383-398. Benoliel R, Eliav E, Iadarola MJ. Neuropeptid e Y in trigeminal ganglion following chronic constriction injury of the rat infraorbital ne rve: is there correlation to somatosensory parameters? Pain 2001;91:111-121. Beyreuther B, Callizot N, Sthr T. Antinocicep tive efficacy of lacosamide in a rat model for painful diabetic neuropathy. Euro J Pharmacol 2006;539:64-70. 126

PAGE 127

Beyreuther BK, Callizot N, Brot MD, Feldman R, Bain SC, Sthr T. Antinociceptive efficacy of lacosamide in rat models for tumorand chemotherapy-induced cancer pain. Euro J Pharmacol 2007;565:98-104. Biggs JE, Yates JM, Loescher AR, Clayton NM, Boissonade FM, Robinson PP. Changes in vanilloid receptor 1 (TRPV1) expression following lingual nerve injury. Euro J Pain 2007;11:192-201. Blcskei K, Helyes Z, Szab Sndor K, Elekes K, Nmeth J, Almsi R, Pintr E, Petho G, Szolcsnyi J. Investigation of the role of TRPV1 receptors in acute and chronic nociceptive processes us ing gene-deficient mice. Pain 2005;117:368-376. Boroujerdi A, Kim HK, Lyu YS, Kim D-S, Figu eroa KW, Chung JM, Luo ZD. Injury discharges regulate calcium channel alpha-2-delta-1 s ubunit upregulation in th e dorsal horn that contributes to initiation of neuropathic pain. Pain 2008;6 Oct. 2008 [In Press]. Bowsher D. Dynamic mechanical allodynia in neuropathic pai n. Pain 2005;116:164-165. Cahusac PMB, Noyce R. A pharmacological study of slowly adapting mechanoreceptors responsive to cold thermal stim ulation. Neuroscience 2007;148:489-500. Caterina MJ, Leffler A, Malmberg AB, Martin WJ, Trafton J, Petersen-Zeitz KR, Koltzenburg M, Basbaum AI, Julius D. Impaired nociception and pain sensation in mice lacking the capsaicin receptor. Science 2000;288:306-313. Caterina MJ, Rosen TA, Tominaga M, Brake AJ Julius D. A capsaicin-receptor homologue with a high threshold for noxious heat. Nature 1999;398:436-441. Caterina MJ, Schumacher MA, Tominaga M, Ro sen TA, Levine JD, Julius D. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 1997;389:816-824. Chichorro JG, Zampronio AR, Petto Souza GE, Rae GA. Orofacial cold hyperalgesia due to infraorbital nerve constriction injury in rats : Reversal by endothelin receptor antagonists but not non-steroidal anti-infla mmatory drugs. Pain 2006;123:64-74. Choi DS, Choi DY, Whittington RA, Nedeljkovic SS. Sudden amnesia resulting in pain relief: The relationship between memory and pain. Pain 2007;132:206-210. Christoph T, Gillen C, Mika J, Grnweller A, Sc hfer MKH, Schiene K, Frank R, Jostock R, Bahrenberg G, Weihe E, Erdmann VA, Kurreck J. Antinociceptive effect of antisense oligonucleotides against the vanilloid receptor VR1/TRPV1. Neurochem Int 2007;50:281-290. Christoph T, Grnweller A, Mika J, Schfer M KH, Wade EJ, Weihe E, Erdmann VA, Frank R, Gillen C, Kurreck J. Silencing of vanilloid receptor TRPV1 by RNAi reduces neuropathic and visceral pain in vivo. Bioc hem Biophys Res Commun 2006;350:238-243. 127

PAGE 128

Chuang H-h, Neuhausser WM, Julius D. The super-c ooling agent icilin reveals a mechanism of coincidence detection by a temperature-se nsitive TRP channel. Neuron 2004;43:859-869. Clapham DE. TRP is cracked but is CRAC TRP? Neuron 1996;16:1069-1072. Cliff MA, Green BG. Sensitization and desensiti zation to capsaicin and menthol in the oral cavity: Interactions and individual differences. P hysiol Behav 1996;59:487-494. Colburn RW, Lubin ML, Stone JDJ, Wang Y, La wrence D, D'Andrea Michael R, Brandt MR, Liu Y, Flores CM, Qin N. Attenuated cold sensitivity in TRPM8 null mice. Neuron 2007;54:379-386. Colpaert FC, Deseure K, Stinus L, Adriaense n H. High-efficacy 5-Hydroxytryptamine 1A receptor activation counteracts opioid hyperallodynia and affective conditioning. J Pharmacol Exp Ther 2006;316:892-899. Colton C, Zhu M. 2-Aminoethoxydiphenyl borate as a common activat or of TRPV1, TRPV2, and TRPV3 channels. Handb Exp Pharmacol 2007;179:173-187. Crabbe JC, Wahlsten D, Dudek BC. Genetics of mouse behavior: interac tions with laboratory environment. Science 1999;284:1670-1672. Craig AD. PAIN MECHANISMS: labeled lines versus convergence in central processing. Ann Rev Neurosci 2003;26:1-30. Davis JB, Gray J, Gunthorpe MJ, Hatcher JP, Davey PT, Overend P, Harries MH, Latcham J, Clapham C, Atkinson K, Hughes SA, Rance K, Grau E, Harper AJ, Pugh PL, Rogers DC, Bingham S, Randall A, Sheardown SA. Vanilloid receptor-1 is essential for inflammatory thermal hyperalgesia. Nature 2000;405:183-187. de Leeuw R, Albuquerque R, Okeson J, Carlson C. The contribution of neuroimaging techniques to the understanding of supraspinal pain circui ts: Implications for orofacial pain. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;100:308-314. Deseure K, Adriaensen H. Comparison between two types of behavioral variables of non-evoked facial pain after chronic constriction injury to the rat infraorbital nerve. Comp Med 2002;52:44-49. Deseure K, Adriaensen H. Nonevoked facial pain in rats following infraorbital nerve injury: a parametric analysis. Physiol Behav 2004;81:595-604. Dhaka A, Earley TJ, Watson J, Patapoutian A. Visualizing cold spots: TRPM8-expressing sensory neurons and their proj ections. J Neurosci 2008;28:566-575. Dhaka A, Murray AN, Mathur J, Earley TJ, Petrus MJ, Patapoutian A. TRPM8 Is required for cold sensation in mice. Neuron 2007;54:371-378. 128

PAGE 129

Doerner JF, Gisselmann G, Hatt H, Wetzel CH. Transient receptor potential channel A1 is directly gated by calcium ions J Biol Chem 2007;282:13180-13189. Eckert WA, III, Julius D, Basbaum AI. Differentia l contribution of TRPV1 to thermal responses and tissue injury-induced sensit ization of dorsal horn neurons in laminae I and V in the mouse. Pain 2006;126:184-197. Eisenach JC, Lindner MD. Did experimeter bias conceal the efficacy of spinal opioids in previous studies with the spinal nerv e ligation model of neuropathic pain? Anesthesiology 2004;100:905-911. Evans MS, Reid KH, Sharp JBJ. Dimethylsulf oxide (DMSO) blocks conduction in peripheral nerve C fibers: a possible mechanisms of analgesia. Neurosci Lett 1993;150:145-148. Evenden J, Duncan B, Ko T. A comparison of th e effects of psychomimetics and anxiolytics on punished and unpunished responding maintained by fixed interval schedules of food reinforcement in the rat. Behav Parmacol 2006;17:87-99. Field MJ, Bramwell S, Hughes J, Singh L. Detection of static and dynamic components of mechanical allodynia in rat models of neur opathic pain: are they signalled by distinct primary sensory neuron es? Pain 1999;83:303-311. Field MJ, Cox PJ, Stott E, Melrose H, Offord J, Su T-Z, Bramwell S, Corradini L, England S, Winks J, Kinloch RA, Hendrich J, Dolphin AC, Webb T, Williams D. Identification of the {alpha}2-{delta}-1 subunit of voltage-dep endent calcium channels as a molecular target for pain mediating the analgesi c actions of pregabalin. PNAS 2006;103:1753717542. Field MJ, Ryszard JO, Singh L. Pregabalin may repr esent a novel class of a nxiolytic agents with a broad spectrum of activity. Br J Pharmacol 2001;132:1-4. Frederick J, Buck ME, Matson DJ, Cortri ght DN. Increased TRPA1, TRPM8, and TRPV2 expression in dorsal root ganglia by ne rve injury. Biochem Biophys Res Commun 2007;358:1058-1064. Gameiro GH, da Silva Andrade A, de Castro M, Pereira LF, Tambeli CH, Ferraz de Arruda Veiga MC. The effects of restraint stress on nociceptive responses induced by formalin injected in rat's TMJ. Pharmacol Biochem Behav 2005;82:338-344. Gameiro GH, Gameiro PH, da Silva Andrade A, Pereira LF, Arthuri MT, Marcondes FK, de Arruda Veiga MCF. Nociceptionand anxiety-li ke behavior in rats submitted to different periods of restraint stress Physiol Behav 2006;87:643-649. Garcia-Anoveros J, Nagata K. TR PA1. Handb Exp Pharmacol 2007;179:347-362. 129

PAGE 130

Gee NS, Brown JP, Dissanayake VUK, Offord J, Thurlow R, Woodruff GN. The novel anticonvulsant drug, gabapentin (Neurontin), binds to the alpha(2) delta subunit of a calcium channel. J Biol Chem 1996;271:5768-5776. Green BG, McAuliffe BL. Menthol desensitization of capsaicin irritation: Evidence of a shortterm anti-nociceptive effect Physiol Behav 2000;68:631-639. Green BG, Pope JV. Innocuous cooling can produ ce nociceptive sensations that are inihibited during dynamic mechanical contac t. Exp Brain Res 2003;148:290-299. Green BG, Schoen KL. Evidence that tactile stimulation inhibits nociceptive sensations produced by innocuous contact cooling. Behav Brain Res 2005;162:90-98. Green BG, Schoen KL. Thermal and nociceptive se nsations from menthol and their suppression by dynamic contact. Behav Brain Res 2007;176:284-291. Guler AD, Lee H, Iida T, Shimizu I, Tominaga M, Caterina M. Heat-Evoked Activation of the Ion Channel, TRPV4. J Neurosci 2002;22:6408-6414. Hargreaves KM, Dubner R, Brown F, Flores CM, Joris J. A new and sensitive method for measuring thermal nociception in cuta neous hyperalgesia. Pain 1988;32:77-88. Hlzl R, Kleinbhl D, Huse E. Implicit operant learning of pain sensit ization. Pain 2005;115:1220. Hong-Ju Y, He L, Wei-Guo S, Nan Z, Wei-Xiu Y, Zhong-Wei J, Jun-Wei W, Zheng-Hua G, BoHua Z, Zhi-Pu L, Zhe-Hui G. Effect of gaba pentin derivates on m echanical allodynia-like behaviour in a rat model of chronic sciati c constriction injury. Bioorg Med Chem Lett 2004;14:2537-2541. Imamura Y, Kawamoto H, Nakanishi O. Char acterization of heathyperalgesia in an experimental trigeminal neuropathy in rats. Exp Brain Res 1997;116:97-103. Institute of Laboratory Animal Resources US. Guide for the care and use of laboratory animals. Washington, D.C.: National Academy Press, 1996. Jabakhanji R, Foss JM, Berra HH, Centeno MV, Apkarian AV, Chialvo DR. Inflammatory and neuropathic pain animals exhi bit distinct responses to i nnocuous thermal and motoric challenges. Mol Pain 2006;2:5 Jan 2006 . Jensen AA, Mosbacher J, Elg S, Lingenhoehl K, Lohmann T, Johansen TN, Abrahamsen B, Mattsson JP, Lehmann A, Bettler B, Brauner-O sborne H. The anticonvulsant gabapentin (Neurontin) does not act through gamma -ami nobutyric acid-B receptors. Mol Pharmacol 2002;61:1377-1384. 130

PAGE 131

Jhaveri MD, Elmes SJR, Kendall DA, Chapman V. Inhibition of peripheral vanilloid TRPV1 receptors reduces noxious heat-evoked responses of dorsal horn neurons in nave, carrageenan-inflamed and neuropathic rats. Euro J Neurosci 2005;22:361-370. Jordt S-E, Bautista DM, Chuang H-h, McKemy DD, Zygmunt PM, Hogestatt ED, Meng ID, Julius D. Mustard oils and cannabinoids ex cite sensory nerve fi bres through the TRP channel ANKTM1. Nature 2004;427:260-265. Julius D, Basbaum AI. Molecular mechanis ms of nociception. Nature 2001;413:203-210. Karashima Y, Damann N, Prenen J, Talavera K, Segal A, Voets T, Nilius B. Bimodal Action of Menthol on the Transient Receptor Potentia l Channel TRPA1. J Neurosci 2007;27:98749884. Katsura H, Obata K, Mizushima T, Yamanaka H, Kobayashi K, Dai Y, Fukuoka T, Tokunaga A, Sakagami M, Noguchi K. Antisense knock down of TRPA1, but not TRPM8, alleviates cold hyperalgesia after spinal nerve liga tion in rats. Exper Neurol 2006;200:112-123. Kayser V, Aubel B, Hamon M, Bourgoin S. The antimigraine 5HT 1B/1D receptor agonists, sumatriptan, zolmitriptan and dihydroergotamin e, attenuate pain-rel ated behaviour in a rat model of trigeminal pain. Br J Pharmacol 2002;137:1287-1297. Kim JH, Greenspan JD, Coghill RC, Ohara S, Lenz FA. Lesions limited to the human thalamic principal somatosensory nucleus (ventral ca udal) are associated with loss of cold sensations and central pa in. J Neurosci 2007;27:4995-5004. Kim SM, Kim J, Kim E, Hwang SJ, Shin HK, Lee SE Local application of capsaicin alleviates mechanical hyperalgesia afte r spinal nerve transection. Neurosci Lett 2008;433:199-204. Kindt KS, Viswanath V, Macpherson L, Qu ast K, Hu H, Patapoutian A, Schafer WR. Caenorhabditis elegans TRPA-1 functions in mechanosensation. Nat Neurosci 2007;10:568-577. King CD, Devine DP, Vierck CJ, Mauderli A, Yezi erski RP. Opioid modulation of reflex versus operant responses following stress in the rat. Neuroscience 2007;147:174-182. King CD, Devine DP, Vierck CJ, Rodgers J, Ye zierski RP. Differential effects of stress on escape and reflex responses to nociceptive thermal stimuli in the rat. Brain Res 2003;987:214-222. Kobayashi K, Fukuoka T, Obata K, Yamanaka H, Dai Y, Tokunaga A, Noguchi K. Distinct expression of TRPM8, TRPA1, and TRPV1 mRNAs in rat primar y afferent neurons with adelta/c-fibers and colocalization with tr k receptors. J Comp Neurol 2005;493:596-606. 131

PAGE 132

Kosugi M, Nakatsuka T, Fujita T, Kuroda Y, Kumamoto E. Activation of TRPA1 channel facilitates excitatory synaptic transmission in substantia gelatinosa neurons of the adult rat spinal cord. J neurosci 2007;27:4443-4451. Kwan KY, Allchorne AJ, Vollrath MA, Christen sen AP, Zhang D-S, Woolf CJ, Corey DP. TRPA1 contributes to cold, mechanical, and ch emical nociception but is not essential for hair-cell transduction. Neuron 2006;50:277-289. Lang PM, Schober GM, Rolke R, Wagner S, Hilg e R, Offenbcher M, Treede R-D, Hoffmann U, Irnich D. Sensory neuropat hy and signs of central sens itization in patients with peripheral arterial dis ease. Pain 2006;124:190-200. Lee H, Iida T, Mizuno A, Suzuki M, Caterina MJ. Altered therma l selection behavior in mice lacking transient receptor potential vanilloid 4. J Neurosci 2005;25:1304-1310. Leffler A, Linte RM, Nau C, Reeh P, Babes A. A high-threshold heat-activated channel in cultured rat dorsal root ganglion neurons resembles TRPV2 and is blocked by gadolinium. Euro J Neurosci 2007;26:12-22. Levine JD, Alessandri-Haber N. TRP channels: Targets for the relief of pain. Biochim Biophys Acta 2007;1772:989-1003. Li C-Y, Zhang X-L, Matthews EA, Li K-W, Kurwa A, Boroujerdi A, Gross J, Gold MS, Dickenson AH, Feng G, Luo ZD. Calcium ch annel [alpha]2[delta]1 subunit mediates spinal hyperexcitability in pa in modulation. Pain 2006;125:20-34. Liang Y-C, Huang C-C, Hsu K-S. The synt hetic cannabinoids a ttenuate allodynia and hyperalgesia in a rat mode l of trigeminal neuropat hic pain. Neuropharmacology 2007;53:169-177. Lim EJ, Jeon HJ, Yang GY, Lee MK, Ju JS, Han SR, Ahn DK. Intracisternal administration of mitogen-activated protein kinase inhibito rs reduced mechanical allodynia following chronic constriction injury of infraorbital nerve in ra ts. Prog Neuro-Psychopharmacol Biol Psychiatry 2007;31:1322-1329. Ling B, Authier N, Balayssac D, Eschalier A, Coudore F. Behavioral and pharmacological description of oxaliplatin-induced painfu l neuropathy in rat. Pain 2007;128:225-234. Ling B, Coudor F, Decalonne L, Eschalier A, Au thier N. Comparative antiallodynic activity of morphine, pregabalin and lidocaine in a rat model of neuropathic pain produced by one oxaliplatin injection. Ne uropharmacology 2008;55:724-728. Liu XG, Morton CR, Azkue JJ, Zimmermann M, Sa ndkhler J. Long-term depression of C-fibreevoked spinal field potentials by stimulation of primary afferent A delta-fibres in the adult rat. Euro J Neurosci 1998;10:3069-3075. 132

PAGE 133

Ma W, Zhang Y, Bantel C, Ei senach JC. Medium and large in jured dorsal root ganglion cells increase TRPV-1, accompanied by increased [alpha]2C-adrenoceptor co-expression and functional inhibition by clonidine. Pain 2005;113:386-394. Macpherson LJ, Hwang SW, Miyamoto T, Dubin AE, Patapoutian A, Story GM. More than cool: Promiscuous relationships of menthol and other sensory compounds. Mol Cell Neurosci 2006;32:335-343. Madrid R, Donovan-Rodriguez T, Meseguer V, Ac osta MC, Belmonte C, Viana F. Contribution of TRPM8 Channels to Cold Transduction in Primary Sensory Neurons and Peripheral Nerve Terminals. J Neurosci 2006;26:12512-12525. Mauderli AP, Acosta-Rua A, Vierck CJ. An op erant assay of thermal pain in conscious, unrestrained rats. J Neur osci Methods 2000;97:19-29. Mauderli AP, Vierck CJ, Jr., Cannon RL, Rodrigue s A, Shen C. Relationships between skin temperature and temporal summation of h eat and cold pain. J Neurophysiol 2003;90:100109. McKemy DD. How cold is it? TRPM8 and TRPA1 in the molecular logic of cold sensation. Mol Pain 2005;1:22 April 2005 . McKemy DD, Neuhausser WM, Julius D. Identificati on of a cold receptor re veals a general role for TRP channels in thermosensation. Nature 2002;416:52-58. Mogil JS, Ritchie J, Sotocinal SG, Smith SB, Croteau S, Levitin DJ, Naumova AK. Screening for pain phenotypes: Analysis of three congenic mouse strains on a battery of nine nociceptive assays. Pain 2006;126:24-34. Moqrich A, Hwang SW, Earley TJ, Petrus MJ, Murray AN, Spencer KSR, Andahazy M, Story GM, Patapoutian A. Impaired thermosensat ion in mice lacking TRPV3, a heat and camphor sensor in the sk in. Science 2005;307:1468-1472. Morin C, Bushnell MC. Temporal and qualitative properties of cold pain and heat pain: a psychophysical study. Pain 1998;74:67-73. Munns C, AlQatari M, Koltzenbur g M. Many cold sensitive peripheral neurons of the mouse do not express TRPM8 or TRPA1. Cell Calcium 2007;41:331-342. Nagata K, Duggan A, Kumar G, Garcia-Anoveros J. Nociceptor and ha ir cell transducer properties of TRPA1, a channel for pain and hearing. J Neurosci 2005;25:4052-4061. Neubert JK, Mannes AJ, Karai LJ, Jenkins AC, Zawatski L, Abu-Asab M, Iadarola MJ. Perineural resiniferatoxin se lectively inhibits inflamma tory hyperalgesia. Mol Pain 2008;4:doi: 10.1186/1744-8069-1184-1183. 133

PAGE 134

Neubert JK, Rossi HL, Pogar J, Jenkins AC, Ca udle RM. Effects of muand kappa-2 opioid receptor agonists on pain and rearing beha viors. Behav Brain Funct 2007;3:20 Sept 2007 . Neubert JK, Widmer CG, Malphurs W, Rossi HL, Vierck JCJ, Caudle RM. Use of a novel thermal operant behavioral assay for characterization of orofacial pain sensitivity. Pain 2005;116:386-395. Nicolas LB, Klein S, Prinssen EP. Defensivelike behaviors induced by ultrasound: further pharmacological characterization in Lister-hooded rats. Psychopharmacology 2007;194:243-252. Nutt D, Mandel F, Baldinetti F. Early onset anxi olytic efficacy after a single dose of pregabalin: double-blind, placeboand active-comparator co ntrolled evaluation using a dental anxiety model. J Psychopharmacol 2008;17 July 2008 [In Press]. Obata K, Katsura H, Mizushima T, Yamanaka H, Kobayashi K, Dai Y, Fukuoka T, Tokunaga A, Tominaga M, Noguchi K. TRPA1 induced in sensory neurons contributes to cold hyperalgesia after inflammation and nerv e injury. J Clin Invest 2005;115:2393-2401. Okazawa M, Inoue W, Hori A, Hosokawa H, Matsumura K, Kobayashi S. Noxious heat receptors present in cold-sensory ce lls in rats. Neurosci Lett 2004;359:33-36. Peier AM, Moqrich A, Hergarden AC, Reeve AJ, Andersson DA, Story GM, Earley TJ, Dragoni I, McIntyre P, Bevan S, Patapoutian A. A TRP Channel that Senses Cold Stimuli and Menthol. Cell 2002;108:705-715. Pogatzki-Zahn EM, Shimizu I, Caterina M, Raja SN. Heat hyperalgesia after incision requires TRPV1 and is distinct from pure in flammatory pain. Pain 2005;115:296-307. Pomonis JD, Harrison JE, Mark L, Bristol DR, Valenzano KJ, Walker K. N-(4Tertiarybutylphenyl)-4-(3-cholorphyridin-2-yl)tetrahydropyrazi ne -1(2H)-carbox-amide (BCTC), a Novel, Orally Effective Vanilloid Receptor 1 Antagonist with Analgesic Properties: II. In Vivo Characterization in Rat Models of Inflammatory and Neuropathic Pain. J Pharmacol Exp Ther 2003;306:387-393. Proudfoot CJ, Garry EM, Cottrell DF, Rosie R, Anderson H, Robertson DC, Fleetwood-Walker SM, Mitchell R. Analgesia mediated by the TR PM8 cold receptor in chronic neuropathic pain. Curr Biol 2006;16:1591-1605. Qin N, Neeper MP, Liu Y, Hutchinson TL, L ubin ML, Flores CM. TRPV2 is activated by cannabidiol and mediates CGRP release in cultured rat dorsal root ganglion neurons. J Neurosci 2008;28:6231-6238. 134

PAGE 135

Rainville P, Feine J, Bushnell M, Duncan G. A psychophysical comparison of sensory and affective responses to four modalities of experimental pain. Somatosens Mot Res 1992;9:256-277. Roeska K, Doods H, Arndt K, Treede R-D, C eci A. Anxiety-like behaviour in rats with mononeuropathy is reduced by the analgesic drugs morphine and gabapentin. Pain 2008;17 June 2008 [In Press]. Rolls ET, Grabenhorst F, Parris BA. Warm pl easant feelings in the brain. NeuroImage 2008;41:1504-1513. Ross RA. Anandamide and vanilloid TRPV 1 receptors. Br J Pharmacol 2003;140:790-801. Rossi HL, Neubert JK. Effects of environmental enrichment on thermal sensitivity in an operant orofacial pain assay. Be hav Brain Res 2008;187:478-482. Rossi HL, Vierck CJ, Caudle RM, JK N. Characterization of cold sensitivity and thermal preference using an opernat orofacial assay. Mol Pain 2006;2:13 Dec 2006 . Saad NE, Al Amin H, Abdel Baki S, Safieh -Garabedian B, Atweh SF, Jabbur SJ. Transient attenuation of neuropathic manifestations in rats following lesion or reversible block of the lateral thalamic somatosensory nuclei. Exper Neur ol 2006;197:157-166. Samuelsson M, Leffler A-S, Johansson B, Hansson P. On the repeatability of brush-evoked allodynia using a novel semi-quantitative method in patients with peripheral neuropathic pain. Pain 2007;130:40-46. Sawada Y, Hosokawa H, Hori A, Matsumura K, Kobayashi S. Cold sensitivity of recombinant TRPA1 channels. Brain Res 2007;1160:39-46. Seifert F, Maihofner C. Central mechanisms of experimental and chronic neuropathic pain: Findings form functional imaging studies. Cell Mol Life Sci 2008;15 Sept 2008 [In Press]. Selak I. Pregablin (Pfizer). Cu rr Opin Investig Drugs 2001;2:828-834. Simone DA, Kajander KC. Excitation of rat cu taneous nociceptors by noxious cold. Neurosci Lett 1996;213:53-56. Steiner AA, Turek VF, Almeida MC, Burmeister JJ, Oliveira DL, Roberts JL, Bannon AW, Norman MH, Louis J-C, Treanor JJS, Gavva NR, Romanovsky AA. Nonthermal activation of transient recep tor potential vanillo id-1 channels in abdominal vscera tonically inhibits autonomic cold-defen se effectors. J Neurosci 2007;27:7459-7468. 135

PAGE 136

Story GM, Peier AM, Reeve AJ, Eid SR, Mosbacher J, Hricik TR, Earley TJ, Hergarden AC, Andersson DA, Hwang SW, McIntyre P, Jegl a T, Bevan S, Patapoutian A. ANKTM1, a TRP-like channel expressed in nociceptive neur ons, is activated by cold temperatures. Cell 2003;112:819-829. Strotmann R, Schultz G, Plant TD. Ca2+-dep endent potentiation of the nonselective cation channel TRPV4 is mediated by a C-terminal calmodulin binding site. J Biol Chem 2003;278:26541-26549. Sufka K. Conditioned place preference paradi gm: a novel approach for analgesic drug assessment against chroni c pain. Pain 1994;58:355-366. Takashima Y, Daniels RL, Knowlton W, Teng J, Liman ER, McKemy DD. Diversity in the neural circuitry of cold sens ing revealed by genetic axonal labeling of tran sient receptor potential melastatin 8 neur ons. J Neurosci 2007;27:14147-14157. Tanimoto-Mori S, Nakazato-Imasato E, Toide K, Kita Y. Pharmacological investigation of the mechanisms underlying cold allodynia using a new cold plate procedure in rats with chronic constriction injuri es. Behav Parmacol 2008;19:85-90. Tran TD, Matre D, Casey KL. An inhibitory inte raction of human cortical responses to stimuli preferentially exciting A[delta] or C fibers. Neuroscience 2008;152:798-808. Tzabazis A, Klyukinov M, Manering N, Nemen ov MI, Shafer SL, Yeomans DC. Differential activation of trigeminal C or A[delta] nociceptors by infrared di ode laser in rats: behavioral evidence. Brain Res 2005;1037:148-156. Vaccarino AL, Plamondon H, Melzack R. Analgesic and aversive effects of naloxone in BALB/c mice. Exper Neurol 1992;117:216-218. van der Kam EL, Vry JD, Schiene K, Tzschentke TM. Differential effects of morphine on the affective and the sensory component of carra geenan-induced nociception in the rat. Pain 2008;136:373-379. Vierck CJ, Acosta-Rua AJ, Johnson RD. Bilateral chronic constriction of the sciatic nerve: A model of long-term cold hyperalgesia. J Pain 2005;6:507-517. Vierck CJ, Acosta-Rua AJ, Rossi HL, Neubert JK. Sex differences in thermal pain sensitivity and sympathetic reactivity for two strains of rat. J Pain 2008;9:739-749. Vierck CJ, Kline R, Wiley RG. Comparison of ope rant escape and innate reflex responses to nociceptive skin temperatures produced by h eat and cold ttimulation of rats. Behav Neurosci 2004;118:627-635. 136

PAGE 137

Vos BP, Strassman AM, Maciewicz RJ. Behavioral evidence of trigeminal neuropathic pain following chronic constriction injury to th e rat's infraorbital nerve. J Neurosci 1994;14:2708-2723. Walczak J-S, Beaulieu P. Comparison of three models of neuropathic pain in mice using a new method to assess cold allodynia: The double plate technique. Ne urosc Lett 2006;399:240244. Walker KM, Urban L, Medhurst SJ, Patel S, Panesar M, Fox AJ, McIntyre P. The VR1 Antagonist Capsazepine Reverses Mechanical H yperalgesia in Models of Inflammatory and Neuropathic Pain. J Pharmacol Exp Ther 2003;304:56-62. Wasner G, Schattschneider J, Binder A, Baron R. Topical menthol a human model for cold pain by activation and sensitization of C nociceptors. Brain 2004;127:1159-1171. Wexel MM. Central targeting of trigeminal primar y afferent nerve terminals via intracisternal injection of RTX and its effect on pain behavi or. Master of Science Thesis. University of Florida, Gainesville, 2008. Wheeler G. Gabapentin. Pfizer. Curr Opin Investig Drugs 2002;3:470-477. Wilson-Gerwing TD, Dmyterko MV, Zochodne DW, Johnston JM, Verge VMK. Neurotrophin-3 suppresses thermal hyperalgesia associated with neuropathic pain and attenuates transient receptor potential vanilloid receptor-1 expressi on in adult sensory neurons. J Neurosci 2005;25:758-767. Wilson SP, Yeomans DC, Bender MA, Lu Y, Goin s WF, Glorioso JC. An tihyperalgesic effects of infection with a preproenkephalin-en coding herpes virus. PNAS 1999;96:3211-3216. Woodbury CJ, Zwick M, Wang S, Lawson JJ, Ca terina MJ, Koltzenburg M, Albers KM, Koerber HR, Davis BM. Nociceptors lack ing TRPV1 and TRPV2 have normal heat responses. J Neurosci 2004;24:6410-6415. Woolf CJ. Long term alterations in the excitabili ty of the flexion reflex produced by peripheral tissue injury in the chronic decerebrate ra t. Pain 1984;18:325-343. Xing H, Chen M, Ling J, Tan W, Gu JG. TRPM 8 mechanism of cold a llodynia after chronic nerve injury. J Neurosci 2007;27:13680-13690. Xing H, Ling J, Chen M, Gu JG. Chemical and co ld sensitivity of two distinct populations of TRPM8-expressing somatosensory ne urons. J Neurophysiol 2006;95:1221-1230. Xu H, Blair NT, Clapham DE. Camphor activates a nd strongly desensitizes the transient receptor potential vanilloid subtype 1 channel in a vanilloid-independent mechanism. J Neurosci 2005;25:8924-8937. 137

PAGE 138

Xu H, Ramsey IS, Kotecha SA, Moran MM, Chong JA, Lawson D, Ge P, Lilly J, Silos-Santiago I, Xie Y, DiStefano PS, Curtis R, Cl apham DE. TRPV3 is a calcium-permeable temperature-sensitive cation ch annel. Nature 2002;418:181-186. Zanotto KL, Iodi Carstens M, Carstens E. Crossdesensitization of responses of rat trigeminal subnucleus caudalis neurons to cinnam aldehyde and menthol. Neurosci Lett 2008;430:29-33. Zanotto KL, Merrill AW, Carstens MI, Carstens E. Neurons in superficial trigeminal subnucleus caudalis responsive to oral cooling, menthol, and other irritant stimuli. J Neurophysiol 2007;97:966-978. Zohar J, Matar MA, Ifergane G, Kaplan Z, Cohen H. Brief post-stressor treatment with pregabalin in an animal model for PTSD: Shor t-term anxiolytic effects without long-term anxiogenic effect. Euro Ne uropsychopharmacol 2008;18:653-666. Zurborg S, Yurgionas B, Jira JA, Caspani O, Heppenstall PA. Direct activation of the ion channel TRPA1 by Ca2+. Nat Neurosci 2007;10:277-279. 138

PAGE 139

BIOGRAPHICAL SKETCH Heather Rossi graduated cum laude from the University of Nort h Carolina at Asheville in May 2003, with a bachelors degree in biol ogy. Later that fall, she enrolled in the Interdisciplinary Program in the Biomedical Scien ces at the University of Florida. She joined the laboratory of Dr. John Neubert in January 2005, and successfully advanced to doctoral candidacy that October. She has been a memb er of the Society for Neuroscience since 2003 and has presented a poster at the last three annual meetings. She al so has two first-authored and four co-authored publications from her time in Dr. Neuberts laboratory and plans to submit several manuscripts before she graduates in December 2008. She will be the first member of her family to receive a doctoral degree. 139