The roles of TRPM8 in cold sensation: the six sides of TRPM8

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content

THE ROLES OF TRPM8 IN COLD SENSATION: THE SIX SIDES OF TRPM8

by

Wendy Michelle Knowlton

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(NEUROSCIENCE)

August 2012

Copyright 2012 Wendy Michelle Knowlton
ii

To Farhan, for helping me ‘keep it together.’

iii
ACKNOWLEDGEMENTS
Without the assistance and support of a great number of people, this certainly
would not have been possible. My thesis advisor, Dr. David McKemy, has given me the
support and liberty to explore new ideas, experiment, and grow as a scientist, and I am
thankful for his trust in me. I have learned a great deal under his advisement, and will
remember the lessons learned throughout my career.
I would also like to thank the other members of my thesis committee, Drs.
Samantha Butler, Sarah Bottjer, Chien-Ping Ko, and Robert Maxson for their ideas,
support, and feedback. In particular, Dr. Samantha Butler provided me with some
invaluable personal and professional guidance over the years, and I will forever try to
remind myself of her sage advice to “[not] overthink it” and to do the science that I love.
Several other faculty members at USC and elsewhere contributed to my progress
over the years, including Dr. Jonah Chan, who was initially on my guidance committee
during my qualifying examination. I would also like to thank Dr. Alapakkam Sampath for
his input and advice, Dr. Diana Bautista for collaborating with me on my first paper, Mr.
Gene Woon for deciding I would be a scientist, and Dr. Joyce Schroeder for persuading
me to go to graduate school in the first place. I would be remiss not to acknowledge the
financial support of the Neuroscience Graduate Program, the Women in Science and
Engineering program, as well as the CBM training grant headed by Dr. Michael Stallcup.
I have witnessed a great change within the lab since I joined in 2006, including
the changing of buildings and colleges, the coming and going (and coming) of grants, as
well as tenure for Dr. McKemy, and I would like to thank the numerous people I have
iv
had to good fortune to work with. In my early days, I was constantly hounding Luke
Daniels and Yoshio Takashima (now both Dr.) for help with equipment, ideas, and
plentiful other miscellaneous problems and I am grateful for their patience and
mentorship. Over the years I was able to develop my own mentoring and training skills
with undergraduate students Sabeen Chawla, Shawna Kleban, Kavita Renduchintala, and
Rebekah Romanu, and I thank them for their help and perseverance. I have enjoyed the
company, assistance, and camaraderie of a good number of labmates, including Narbeh
Bandary, Daniel McCoy, Yun Li, Erika Lippoldt, Jessica Chen, and especially Radhika
Palkar, all of whom have made direct contributions to this project. It has also been a
pleasure to work with the various members of the Ko, Bottjer, and Butler labs.
My friends in and out of the program have provided me with some much-needed
stress relief over the years, and I would like to thank them all. I would particularly like
to thank Benjamin Files and Nikki Derdzinske for their many invitations for dinner, and
Sum-Yan Ng, Christin Chong, and Xiaofei Yang for our semi-regular brunches. I would
also like to thank the members of my knitting/book/women in science group (KABOOM)
for their friendship and support, as well as my roommate, Teresa Chang, for her
friendship and invaluable help with my dog Mali. Support from my friends from college
(and earlier!), including Anne Armstrong, Susan Price, Kathleen Johnson, Michael Miller,
and Sarah Dreeha has been much appreciated.
My family has been hugely supportive, and I would like to especially thank my
mom Pam and dad Tim for their encouragement, patience, support, love, and
understanding and for helping me through the highs and lows (both real and imagined)
v
of graduate school. My brother has provided some much-needed distraction in the
form of a sister-in-law and five wonderful nieces and nephews, all of whom I thank for
their good wishes and encouragement.
Lastly, and most significantly, I want to thank my colleague, collaborator, and
companion Farhan Baluch, who has helped me in so many ways I cannot even begin to
list. He has been a tremendous source of energy, encouragement, friendship, and love,
and I am truly grateful for and touched by his support.
vi
TABLE OF CONTENTS
DEDICATION ii

ACKNOWLEDGEMENTS iii

LIST OF TABLES ix

LIST OF FIGURES x

ABSTRACT xii

CHAPTER ONE Introduction to TRPM8 1
Introduction 1
Molecular Structure 1
Gating Mechanisms 4
Cellular Modulation 9
Neurochemical phenotype 10
Expression patterns 12
Health implications 13
Conclusion 16

CHAPTER TWO Materials and Methods 18
Animals 18
TRPM8
DTR-GFP
transgenic mice 19
In vitro methods 26
Behavioral assays 29

CHAPTER THREE Innocuous Cold Sensing 38
Introduction 38
Temperature preference and cold avoidance 40
Evaporative cooling 47
Conclusion 49

vii
CHAPTER FOUR Noxious Cold Sensing 51
Introduction 51
Icilin-induced wet dog shaking 53
Icilin-induced paw flinching 54
Neural activation 56
Temperature preference and cold avoidance 60
Cold plate 63
Conclusion 65

CHAPTER FIVE Cold Hypersensitivity with Injury 67
Introduction 67
TRPM8-expressing cells and inflammatory injury 68
PBMC attenuates inflammatory cold hypersensitivity 70
TRPM8-expressing cells and neuropathic injury 70
PBMC attenuates neuropathic cold hypersensitivity 74
Conclusion 75

CHAPTER SIX Thermoregulation 77
Introduction 77
M8KO mice are deficient in icilin-induced hyperthermia 77
PBMC induces hypothermia in a dose- and TRPM8-dependent manner 79
Conclusion 83

CHAPTER SEVEN Cooling-Induced Analgesia 86
Introduction 86
Cooling-induced analgesia 88
Cooling-induced analgesia is dependent on TRPM8 and TRPM8-expressing
cells 89
Conclusion 93

CHAPTER EIGHT Antipruritis 95
Introduction 95
Development of spontaneous skin lesions in ablated mice 97
Induced itch 99
Itch hypersensitivity is TRPM8-dependent 103
Conclusion 105

CHAPTER NINE Conclusion 106

REFERENCES 113

viii
APPENDIX A TRPM8 Haploinsufficiency 131
Introduction 131
Decreased Fos expression in M8-het mice 134
Sensory behaviors 134
Quantitative assessment of TRPM8 expression in M8-het sensory ganglia 140
Conclusion 141

APPENDIX B TRPA1 and Noxious Cold Sensing 144
Introduction 144
TRPM8/TRPA1 double-null mice 149
Conclusion 155

ix
LIST OF TABLES
CHAPTER ONE Introduction to TRPM8
Table 1.1 TRPM8 cold thresholds across species 5
Table 1.2 TRPM8 activators 7
Table 1.3 TRPM8 inhibitors 8

CHAPTER TWO Materials and Methods
Table 2.1 Control line behaviors 24

x
LIST OF FIGURES
CHAPTER TWO Materials and Methods
Figure 2.1 GFP fluorescence in DTR ganglia 20
Figure 2.2 Loss of GFP and TRPM8 after DTx injection 21
Figure 2.3 Specificity of TRPM8-cell ablation 23
Figure 2.4 Non-cold behaviors preserved in ablated mice 25
Figure 2.5 MouseChaser heatmaps of wildtype mice 32

CHAPTER THREE Innocuous Cold Sensing
Figure 3.1 Temperature preference vs. cold avoidance 41
Figure 3.2 Temperature preference in the innocuous range 42
Figure 3.3 Cold avoidance in the innocuous range 43
Figure 3.4 MouseChaser heatmaps of innocuous cold 44
Figure 3.5 Evaporative cooling 48
Figure 3.6 Evaporative cooling with PBMC 49

CHAPTER FOUR Noxious Cold Sensing
Figure 4.1 Wet dog shakes 53
Figure 4.2 Icilin-induced nocifensive responses 55
Figure 4.3 Fos induction with cold and chemicals 57
Figure 4.4 Quantification of Fos induction with cold and chemicals 58
Figure 4.5 TRPM8-dependent Fos expression 59
Figure 4.6 Temperature preference 61
Figure 4.7 Noxious cold avoidance 62
Figure 4.8 MouseChaser heatmaps of noxious temperatures 63
Figure 4.9 Cold plate 65

CHAPTER FIVE Cold Hypersensitivity with Injury
Figure 5.1 Allodynia and hyperalgesia 68
Figure 5.2 Cold hypersensitivity with CFA 69
Figure 5.3 Mechanical hypersensitivity with CFA 70
Figure 5.4 PBMC treatment of CFA-induced cold hypersensitivity 71
Figure 5.5 Cold hypersensitivity with CCI 72
Figure 5.6 Mechanical hypersensitivity with CCI 73
Figure 5.7 PBMC treatment of CCI-induced cold hypersensitivity 74

xi
CHAPTER SIX Thermoregulation
Figure 6.1 Icilin-induced hyperthermia 78
Figure 6.2 Capsaicin-induced hypothermia 79
Figure 6.3 DS vehicle control 80
Figure 6.4 SPS vehicle control 81
Figure 6.5 10mg/kg PBMC-induced hypothermia 82
Figure 6.6 20mg/kg PBMC-induced hypothermia 83

CHAPTER SEVEN Cooling-Induced Analgesia
Figure 7.1 Gate theory of pain for extrinsic analgesia by sensory
activation 88
Figure 7.2 Cooling-induced analgesia in wildtype mice 91
Figure 7.3 Lack of cooling-induced analgesia in M8KO mice 92
Figure 7.4 Lack of cooling-induced analgesia in ablated mice 93

CHAPTER EIGHT Antipruritis
Figure 8.1 Alloknesis and hyperknesis 95
Figure 8.2 Spontaneous lesions in ablated mice 98
Figure 8.3 Histamine hyperknesis 100
Figure 8.4 Chloroquine hyperknesis 100
Figure 8.5 BAM8-22 alloknesis 101
Figure 8.6 Loss of serotonin responses 102
Figure 8.7 TRPM8-dependent alloknesis and hyperknesis 103
Figure 8.8 Partially TRPM8-dependent serotonin effect 104

APPENDIX A TRPM8 Haploinsufficiency
Figure A.1 Dorsal horn Fos expression and TRPM8 heterozygosity 133
Figure A.2 Evaporative cooling and cold plate and TRPM8 heterozygosity 135
Figure A.3 Temperature preference and avoidance and TRPM8
heterozygosity 137
Figure A.4 Icilin-induced wet dog shakes and TRPM8 heterozygosity 139
Figure A.5 Decreased TRPM8 expression in M8-het ganglia 141

APPENDIX B TRPA1 and Noxious Cold Sensing
Figure B.1 Increased baseline calcium with TRPA1 expression 146
Figure B.2 DKO temperature preference and cold avoidance 150
Figure B.3 Icilin-induced flinching and TRPA1 152
Figure B.4 Dorsal horn Fos expression and TRPA1 154

xii
ABSTRACT
Our ability to detect temperatures is conferred by thermally-activated sensory
neurons of the peripheral nervous system which innervate the skin, with cell bodies
located in the dorsal root and trigeminal ganglia. These thermosensitive neurons
express thermally sensitive ion channels of the transient receptor potential (TRP) family,
including TRPV1, which responds to hot temperatures as well as the ‘hot’ ingredient in
chili peppers, capsaicin, and TRPM8, which responds to cold temperatures and the ‘cool’
ingredient in mint, menthol. Beyond thermosensation, these channels have been
implicated in a variety of functions including pain, thermoregulation, pain relief, and
itch.
Here we examine the roles of the cold-sensitive channel TRPM8 in sensory
behavioral responses. We use three main approaches in this study: the disruption of
the Trpm8 gene in mice, the ablation of TRPM8-expressing cells in adulthood using BAC
transgenesis and the simian diphtheria toxin receptor transgene, and administration of
the potent TRPM8 antagonist PBMC.
We found six behaviors are affected by TRPM8 expression using one or more of
these methods. These behaviors fall into two general categories: active, or behaviors
where activation of TRPM8-expressing cold-sensitive neurons generates to behavioral
responses, and inhibitory, where the activation of TRPM8 cells leads to the inhibition of
behavioral responses. Active behaviors include responses to innocuous cool, noxious
cold, thermoregulation, and cold hypersensitivity after injury. Inhibitory behaviors
include inhibition of pain by mild cooling and the inhibition of itch.
xiii
The data presented here expand the previously established role of TRPM8 as an
in vivo sensor of innocuous cold into a channel involved in a variety of behaviors,
including some which are intuitively opposed (i.e. pain and analgesia). Furthermore, we
have found a previously unappreciated role for TRPM8 in the inhibition of itch, a finding
which will surely create new avenues of drug development for the treatment of chronic
itch conditions. Altogether, these studies suggest that the population of sensory
neurons expressing TRPM8 is quite diverse, and that understanding this diversity may
someday enable us to pharmacologically manipulate specific subsets, leading to novel
treatments for various sensory-related pathologies as well as a better understanding of
the logic of thermosensation and pain.

1
CHAPTER ONE: Introduction to TRPM8
Introduction
TRPM8 is a member of the transient receptor potential (TRP) superfamily of ion
channels which perform a myriad of functions in vertebrates and invertebrates alike.
Originally, TRPM8 was reported as a specific marker for prostate cancer cells (then
deemed Trp-p8) and shortly thereafter as a cold- and menthol-receptor in sensory
neurons (Tsavaler et al. 2001; McKemy, Neuhausser, and Julius 2002; Peier et al. 2002).
This chapter serves to outline the current state of our knowledge of the in vitro
functions of TRPM8, including its structure, function, and potential as a drug target for
non-sensory conditions ranging from asthma to cancer. The in vivo effects of neuronal
TRPM8 will be discussed in the following chapters and can be divided into six categories:
innocuous cool sensing, noxious cold sensing, cold hypersensitivity after injury,
thermoregulation, analgesia, and antipruritis. The capacity for TRPM8 and TRPM8-
expressing cold sensory neurons to participate in such a myriad of sensory behaviors is
truly remarkable and may represent a powerful system for treating such pathologies as
chronic pain and itch.

Molecular Structure
TRPM8 is encoded by a cDNA with an open reading frame of 3312 nucleotides
coding for a protein of 1104 amino acids, yielding a protein of about 128 kDa (Tsavaler
et al. 2001; McKemy, Neuhausser, and Julius 2002; Peier et al. 2002). The mouse and
rat versions are 92% and 93% identical to the human version, respectively (McKemy,
2
Neuhausser, and Julius 2002; Peier et al. 2002). TRPM8 forms a six-pass transmembrane
protein, with a long segment between transmembrane domains S5 and S6 forming a
pore loop when the protein tetramerizes. The 690 amino acid cytoplasmic N-terminus
of TRPM8 contains a coiled-coil domain and four TRPM family homology regions yet is
devoid of any obvious enzymatic domains (McKemy, Neuhausser, and Julius 2002; Peier
et al. 2002; Phelps and Gaudet 2007). The Ser9 and Thr17 residues are sites of
phosphorylation and positions 40-86 are required for proper subunit targeting to the
plasma membrane (Phelps and Gaudet 2007; Bavencoffe et al. 2010). Expression of a
TRPM8 variant missing the N-terminus and first two transmembrane domains was
reported, although the prevalence of this isoform is not yet fully understood (Sabnis et
al. 2008).
TRPM8 has a short cytoplasmic C-terminus compared to other members of the
TRPM ion channel family, containing only 127 amino acids (McKemy, Neuhausser, and
Julius 2002; Peier et al. 2002). It contains a TRP domain, common to all transient
receptor potential channels. The C-terminus also contains a coiled-coil domain which
directs the assembly of TRPM8 into a tetrameric ion channel, and mutations within this
domain prevent channel formation and possibly TRPM8 expression (Erler et al. 2006;
Tsuruda, Julius, and Minor Jr. 2006; Pedretti et al. 2009). Coiled-coil units self-assemble
to form the channel, and the presence of a membrane-tethered form of the C-terminus
coiled-coil domain is sufficient to disrupt channel function (Erler et al. 2006; Tsuruda,
Julius, and Minor Jr. 2006).
3
The C-terminus is also the site of action of many TRPM8 activators. The channel
is activated by cold temperatures ranging from 25-8°C and by cold-mimetic chemicals
like menthol and icilin, which induce the psychophysical sensation of cold (McKemy,
Neuhausser, and Julius 2002; Peier et al. 2002; Bandell et al. 2006). In a set of clever
experiments, Brauchi and colleagues determined that exchanging the TRPM8 C-terminal
domain with that of the related heat-activated channel TRPV1 leads to activation of
TRPM8 by heat (and TRPV1 by cold), leading to the conclusion that the temperature-
sensing domain is modular and located in the C-terminal domains of both channels
(Brauchi et al. 2006; Brauchi et al. 2007). Similar experiments revealed that the C-
terminal domain is also responsible for the activation of TRPM8 by phosphatidylinositol-
4,5-bisphosphate (PIP
2
) (Brauchi et al. 2007).
The channel contains one confirmed glycosylation site at Asn934, which lies in
the pore-loop region between S5 and S6 (Dragoni, Guida, and McIntyre 2006; Erler et al.
2006). Although mutations of this residue lead to loss of glycosylation, this does not
appear to affect the channel’s expression, tetramerization, trafficking, or activation by
various agonists, but rather plays a facilitating role in one or more of these processes.
However, mutating two cysteine residues flanking the glycosylation site (Cys929 and
Cys940) abolishes channel function and leads to subunit dimerization instead of
tetramerization (Dragoni, Guida, and McIntyre 2006).

4
Gating mechanisms
When activated, TRPM8 is an outwardly-rectifying nonselective cation channel,
permeant to sodium, potassium, cesium, and calcium ions (McKemy, Neuhausser, and
Julius 2002; Peier et al. 2002). Its preference for passing cations rather than anions is
conferred by the S6 region (Kuhn, Knop, and Luckhoff 2007). The channel is partially
gated by voltage with a reversal potential of 0mV, and can be activated with membrane
depolarization (Peier et al. 2002; Voets et al. 2004; Matta and Ahern 2007). Mutations
of charged residues within the S4 and S4-S5 linker region, specifically Arg842 and
Lys856, lead to decreased voltage gating, implicating this region as the voltage sensor
(Voets et al. 2007). Channel opening and closing has been explained with an eight-state
kinetic model wherein the channel undergoes conformational changes when gated by
voltage, cold, or pharmacological agonists, allowing cations to flow through the pore
(Brauchi, Orio, and Latorre 2004; Voets et al. 2007).
Voltage and temperature sensing by TRPM8 are closely linked, as temperature
affects the maximum open probability of the channel in response to voltage changes,
and a change in the channel’s ability to sense voltage affects its thermal gating (Brauchi,
Orio, and Latorre 2004; Voets et al. 2004; Voets et al. 2007). The thermal threshold of
heterologously-expressed TRPM8 under normal conditions is around 25°C in mice, rats,
and humans (see Table 1.1), and the current saturates at temperatures lower than 8°C
(McKemy, Neuhausser, and Julius 2002; Peier et al. 2002; Weil et al. 2005). Importantly,
the kinetics of activation of the channel by cold is invariant with cooling rate, indicating
that the process is not simply enzymatic (McKemy, Neuhausser, and Julius 2002). The
5
channel is, however, desensitized with prolonged cold stimulation, which may be
reversed by briefly warming the cell to physiological temperatures.

Table 1.1: TRPM8 cold thresholds across species
Species Temperature threshold Reference
human ~25°C
(Tsavaler et al. 2001; Weil et al. 2005)
rat 22-28°C
(McKemy, Neuhausser, and Julius 2002)
mouse 22-28°C
(Peier et al. 2002)
chicken 25°C
(Chuang, Neuhausser, and Julius 2004; Saito
and Shingai 2006)
canine 17°C
(Y Liu et al. 2006)
Xenopus laevis* 15°C
(Myers, Sigal, and Julius 2009)
Xenopus tropicalis* undetermined
(Saito and Shingai 2006)
crocodile undetermined
(Seebacher and Murray 2007)(Seebacher
and Murray 2007)
*TRPM8 is duplicated in the frog genome, with only the “a” form thus far reported as functional.

The “cooling” ingredient in peppermint, menthol, is the most well-known
chemical agonist of the channel, and indeed was central in one method of the initial
cloning of TRPM8 (McKemy, Neuhausser, and Julius 2002). Menthol modulates the
open probability and ion conductance of the channel and shifts the voltage sensitivity to
physiological voltages (Voets et al. 2004; Hui, Guo, and Feng 2005; Matta and Ahern
2007). It also shifts the temperature threshold of the channel to higher temperatures,
for example from ~25°C to 29.7°C in the presence of 20μM menthol (McKemy,
Neuhausser, and Julius 2002; Peier et al. 2002). Mutations in the Tyr745 residue within
the S2 region render the channel insensitive to menthol (Bandell et al. 2006). As with
cold stimuli, the channel desensitizes with prolonged menthol exposure in a calcium-
6
dependent manner (McKemy, Neuhausser, and Julius 2002; Peier et al. 2002; Daniels,
Takashima, and McKemy 2009).
Icilin (AG-3-5) is another well-characterized agonist of TRPM8 and produces
sensations of cold when applied topically (Wei 1981; Wei and Seid 1983). It activates
TRPM8 with much greater potency than menthol in a calcium-dependent manner, and,
like cold and menthol, its currents desensitize with prolonged stimulation (McKemy,
Neuhausser, and Julius 2002). Rises in intracellular calcium play a critical role in icilin
action on TRPM8, leading to more reliable channel gating (Chuang, Neuhausser, and
Julius 2004). Icilin sensitivity lies in the Gly805, Asp802, and partially in the Asn799
residues in the S3 region, and changes in one or more of these amino acids render the
channel insensitive to icilin. These mutations do not appear to affect channel activation
by cold or menthol, however rendering the channel insensitive to activation by icilin has
revealed that this chemical may interfere with currents induced by these other
activators (Chuang, Neuhausser, and Julius 2004; Kuhn, Kuhn, and Luckhoff 2009).
A number of other compounds have been reported to activate TRPM8 (Eid and
Cortright 2009). These include eucalyptol and the menthol-derived WS compounds as
well as many others as listed in Table 1.2 (McKemy, Neuhausser, and Julius 2002;
Behrendt et al. 2004; Beck et al. 2007; Bodding, Wissenbach, and Flockerzi 2007; Ma et
al. 2008). Thus far, WS-12 is the most potent known agonist of TRPM8, with an EC
50

value of 193nM (Bodding, Wissenbach, and Flockerzi 2007). Endogenous activators
have also been found, such as lysophospholipids resulting from activity of calcium-
independent phospholipases (iPLA2), which are proposed as the primary activators of
7
TRPM8 expressed outside of the sensory nervous system (Vanden Abeele et al. 2006;
Benedikt et al. 2007).

Table 1.2: TRPM8 activators
Activator EC
50
Reference
Chemical Agonists
WS-12 193nM
(Beck et al. 2007; Bodding, Wissenbach, and Flockerzi 2007;
Ma et al. 2008)
Icilin 360nM
(McKemy, Neuhausser, and Julius 2002)
CPS-113 1.2µM
(Bodding, Wissenbach, and Flockerzi 2007)
FrescolatML 3.3µM
(Behrendt et al. 2004)
CPS-369 3.6µM
(Bodding, Wissenbach, and Flockerzi 2007)
WS-3 3.7µM
(Behrendt et al. 2004)
Menthol 4.1µM
(McKemy, Neuhausser, and Julius 2002; Peier et al. 2002;
Behrendt et al. 2004)
WS-148 4.1µM
(Bodding, Wissenbach, and Flockerzi 2007)
FrescolatMAG 4.8µM
(Behrendt et al. 2004)
WS-30 5.6µM
(Bodding, Wissenbach, and Flockerzi 2007)
Cooling Agent 10 6µM
(Behrendt et al. 2004)
PMD 38 31µM
(Behrendt et al. 2004)
WS-23 44µM
(Behrendt et al. 2004)
Coolact P 66µM
(Behrendt et al. 2004)
Eucalyptol 3.4mM
(McKemy, Neuhausser, and Julius 2002)
Geraniol 5.9mM
(Behrendt et al. 2004)
Linalool 6.7mM
(Behrendt et al. 2004)
Hydroxycitronellal 19.6mM
(Behrendt et al. 2004)

Physiological activators

Cold temperatures -
(McKemy, Neuhausser, and Julius 2002; Peier et al. 2002)
Voltage -
(McKemy, Neuhausser, and Julius 2002; Peier et al. 2002)
PIP
2
-
(B. Liu and Qin 2005)
Lysophospholipids -
(Vanden Abeele et al. 2006; Andersson, Nash, and Bevan
2007)

The membrane phospholipid PIP
2
is required for channel action, binding to the C-
terminus of the protein (Brauchi et al. 2007). A two percent solution of ethanol
interferes with the interaction between PIP
2
and the channel, effectively inhibiting
8
channel action (Benedikt et al. 2007). The channel can be directly blocked by the TRPV1
antagonists BCTC, thio-BCTC, and capsazepine and the general calcium-entry channel
blocker SKF96365, and we recently reported PBMC as a TRPM8-specific antagonist that
does not act on other TRP channels (see Table 1.3) (Behrendt et al. 2004; Weil et al.
2005; Madrid et al. 2006; Malkia et al. 2007; Knowlton et al. 2011) . Natural products
such as polyvalent cations, phytocannabinoids, and polyunsaturated fatty acids (PUFAs)
are known to inhibit the channel’s function, as does the antifungal drug clotrimazole

Table 1.3: TRPM8 inhibitors
Inhibitor IC
50
Reference
Chemical blockers
PBMC 0.6nM
(Knowlton et al. 2011)
CTPC 131nM
(Weil et al. 2005)
Clotrimazole 200nM
(Meseguer et al. 2008)
SB-452533 571nM
(Weil et al. 2005)
AMTB 588nM
(Lashinger et al. 2008)
BCTC 800nM
(Behrendt et al. 2004; Weil et al. 2005;
Madrid et al. 2006; Malkia et al. 2007)
SKF96365 1µM
(Malkia et al. 2007)
NADA 2µM
(De Petrocellis et al. 2007)
anandimide 3µM
(De Petrocellis et al. 2007)
thio-BCTC 3.1µM
(Behrendt et al. 2004)
arachidonic acid 3.2µM
(Andersson, Nash, and Bevan 2007)
capsazepine 18µM
(Behrendt et al. 2004; Weil et al. 2005)
1,10-phenanthroline 140µM
(Malkia et al. 2007)

Physiological inactivators Action

low pH Surface charge screening
(Andersson, Chase, and Bevan 2004)
PLL PIP
2
scavenger
(B. Liu and Qin 2005)
Ruthenium red
(intracellular)
PIP
2
scavenger
(B. Liu and Qin 2005)
Spermine (intracellular) PIP
2
scavenger
(B. Liu and Qin 2005)
Ethanol PIP
2
-binding interference
(Weil et al. 2005; Benedikt et al. 2007)
PLC activation PIP
2
hydrolysis
(B. Liu and Qin 2005; Rohacs et al. 2005;
Daniels, Takashima, and McKemy 2009)
PKC activation Channel dephosphorylation
(Premkumar et al. 2005; Abe et al. 2006)
PKA activation Channel dephosphorylation
(De Petrocellis et al. 2007; Bavencoffe et
al. 2010)

9
(B. Liu and Qin 2005; Andersson, Nash, and Bevan 2007; De Petrocellis et al. 2007;
Meseguer et al. 2008). Extracellular acidity also appears to inhibit channel function,
reportedly through charge screening at the extracellular surface of the channel
(Andersson, Chase, and Bevan 2004; Behrendt et al. 2004; Mahieu et al. 2010).

Cellular modulation
A handful of cellular processes have been implicated in the modulation of
TRPM8, particularly in the context of thermal adaptation. The best-understood
mechanism of TRPM8 functional regulation is by the phospholipid PIP
2
. PIP
2
association
with the C-terminus is required for channel function, and its depletion (such as that seen
in an excised membrane patch) leads to the rundown of TRPM8 currents (Voets et al.
2004; B. Liu and Qin 2005). Addition of PIP
2
-scavenging molecules to the intracellular
surface of the cell or treatment with inhibitors of PIP
2
synthesis significantly interferes
with channel activity (Liu and Qin 2005). Activation of PLC, either by rises in intracellular
calcium or activation by other proteins, leads to PIP
2
cleavage, thereby desensitizing the
channel (Rohacs et al. 2005; Daniels, Takashima, and McKemy 2009). Indeed, the
negative feedback loop of TRPM8-mediated calcium flux leading to channel
desensitization is responsible for thermal adaptation in sensory neurons (Daniels,
Takashima, and McKemy 2009).
In addition to modulation by phosphates such as PIP
2
, TRPM8 has been found to
associate with inorganic polyphosphates in vitro (Zakharian et al. 2009). Activation of
PKC leads to activation of phosphatases which dephosphorylate TRPM8 and thereby
10
desensitize the channel (Premkumar et al. 2005; Abe et al. 2006). Similarly, activation of
the PKA pathway leads to channel desensitization (De Petrocellis et al. 2007). Thus far
the only protein shown to directly associate with TRPM8 is Pirt, which interacts with
phosphoinositides and presumably (as data regarding function of the interaction are
lacking) desensitizes the channel (A. Y. Kim et al. 2008). Finally, regulation of TRPM8
currents in neurons can occur at the whole-cell level, with the relative strengths of
TRPM8 and potassium currents determining a neuron’s thermal threshold (Madrid et al.
2009).

Neurochemical phenotype
TRPM8-expressing neurons are phenotypically diverse, expressing a wide array
of different neurochemical markers (Takashima et al. 2007; Takashima, Ma, and
McKemy 2010). Up to fifty percent of TRPM8-expressing sensory neurons also express
TRPV1 both under normal and injury conditions (McKemy, Neuhausser, and Julius 2002;
Abe et al. 2005; Hjerling-Leffler et al. 2007; A Dhaka et al. 2008). Some TRPM8 neurons
express other markers of nociceptive phenotype including CGRP and Substance P while
others do not, although the remaining cells do not bind the non-nociceptor marker IB
4

(Abe et al. 2005; Xing et al. 2006; Takashima et al. 2007; A Dhaka et al. 2008; Axelsson et
al. 2009; Hayashi et al. 2009).
TRPM8 is found in around twelve percent of small-diameter sensory neurons in
the dorsal root and trigeminal ganglia (McKemy, Neuhausser, and Julius 2002; Peier et
al. 2002; Takashima et al. 2007). TRPM8 is found in both Aδ- and C-fibers as indicated
11
by its colocalization with both NF200/NF150 and peripherin, markers of each fiber type,
respectively (Kobayashi et al. 2005; Takashima et al. 2007; A Dhaka et al. 2008; Hayashi
et al. 2009). Cold-sensitive TREK potassium channels are found in up to thirty percent of
TRPM8-expressing neurons, and the K
v
1 family of potassium channels are also
coexpressed to varying degrees (Madrid et al. 2009; Yamamoto, Hatakeyama, and
Taniguchi 2009). Intriguingly, TRPM8 agonists have been shown to functionally affect a
population of slowly adapting mechanoreceptors, although until molecular markers of
mechanosensory phenotype are identified it is unknown what percentage of these two
modalities of somatosensation overlap (Cahusac and Noyce 2007).
During embryonic development, TRPM8 activity and mRNA is first seen in
sensory ganglia around embryonic day 16.5 in mice (Hjerling-Leffler et al. 2007;
Takashima, Ma, and McKemy 2010). TRPM8-expressing neurons arise from a
population of TRPV1-expressing cells and depend on proper expression of the NGF
receptor TrkA during earlier stages, likely due to TrkA-dependent induction of the
transcription factor Runx1 (Peier et al. 2002; Hjerling-Leffler et al. 2007; Luo et al. 2007;
Takashima, Ma, and McKemy 2010). TRPM8-expressing neurons require expression of
cadherin-8 for the formation of proper contacts with spinal cord neurons (Suzuki et al.
2007). Interestingly, TRPM8 function is not required for the development of these
neurons as mice homozygous for replacement of the TRPM8 gene with GFP show
normal patterns of target innervation and colocalization with various markers (A Dhaka
et al. 2008).

12
Expression patterns
As a physiological sensor of environmental temperatures, TRPM8 should be
positioned at the body-environment interface—i.e. the skin. Indeed, TRPM8-expressing
nerve fibers are found in free nerve terminations in multiple layers of the epidermis and
near sebaceous glands in the skin, making it well-positioned to mediate sensations of
both innocuous cool and painful cold (Takashima et al. 2007; A Dhaka et al. 2008;
Axelsson et al. 2009; Stucky et al. 2009). TRPM8-expressing nerve terminals are also
found in the oral cavity and tongue, although not in taste buds (Abe et al. 2005;
Takashima et al. 2007; A Dhaka et al. 2008). TRPM8 is also found in both Aδ- and C-
fibers within the pulp and dentin of the tooth and is thus implicated in conditions of
tooth hypersensitivity to cold (Park et al. 2006; Takashima et al. 2007). The cornea is
innervated by TRPM8-expressing nerve fibers as well, which are required for
maintaining ocular wetness (Madrid et al. 2006; Carr et al. 2009; Parra et al. 2010).
TRPM8 expression is required for normal cold sensation in the mouse, as will be
discussed in following chapters (Bautista et al. 2007; Colburn et al. 2007; Daniels and
McKemy 2007; Ajay Dhaka et al. 2007; Knowlton et al. 2010).
Before the discovery of endogenous activators of TRPM8, the channel’s
expression in tissues not involved in thermosensation was puzzling. For example,
TRPM8 expression was found in the male urogenital tract, including the bladder
urothelium, the prostate, the testes, and even in sperm, which all have no known
thermosensory function but potentially implicate the channel in male fertility (Tsavaler
et al. 2001; R. J. Stein et al. 2004; S. Du et al. 2008; Everaerts et al. 2008; Sabnis et al.
13
2008; De Blas et al. 2009). The channel has also been found in the vasculature, including
the aorta and the pulmonary artery, where it may play a role in regulating vascular tone
(Yang et al. 2006; Seebacher and Murray 2007; Johnson et al. 2009). Expression is found
in the lungs, both in afferent nerves and in epithelial cells, however the epithelial
version of the channel is truncated and localized to the endoplasmic reticulum (Sabnis
et al. 2008; Xing et al. 2008). The truncated form is also found in the tongue and kidney,
while full-length TRPM8 has been reported in brain, liver, and nodose, jugular, and
petrosal ganglia (Zhang et al. 2004; Sabnis et al. 2008; J. Du et al. 2009; Hondoh et al.
2010; Zhao, Sprunger, and Simasko 2010). However, the mechanisms, functions, and
significance of TRPM8 activation by lysophospholipids and other endogenous ligands in
these presumably non-thermosensory tissues remain to be determined.

Health implications
TRP channels in general have been implicated in a number of diseases,
particularly the heat- and capsaicin-receptor TRPV1. The main focus within the TRPM8
field has been on pain and thermosensation, which will be discussed in later chapters,
but the expression of the channel in a number of tissues leads to the possibility that
TRPM8 is involved in a number of pathological conditions. Below we discuss the three
main non-thermal health-related implications of TRPM8 expression, although the list
will surely grow as we learn more about this channel.

14
Cancer
TRPM8 was originally identified in prostate cancer (Tsavaler et al. 2001), and has
recently been identified as a biomarker of metastatic disease detectable in urine (Bai et
al. 2010). Although prostate cells normally express low amounts of TRPM8, expression
is much higher in cancerous cells (Tsavaler et al. 2001; Bidaux et al. 2007). This
overexpression appears to be androgen-regulated and may be involved in cancer cell
survival (Zhang and Barritt 2004; Bai et al. 2010). TRPM8 expression has been found in a
number of other cancers, including melanoma, breast adenocarcinoma, colorectal
cancer, lung cancer, and bladder cancer (Tsavaler et al. 2001; Yamamura et al. 2008; Li
et al. 2009). In the cases of melanoma and bladder cancer, activation of TRPM8 by
menthol leads to a dose-related decrease in cell viability (Yamamura et al. 2008; Li et al.
2009). Menthol-based anticancer drugs are an interesting possibility for treatment;
however, as TRPM8 is implicated in survival of other types of cancer cells and metastatic
disease, agonist-based drugs might have an effect opposite of what is desired (Slominski
2008). Therefore a clearer picture of TRPM8 expression and the mechanism of its
involvement in specific cancers need to be further elucidated (Prevarskaya, Zhang, and
Barritt 2007).
In addition to TRPM8’s role in cancer cell biology, TRPM8 is involved in side
effects of the chemotherapy drug oxaliplatin (Gauchan et al. 2009; Knowlton et al.
2011). It would therefore be reasonable to speculate that delivery of a TRPM8-
inhibiting drug alongside oxaliplatin treatment may alleviate these symptoms, again
15
with the caveat that TRPM8 antagonism may have different effects on different cancers
and may therefore be appropriate in certain diseases and not others.
Bladder conditions
In addition to bladder cancer, TRPM8 has been found to be involved in bladder
processes such as the voiding reflex. TRPM8-mediated cooling of the bladder induces
micturition, and blockade of TRPM8 decreases bladder contractions (Tsukimi et al. 2005;
Lashinger et al. 2008). Stimulation of TRPM8 in the skin of the leg is sufficient to trigger
the voiding reflex in rats (Z. Chen et al. 2009). In painful bladder syndromes, TRPM8
expression is heightened and correlates well with symptom severity (Mukerji et al.
2006). Taken together, the data suggest that TRPM8 blockade would reduce the
symptoms of painful bladder as well as decrease inopportune bladder voiding reflexes in
conditions where voluntary control of this reflex is impaired.
Asthma
As mentioned above, a truncation variant of TRPM8 was recently discovered in
bronchial epithelial cells (Sabnis et al. 2008). Cold-induced asthma is a common
condition, and whether this truncated TRPM8 or full-length TRPM8 in the respiratory
vagal afferents is responsible for this phenomenon is undetermined (Giesbrecht and
Younes 1995; Sabnis et al. 2008; Xing et al. 2008). Regardless, activation of TRPM8 in
these tissues by cold air leads to symptoms of asthma, including increased mucus
production, bronchoconstriction, and increased cytokine expression (Sabnis et al. 2008).
TRPM8 activation has been previously shown to affect some motor circuits, therefore
therapies directed at blocking TRPM8 activation may alleviate both the
16
bronchoconstriction and coughing symptoms as well as prevention of the increase in
expression of inflammatory molecules (Mandadi et al. 2009).

Conclusion
Although an astonishing amount of data has been gathered on TRPM8 since its
discovery, a number of key questions remain. First, at the structural level, the
conformational mechanism of cold activation remains to be fully understood, as well as
the function of the N-terminal domain. At the cellular level, it remains to be seen
whether TRPM8 interacts with any additional proteins or signaling pathways that may
be important in the regulation and function of the channel. In neurons, it is apparent
that TRPM8 cells are diverse and a better understanding of this heterogeneity would be
particularly useful in dissecting the logic and function of these cells within their neural
circuit(s). The purpose and mechanism of TRPM8 expression outside of the peripheral
nervous system is still a mystery and a better understanding may prove vital to avoiding
unwanted side effects with any TRPM8-directed therapies.
With this background in the in vitro actions, cellular-level modulation,
neurochemical phenotype, and expression pattern of TRPM8, we will now turn to the six
main behavioral roles of TRPM8 and TRPM8-expressing neurons. These roles can be
generally broken down into two categories: cold sensing (innocuous, noxious,
pathological, and thermoregulatory), and inhibition of other modalities (analgesia and
antipruritis). Some actions (i.e. cold hypersensitivity and analgesia) seem paradoxical,
17
thus future studies identifying the precise subcircuits of TRPM8 neurons will be key in
unlocking the mysteries of this enigmatic channel.
18
CHAPTER TWO: Materials and Methods
Animals
Mouse lines
Wildtype (abbreviated WT) mice indicated in these studies, as well as all
genetically modified animals, were on the C57Bl/6J background (Charles River). TRPM8-
knockout (abbreviated M8KO) mice were a kind gift from Drs. Sven Jordt and David
Julius. TRPM8-heterozygous (TRPM8
+/-
; abbreviated M8-het) mice were obtained by
breeding M8KO mice to WT mice. TRPA1-knockout (abbreviated A1KO) mice and
TRPM8/TRPA1 double-knockout (abbreviated DKO) mice were a kind gift of Drs. Diana
Bautista and David Julius. Mice expressing the simian diphtheria toxin receptor gene
(DTR) were produced via bacteria artificial chromosome (BAC) transgenesis and are
described below. All transgenic or genetically modified mice were backcrossed onto the
C57Bl/6 background for at least four generations whenever possible. All lines were
normal in overall appearance and viability, and matings produced offspring with
expected Mendelian ratios.
Genotyping
Genotype was determined using a small sample of the animals tail, which was
enzymatically digested in a solution of 100mM Tris, 5mM EDTA, 0.2% SDS, 200mM NaCl,
and 500ug Proteinase K overnight at 55°C. DNA was precipitated with isopropanol and
resuspended in 10mM Tris and 100µM EDTA. Gene products were amplified with the
Advantage 2 PCR kit for BAC-transgenic animals using the following primers:
M8- 9456f: 5’-GCAAACAGAAGAGACATCGCTAGC-3’
19
M8- 10908r: 5’-CCTATGAAGCAGGGAGTTTTATTGC-3’
or with the Qiagen Taq PCR Core Kit for M8KO products with the following primers:
P1-f: 5’-CCTTGGCTGCTGGATTCACACAGC-3’
M-r: 5’-CAGGCTGAGCGATGAAATGCTGATCTG-3’
Products were visualized by agarose electrophoresis.
Care
Animals used for behavioral analysis were aged at least eight weeks. Animals
were provided standard mouse chow and water ad libitum and were housed no more
than five per cage on a 12 hour light/dark cycle at ~21°C. Cages and animal health were
monitored by the USC Department of Animal Resources staff and veterinarians. All
procedures and tests were approved by the University of Southern California
Institutional Animal Care and Use Committee and conducted in accordance with the
recommendations of the International Association for the Study of Pain and the NIH
Guide for the Care and Use of Laboratory Animals.

TRPM8
DTR-GFP
transgenic mice
Generation
The Trpm8 BAC clone (553 H3 of the RPCI.22 BAC library, Invitrogen) was
modified by homologous recombination, targeting the simian diphtheria toxin receptor
fused with GFP (abbreviated DTR) transgene (a gift from Dr. Richard Lang (Jung et al.
2002) to sequences corresponding to the second coding exon in the Trpm8 gene. DNA
sequences flanking this site were PCR-amplified with PfuUltra HF polymerase
20
(Stratagene) and subcloned into the pLD53.SCA-E-B shuttle vector (a gift from Dr.
Nathaniel Heintz) in which the GFP cassette was replaced with DTR. Shuttle vector DNA
was electroporated into electrocompetent bacteria and cells and correct targeting was
determined by PCR analysis using transgene and BAC sequence primer pairs (Takashima
et al. 2007). Targeted BAC DNA was purified by sepharose column and injected into the
pronucleus of fertilized ova at the USC Transgenic Core Facility. Transgenic founder
mice were identified by PCR and backcrossed into the C57Bl/6 strain.
Characterization
Since the DTR transgene is fused to GFP, both trigeminal and dorsal root ganglia
(TG and DRG, respectively) exhibit green fluorescence driven by the TRPM8 promoter in
the transgenic animals (Figure 2.1), comparable to our previously described TRPM8
GFP

BAC-transgenic line (Takashima et al. 2007).
Figure 2.1: GFP fluorescence in DTR ganglia

Trigeminal (TG) and dorsal root (DRG) ganglia taken from
TRPM8DTR-GFP transgenic animals (DTR mice) exhibit green
fluorescence in TRPM8-expressing sensory neurons.
21
The simian form of the receptor is 10
5
times more sensitive to diphtheria toxin
(DTx) than is the native rodent form, thus two low doses (50µg/kg; List Biological
Laboratories) of the toxin administered intraperitoneally (i.p.) over three days leads to
preferential binding of the simian receptor and not the murine form, which results in
Figure 2.2: Loss of GFP and TRPM8 after DTx injection

A) The GFP fluorescence seen in naïve DTR mice is lost when the animals are given two doses of
diphtheria toxin (DTx). Non-transgenic animals, such as a TRPM8
GFP
mouse (Takashima et al.
2007) do not have the same loss of fluorescence, suggesting that the toxin’s effects are specific
to the transgenic simian form of the diphtheria toxin receptor expressed in DTR mice. B)
Quantitative PCR reveals that TRPM8 is lost over the course of two injections of DTx. The levels
of TRPM8 in vehicle-injected animals (n=3) was set to 100% and the levels of TRPM8
normalized to this value. After one injection of DTx (n=3), TRPM8 is detected at only 6.4±0.1%
of normal using the M8 primer set. After two injections (n=3), TRPM8 goes down to 0.1±0.05%
of normal.
A
22
selective ablation of TRPM8-expressing afferents and loss of GFP in the ganglia (Figure
2.2A). Quantitative PCR from RNA extracts of sensory ganglia indicate that the
administration of two low doses of the toxin leads to more than 99% loss of TRPM8
transcripts (Figure 2.2B). Control (nontransgenic littermates injected with DTx) animals
show no such change in TRPM8 transcript levels, nor did they exhibit any other obvious
alterations in appearance or phenotype. The ablation of TRPM8 cells was specific, as no
obvious alterations in the levels markers of types of other sensory ganglia was observed
using microarray analysis (Figure 2.3). All mice were at least eight weeks of age at time
of injection to allow the animals to develop normally and allowed at least four weeks of
recovery before behavioral testing to avoid any confounding inflammation in the
ganglia.
23

Figure 2.3: Specificity of TRPM8-cell ablation

Microarray analysis comparing transcripts of pooled ganglia from control (n=8) and ablated
(n=8) mice show that the DTx-induced cell ablation is specific to TPRM8 cells. TRPM8 is the
most highly affected gene, with a net 20-fold decrease in the ablated samples as compared to
control samples, while three genes known to be expressed to varying degrees in TRPM8 cells
(Cadherin-8, Plcδ4, and TREK-1) show smaller changes. Markers of other types of sensory cells,
such as TRPV1, TRPA1, Mrgprd, and Na
v
1.8 remained unchanged. Graph courtesy of David
McKemy.
24
Behavior controls
Control animals’ behavioral responses to a variety of assays outlined below was
compared to WT mice to ensure that a) BAC transgenesis and expression of the
transgene within the cells does not alter baseline behaviors, and that b) injection of the
toxin into nontransgenic mice does not affect sensory behaviors. No significant
differences in a variety of behavioral assays was observed between nontransgenic
littermates, uninjected TRPM8
DTR-GFP
mice, injected nontransgenic littermates, and WT
mice; the results are summarized in Table 2.1. Also, no significant differences in
behavioral responses in ablated mice was observed outside of the cold-related
behaviors discussed in the following chapters (Figure 2.4).

Table 2.1: Control line behaviors
Assay WT (C57Bl/6)
Uninjected
transgenic
(DTR
+
)
Uninjected WT
littermate (DTR
-
)
Injected WT
littermate
(DTR
-
+DTx)
Acetone score 2.1±0.1 (n=8) 2.1±0.1 (n=12) 2.2±0.1 (n=14) 2.1±0.1 (n=6)
0°C plate flinch
latency (s)
6.8±0.9 (n=8 13.5±2.2 (n=8) 12.3±1.2 (n=12) 6.2±2.1 (n=6)
0°C plate lick
latency (s)
15.6±2.5 (n=8) 22.4±2.3 (n=8) 18.6±3.2 (n=12)
21.3±2.5
(n=6)
Wet dog shakes
(per 20mins.)
18.8±3.9 (n=8)
12.3±2.9
(n=12)
15.3±3.3 (n=14)
26.3±7.0
(n=6)
von Frey paw
withdrawal
threshold (g)
6.5±0.3 (n=8) 6.2±0.2 (n=22) 6.4±0.1 (n=13) 6.6±0.2 (n=6)
48°C plate flinch
latency (s)
9.1±1.8 (n=8)
15.5±2.9
(n=12)
12.0±1.4 (n=14)
13.5±0.8
(n=6)
52°C plate flinch
latency (s)
7.9±1.0 (n=8) 5.8±0.7 (n=12) 6.6±0.6 (n=14) 8.5±0.8 (n=6)
Rotarod drop
latency (s)
14.9±0.8 (n=8)
17.4±1.6
(n=12)
17.3±0.8 (n=29)
15.5±1.3
(n=6)
Grip strength (g) 126.2±8.3 (n=8)
127.6±4.1
(n=12)
123.0±2.9 (n=29)
124.1±6.8
(n=6)
25

Figure 2.4: Non-cold behaviors preserved in ablated mice

Compared to injected WT littermate (Control) mice, ablated mice exhibit normal behaviors in a
variety of assays: A) 48°C and 52°C hot plate flinching latencies of 14.2±1.3s and 8.9±1.2s,
respectively, ablated (n=12) vs. 13.5±0.8s and 8.5±0.8s control (n=6). B) Grip strength limit of
121.4±3.5g ablated (n=27) vs. 124.1±6.8g control (n=6). C) Rotarod drop latency of 16.4±1.2s
ablated (n=27) vs. 15.5±1.3s control (n=6). D) Dynamic stroke response frequency of 27.5±6.5%
ablated (n=12) vs. 30±4.5% control (n=11). E) von Frey paw withdrawal latency 6.8±0.1g ablated
(n=12) vs. 6.6±0.2g control (n=6). Student’s t-test vs. control: n.s. p>0.05.
26
In vitro methods
Heterologous expression
Complementary DNA (cDNA) of rat TRPA1 (gifts from Dr. David Julius) was
transfected into the human embryonic kidney cell line 293-T (HEK293T) using TransIT-
LT1 reagent (Mirus) following the manufacturer’s instructions. Cells were maintained in
a 37°C incubator in 5% CO
2
in DMEM containing 10% fetal bovine serum and 1%
penicillin-streptomycin.
Calcium imaging
Cells were plated on coverslips coated with Matrigel (BD Biosciences) and
incubated overnight at 37°C 5% CO
2
. The next day, cells were incubated for 1 hour with
a 1µM solution of calcium-sensitive indicator Fura-2 (Invitrogen). Cells were challenged
with 20µM mustard oil and 50mM KCl. Data were gathered and analyzed with
MetaFluor imaging software.
Quantitative PCR
RNA transcripts were purified from sensory ganglia using the RNAeasy Mini Kit
(Qiagen). cDNA was made from these extracts from 1µg of RNA using the iScript cDNA
synthesis kit (BioRad) according to the manufacturers’ instructions. Quantitative PCR
(qPCR) was performed using a BioRad CFX96, RT-PCR detection system and the SsoFast
qPCR kit (BioRad) according to the manufacturer’s instructions. The primers used for
TRPM8 transcripts were:
M8-f: 5’-GCTGCCTGAAGAGGAAATTG-3’
M8-r: 5’-GCCCAGATGAAGAGAGCTTG-3’
27
M8probe1-f: 5’- TGTATCTCGGAGCACAGACG -3’
M8probe1-r: 5’- GTGAGAATCCACGCACCTTT -3’
M8 ISHPrim2-f: 5’- TGCCTGCTGTTTCAGAAATG -3’
M8 ISHPrim2-r: 5’- GCACCCCTCTTCAGGTGTAA -3’
The M8 primer set probed exons 10-14 (immediately before the modification made in
the M8KO line) while the M8probe1 set probed exons 3-7 and the M8 ISHPrim2 set
probed exons 23-27. The reference gene used was GAPDH, amplified with the following
primer set:
GAPDH-f: 5’- TGTAGACCATGTAGTGAGGTCA-3’
GAPDH-r: 5’- AGGTCGGTGTGAACGGATTTG-3’
Microarray analysis
Trigeminal and dorsal root ganglia from injected wildtype littermate and ablated
mice (n=8 each group) were quickly dissected and stored in a pooled sample in RNAlater
buffer (Qiagen). RNA transcripts were purified with the RNAeasy Mini Kit (Qiagen) and
5µl reserved for qPCR analysis. RNA was taken to the USC Microarray core and
processed using an Affymetrix GeneChip Mouse 430 2.0 array. Data files were
processed using the Partek software suite.
Immunohistochemistry for Fos expression
Animals’ hindpaws were peripherally stimulated two hours prior to tissue
harvest. For cold stimulations, mice were deeply anesthetized with 50 mg/kg
pentobarbital i.p. and one hindpaw dipped into an ice-water bath (~0°C) for 30 seconds
of every two minutes, repeated 15 times over 30 minutes total. For pharmacological
28
experiments, mice were briefly anesthetized with isoflurane and given a 10μl
intraplantar injection of one of the following solutions into one hindpaw using a 27
gauge needle: 2.4mg/ml icilin, 800μg/ml menthol, 500μg/ml capsaicin, 5% mustard oil,
or vehicle (80% DMSO/20% PBS, pH 7.4). Two hours after the initial stimulus,
anesthetized mice were transcardially perfused with 30mL PBS (pH 7.4) followed by
30mL ice-cold 4% paraformaldehyde in PBS (PFA). Spinal cords were removed using PBS
injected into the caudal spinal column using a dulled 18 gauge needle and lumbar
segment 4-6 tissue was post-fixed in 4% PFA for one hour at 4°C, cryoprotected in 30%
sucrose in PBS overnight at 4°C, and then frozen in OCT medium (Sakura). Twenty
micron thick sections were cut, mounted onto slides, and stored at -80°C.
Samples were thawed, washed with PBS, incubated in 1% H
2
O
2
for 30 minutes,
washed, and blocked in 10% normal goat serum (NGS) in 0.3% Triton X-100 in PBS (Tri-X)
for one hour. Samples were then incubated with 1:1250 rabbit-anti-Fos antibody (K-25,
Santa Cruz) in 1% NGS in Tri-X in a humidified chamber at room temperature overnight.
Samples were processed with the Vector Labs Elite ABC kit (rabbit) according to the
manufacturer’s instructions. Samples were incubated in 1:20 diaminobenzadine (DAB;
Invitrogen) in PBS for 15 minutes, in 1:20 DAB in 0.003% H
2
O
2
in PBS for five minutes,
then in 1:20 DAB in 0.015% H2O2 in PBS for five minutes. Finally, slides were washed,
cover-slipped, and analyzed.
Brightfield images were captured on an Olympus IX70 fluorescent microscope
with Sutter Lambda LS light source, Roper CoolSnap ES camera, and the MetaImaging
Software suite. Nuclei counts were obtained by using the ImageJ software (National
29
Institutes of Health) with the Automatic Nuclei Counter plug-in (available at
http://www.bioimage.ucsb.edu/downloads/automatic-nuclei-counter-plug-in-for-
imagej), threshold values set between 7.0-10.0 and cell diameter of 7-9 pixels, counted
by outlining the region of interest (dorsal horn) within the image. In addition, all
computer generated data were verified manually on randomly picked samples.
Immunofluorescence histochemistry
Immunohistochemistry was performed on cryopreserved sections of DRG, TG,
and spinal cords from PFA-perfused mice. A chicken anti-GFP antibody (Aves abs) was
diluted 1:500 in Tri-X with 10% normal goat serum and incubated on sections overnight
at 4°C. The following day, sections were washed and incubated for one hour at room
temperature with Alexa488 secondary antibody diluted 1:2000 in Tri-X. Digital images
were acquired on a Zeiss AxoImager Z1 with Apotome attachment.

Behavioral assays
All testing was performed during the animals’ light cycle with the experimenter
blind to the genotype. All animals were handled extensively prior to the
commencement of behavioral testing and acclimated to the testing room. Animals
were housed in their home cages immediately prior to and following testing, and were
pre-habituated to the behavioral apparatus where appropriate. For most assays,
responses were video-recorded and later quantified by an experimenter blind to
genotype. Assays were performed on different days whenever possible so as to allow
sufficient recovery time and minimize testing fatigue. Data are presented as mean ±
30
standard error. All results were analyzed using a two-tailed independent or dependent
Student’s t-test, as appropriate, taking care to correct for unequal variances when
necessary. All data are graphed as averages± standard error of the mean.
Acetone evaporation
The evaporative cooling assay was performed as follows: Mice were acclimated
for fifteen minutes in an elevated, four-place chamber with a mesh floor. A syringe with
a piece of rubber tubing attached to the end was filled with acetone and the plunger
depressed so that a small drop of acetone formed at the top of the tubing. The syringe
was raised to the mouse’s hindpaw from below, depositing the acetone drop on the
paw. Mice were tested four at a time with an inter-stimulation period of four minutes
per mouse, alternating paws between stimulations. Responses were video recorded for
later quantification by an observer blind to the experiment conditions. Behaviors were
scored according to the magnitude of the response along the following scale: 0-no
response; 1-brief lift, sniff, flick, or startle; 2-jumping, paw shaking; 3-multiple lifts, paw
lick; 4-prolonged paw lifting, licking, shaking, or jumping; 5-paw guarding. The scale was
designed so that the extreme values (0 and 5) occurred only rarely.
For behavioral testing with PBMC, the drug was administered i.p. at the doses
listed. Animals were allowed to settle for one hour following PBMC injections.
Preference and avoidance
For the two-plate temperature choice test mice were placed in a chamber
containing two identical, adjacent floor platforms with one always set to 30°C (HOME
plate) and the other (TEST plate) set to one of the following different temperatures:
31
30°C, 25°C, 20°C, 15°C, 10°C, or 5°C. Mice were always initially placed on the home
plate and were free to explore for five minutes. Each test was videotaped for later
analysis of the total time spent on each surface (temperature preference) and the
number of times an animal moved to an adjacent platform (temperature avoidance)
over the test period. Between trials, animals were rehabituated to the chamber with
each surface set to 30°C, and the control surface was alternated between trials.
Computer analysis of temperature preference was performed with the MouseChaser
program (designed especially for this project by Farhan Baluch) and matrix files of
mouse locations were compiled and converted into heatmaps using the Origin 8.1
software (OriginLab; Figure 2.5).
32
Cold plate
Animals were placed on a plate chilled to 0°C for five minutes and video-
recorded. Latency to front paw flinching and front paw shaking was later quantified.
Wet dog shakes
Mice were injected i.p. with 10µl/g body weight of a 5mg/ml solution of icilin in
1% Tween-80 and video-recorded. The number of wet dog-shakes was counted over
twenty minutes.
Chemical-induced paw flinching
Behaviors of mice after unilateral hindpaw injections of 2.4mg/ml icilin (10 µl)
were video-recorded over the first twenty minutes following injection. The number of
paw licking, shaking, lifting, and other flinching behaviors were quantified from the
Figure 2.5: MouseChaser heatmaps of wildtype mice
Using the MouseChaser program, each mouse’s progress in plate exploration in the two-
temperature choice assay can be recorded and compiled into heatmaps. Here we have
compiled data from wildtype mice (n=16) exploring assays with the home plate (left side) set to
30°C and the test plate (right side) set to temperatures ranging from 5°C to 50°C. Yellow
dashed line indicates the plate divider.
33
videos, as was the cumulative time of the twenty minute period the mice spent
attending to the injected paw.
Hot plate
Animals were placed in a chamber atop a heated plate set to either 48°C or 52°C
and responses were video recorded. Animals were removed after sixty or thirty
seconds, respectively, so as to prevent tissue damage.
von Frey paw withdrawal thresholds
Mice were acclimated to an elevated mesh platform for fifteen minutes. An
electronic von Frey apparatus (IITC) was fitted with a semiflexible tip and raised to the
plantar surface of the hindpaw. The force at which the mouse removed the paw was
measured, and average paw withdraw threshold was measured either per mouse or per
foot (depending on experiment) from five trials per foot, alternating paws, with at four
minutes between trials.
Dynamic stroke
A cotton tipped-swab was puffed up until it was approximately three times its
original volume (Garrison, Dietrich, and Stucky 2012). Animals were acclimated on an
elevated mesh platform for fifteen minutes, then the cotton swab was quickly (<1s)
swiped across the plantar surface of each hindpaw. Removal or shaking of the hindpaw
was counted as a positive response, and the rate of response was calculated over ten
trials, five per foot, alternating paws, and at least 30 seconds apart.

34
Grip strength
Animals’ grip strength limits were measured using a grip strength monitor
(Chantillon) fitted with a triangular grip wire. An animal was suspended by the tail in
front of the wire and allowed to grip it while the experimenter gently pulled the mouse
backward. The maximum force exerted by the mouse at time of release was recorded
and averaged over five trials.
Accelerating Rotarod
Naïve mice were placed on a Rotarod device (Letica Scientific Instruments) while
it was slowly turning at four rotations per minute. When animal was oriented in the
correct direction and stably maintaining its grip on the rod, the device was activated to
maximum speed (40 rotations per minute) over sixty seconds. Time to drop was
averaged over five trials.
Thermal telemeter implantation and core temperature monitoring
Mice were implanted with G2 e-mitters (Mini Mitter, Bend, OR) according to the
manufacturer’s instructions. Under sterile conditions, mice were anesthetized with 4%
isoflurane and maintained with 2% isoflurane in oxygen. The ventral surface was shaved
and sterilized and a 2cm incision was made in the skin, and a 1.5cm incision made in the
abdominal wall. A chemically-sterilized e-mitter was gently nestled amongst the small
intestines with care not to compress any vital organs and anchored to the abdominal
wall with 5-0 Vicryl sutures. The peritoneum was sutured, and the overlying skin closed
with tissue adhesive. The animals were allowed to recover for 30 minutes in a warmed
recovery cage before returning to their home cages. The animals were given 0.03mg/kg
35
buprenorphine fifteen minutes prior to surgery and again every twelve hours post-
surgery for a total of 48 hours. The animals’ health and recovery were monitored by
experimenters and USC Department of Animal Resources staff. Animals were allowed
to recover from surgery for at least one week to ensure the absence of infection and
fever.
On the day of experiments, animals were acclimated to the experiment room at
least one hour prior to the commencement of temperature monitoring. The VitalView
software package (Mini Mitter, Bend, OR) was used for automated temperature
monitoring, with the temperature recording limits set to 40-30°C and the monitoring
period set to every five minutes. Animals were provided standard mouse chow and
water ad libitum during the testing period and allowed at least two days recovery
between experiments. Baseline temperatures were calculated by averaging the
temperature readings over the thirty minutes immediately prior to injection. The
change in core temperature (ΔT) was calculated by subtracting the baseline
temperature from the observed temperature. Care was taken to perform experiments
at the same time of day so as to minimize circadian influences on temperature readings.
Icilin (Tocris Bioscience) was dissolved to a concentration of 24mg/ml in DMSO
and then diluted to 1mg/ml in 20% DMSO/80% saline (DS vehicle) and administered at
10mg/kg. A stock solution of 10mg/ml capsaicin in ethanol was diluted to 0.1mg/ml in
the same vehicle and administered at a dose of 1mg/kg. The vehicle was administered
at the same volume (10ml/kg) as both icilin and capsaicin. These solutions were
administered subcutaneously to the dorsal surface of the animal. PBMC was a provided
36
as a generous gift from Pfizer Inc. (Sandwich, Kent, U.K.) and a concentrated stock
prepared in DMSO and then suspended in a vehicle solution of 10% Solutol (Sigma-
Aldrich), 20% PEG-200 in normal saline (SPS vehicle) to a concentration of 2.5mg/ml.
PBMC and SPS vehicle solutions were administered either subcutaneously or
intraperitoneally at doses of 2, 10, or 20mg/kg, and any solutions administered i.p. were
warmed to 37°C prior to injection.
Injury models
Inflammatory injury was induced by unilateral intraplantar injection of 20µl of
complete Freund’s adjuvant (CFA). Mice were tested at two days post-injection.
The chronic constriction injury (CCI) model of neuropathic pain was induced as
follows: Under sterile conditions, mice were anesthetized with 5% isoflurane and
anesthesia maintained with 3% isoflurane in oxygen. The animal was positioned so that
the right flank was accessible and the leg supported with a roll of gauze. The flank
surface was closely shaved and sterilized and a 2cm incision was made in the skin. The
muscle was gently retracted until the sciatic nerve was revealed. Three 6-0 chromic gut
sutures were loosely tied around the nerve about 1mm apart. The muscle was replaced,
and the skin was closed with tissue adhesive. The animals were allowed to recover for
30 minutes in a warmed recovery cage before returning to their home cages. Animals
were monitored for infection and proper wound healing and were tested on day seven
post-surgery, as well as day ten or eleven for analgesia experiments.

37
Cooling analgesia
Between ten and fourteen days after CCI surgery, animals were tested for von
Frey paw withdrawal thresholds with the electronic von Frey apparatus as described
above. Animals were then placed in a shallow (~0.5cm deep) water bath held at 17°C
for five minutes. Mice were then re-tested with the electronic von Frey every five
minutes for a total of sixty minutes.
Induced pruritis
Mice were individually acclimated for one hour each day over three days in
triangular chambers made of two mirror panels and one pane of glass. On the second
day of acclimation, the animals’ cheeks were shaved. Following acclimation, mice were
acclimated again for one hour, followed by injection of 10µl of one of the following
pruritogens into one cheek: histamine (Sigma-Aldrich), 48/80 (Sigma-Aldrich),
chloroquine (Sigma-Aldrich), BAM 8-22 (Abcam), SLIGRL-NH2 (Abcam) and 3mg/ml
serotonin (Sigma-Aldrich). Responses were video recorded for later analysis, and the
utmost care was taken to not disturb the animals during this period. The number of
scratches directed at the injection site was later quantified by an observer blind to
genotype and condition. The same animal received injection into the same cheek at
intervals no shorter than six weeks to allow for any tissue damage to heal while
conserving the number of mice used in these experiments.
38
CHAPTER THREE: Innocuous Cold Sensing
Introduction
Sensing temperatures within the innocuous range may be the most important
aspect of cold sensing in regards to thermal homeostasis since cooling is the first sign
that an animal is experiencing heat loss. In humans, the innocuous range of
temperatures is generally considered between 30°C to around 15°C, below which the
temperature is reported as painful (Morin and Bushnell 1998). Cold sensation as a
whole is mediated by both lightly-myelinated Aδ fibers, which are generally considered
sensors of innocuous stimuli, as well as unmyelinated C fibers, which are considered
nociceptors (Bessou and Perl 1969). Our previous work has shown an overlap of TRPM8
with markers of both fiber types in sections of mouse sensory ganglia, thus TRPM8 is
well poised for involvement in innocuous cool sensing (Takashima et al. 2007).
To test whether TRPM8 confers the ability to sense innocuous cool temperatures
in vivo, the first step was to generate an animal in which the Trpm8 gene was deleted
(M8KO mice). This was performed by three different groups with slightly differing
approaches to behavioral assays of innocuous cold sensation. Dhaka and colleagues
measured responses to the two-temperature choice assay, the proportion of time spent
on each plate interpreted as the mouse’s ability to detect and avoid cool temperatures
(Ajay Dhaka et al. 2007). They showed that normal mice preferred the warmer
temperatures of 26°C vs. 30°C, 22°C vs. 26°C, and 18°C vs. 31°C, while M8KO mice
showed no preference for either plate during the first fifteen minutes of testing.
Colburn et al. used the same assay with the plates set at 25°C vs. 15°C and also found
39
that TRPM8-null mice spent equal amounts of time on each plate, a deficiency in
preference for the warmer temperature as compared to wildtype mice (Colburn et al.
2007). Bautista and coworkers also performed the two-temperature choice assay, but
at temperature comparisons of 30°C vs. 25°C, 20°C, and 15°C, again finding that M8KO
mice have deficits in preference for the 30°C side in all three comparisons (Bautista et al.
2007).
Additionally, Bautista et al. performed the evaporative cooling assay, finding that
M8KO mice exhibited significantly fewer paw flinches (although not zero) in response to
a small amount of acetone applied to the hindpaw. Although the Colburn group only
used this test for injury conditions (see Chapter 5), they did measure the skin
temperatures with acetone application, reporting that subcutaneous temperatures of
~16°C were reached, just within the innocuous cool range. If the results from the two-
temperature preference assays were so clear, across all three groups, then why were
the animals still showing responses (albeit reduced ones) to evaporative cooling?
Testing behavioral responses to innocuous temperatures is challenging in lieu of
verbal reporting, which is of course impossible in mice (at least for innocuous testing—
see Williams 2008 for an intriguing study on noxious responses). Noxious responses are
more readily to measureable, and a simple lack of noxious response is ambiguous as it
can signal either innocuous (i.e. lack of pain) sensing or lack of sensation. Thus it has
been useful to measure innocuous responses in ways that place them on a continuum
ranging from none (null) to noxious or extreme aversion.

40
Temperature preference and cold avoidance
For the two temperature choice assay, we were concerned that the
quantification scheme of measuring the proportion of time a mouse spent on each
plate, or preference, was potentially confounded by the competing drives of cold
avoidance and heat seeking (McCoy, Knowlton, and McKemy 2011). That is, while mice
prefer a temperature closer to their thermoneutral temperature over temperatures as
innocuous as room temperature (~20°C), their behavior may reflect both the
thermoregulatory drive to stay warm and expend as little energy as possible to maintain
normal body temperatures as well as the purely sensory-driven experience of avoiding
cool temperatures (Figure 3.1). As 20°C may pose only a mild thermoregulatory threat,
this may not be as important a point in regards to innocuous cool temperatures as with
noxious cold temperatures, as will be discussed in Chapter 4. However, because of this
potential interference of thermoregulatory drive in the temperature preference
quantification scheme, we devised an additional method of quantification, namely the
number of times the animal chose to change plates, or midline crossings, as a measure
of temperature avoidance. In terms of operant learning, if the animal is motivated to
avoid cool temperatures and depends on sensory experience to do so, then the number

41

Figure 3.1: Temperature preference vs. Cold Avoidance

A) For preference, the proportion of the testing period that the mice spend on each plate is
measured, with a 50% reading for any given plate indicating no preference (e.g. when both
plates are set to 30°C). When the test plate is 30 °C both wildtype (WT) and M8KO (KO) mice
show no preference, or 50% is spent on each plate. As the test plate is cooled, WT mice show a
preference for the 30°C plate, while KO mice show no preference until noxious cold
temperatures. This can be explained by the presence of two partially overlapping drives: a
discomfort avoidance drive and a second drive to maintain proper bodily temperatures by
reducing the thermoregulatory burden. At noxious cold temperatures, the thermoregulatory
drive is enough to affect behavior. B) For avoidance, the number of plate crosses are
quantified, since these are dependent only on whether the mouse sensed that the cooler plate
is not an ideal place to be (i.e. whether the cold was detected by peripheral thermosensors).
While WT mice stop crossing the plate midline as the test plate is cooled, KO mice continue to
cross, indicating that the initial sensory signal is defective.
42
of times the animal chooses to sample the cool surface is a measure of whether or not
the stimulus was detected and experienced as repulsive.
At test temperatures ranging from 30°C to 15°C (vs. the invariable 30°C ‘home’
plate), wildtype mice (n=16) show an increasing preference for the home plate,
spending 46.7±5.9% of the time on the test plate when it was set to 30°C, and
decreasing to 23.9±4.6% at 25°C, 2.3±0.8% at 20°C, and 1.5±0.7% at 15°C (Figure 3.2).
Similarly, the rate of plate crossing decreased with temperature, with wildtype mice
crossing 70.4±17.9% at 25°C, 10.4±2.9% at 20°C, and 6.5±3.8% at 15°C, as compared to
baseline crossing rates when both plates were set to 30°C (i.e. individual crossing rate at
Figure 3.2: Temperature preference in the innocuous range

Wildtype (WT) mice (n=16) spend 46.7±5.9% of the testing time on the test plate when it is set
to 30°C, which decreases to 23.9±4.6% at 25°C, 2.3±0.8% at 20°C, and 1.5±0.7% at 15°C. M8KO
mice (n=8) spend 55.2±5.0% at 30°C, 51.6±3.3% at 25°C, 45.9±7.5% at 20°C, and 20.3±3.1% at
15°C, all of which are significantly more than WT mice. Ablated mice (n=12) spend 51.0±2.8%
at 30°C, 57.7±3.3% at 25°C, 56.1±5.2% at 20°C, and 44.3±4.9% at 15°C, all of which were
significantly different from wildtype. Only at 15°C were there any differences between M8KO
and ablated mice (p=0.000652). Student’s t-test vs. WT: **p<0.01, ***p<0.001, n.s. p>0.05.
43
30°C vs. 30°C set to 100% to control for individual differences in general activity; Figure
3.3). In comparison, TRPM8-null mice (n=8) showed no significant preference for the
30°C home plate over the variable temperature plate until the latter reached 15°C, with
mice spending 55.2±5.0%, 51.6±3.3%, 45.9±7.5%, and 20.3±3.1% of the time on test
plates set to 30°C, 25°C, 20°C, and 15°C, respectively (Figure 3.2). Similarly, TRPM8-null
mice crossed 112.9±19.3%, 108.5±23.5%, and 46.8±7.3% at 25°C, 20°C, and 15°C,
respectively, as compared to crossing at 30°C (Figure 3.3). Analysis of heat maps at
Figure 3.3: Cold avoidance in the innocuous range

Individual crossing rates were normalized to when the test plate was set to 30°C. Wildtype (WT)
mice (n=16) crossed 70.4±17.9% at 25°C, 10.4±2.9% at 20°C, and 6.5±3.8% at 15°C as compared
to 30°C. M8KO mice (n=8) crossed 112.9±19.3% at 25°C, 108.5±23.5% at 20°C, and 46.8±7.3%
at 15°C compared to 30°C. The crossing rates at 20°C and 25°C were both significantly different
from WT. Ablated mice (n=12) crossed 109.6±8.6% at 25°C, 109.9±15.8% at 20°C, and
94.3±10.3% at 15°C compared to 30°C. Again, the crossing rates at 20°C and 25°C were both
significantly different from WT, although at 15°C the crossing rates of M8KO and ablated mice
differed significantly (p= 0.003122). Error bars omitted for clarity. Student’s t-test vs. WT:
**p<0.01, ***p<0.001, n.s. p>0.05.
44
temperature comparisons of 30°C vs. 30°C, 20°C and 15°C also demonstrate these
relationships (Figure 3.4).
These results clearly demonstrate the TRPM8 dependence of innocuous cool
sensation in both measures of temperature preference and avoidance at temperatures
of 25°C and 20°C since TRPM8-null mice do not display the normal increase in
preference for the 30°C plate and decrease in crossing rate exhibited by wildtype mice
Figure 3.4: MouseChaser heatmaps of innocuous cold

Visualization of mouse exploration behavior in the two-temperature choice assay with the
home plate (left) set to 30 °C and the test plate set to either 20°C or 15°C. Note the differences
in wildtype (n=16), M8KO (n=8), and ablated (n=12) mice at 20°C and 15°C. Yelow dotted line
indicates the plate divider.

45
at these temperatures. However, although at 15°C wildtype and TRPM8-null mice show
significantly different behaviors in both measures, the loss of TRPM8 does not appear to
account for all aspects of cold sensing at this threshold temperature since the mice
exhibit a moderate amount of preference for the home plate and a reduction in plate
crossing at this temperature.
This can possibly be explained by two scenarios: First, it is possible that 15°C
marks the transition from innocuous cool to noxious cold and TRPM8 is not involved in
noxious cold sensing, thus the mixed results. Indeed, when the original reports on
TRPM8-null mice were published, the field largely assumed this to be the case,
suggesting another TRP channel—TRPA1—as the likely noxious cold sensor (see
Appendix 2). This seemed to us an unlikely explanation since the TRPM8-null mice did
not show completely normal behaviors and therefore implicated TRPM8 at least
partially, and that other molecules such as the TREK1 and TRAAK potassium channels
could also be involved in noxious cold sensing. The second possible explanation is that
avoidance of noxious cold temperatures (even transition temperatures) is a biological
imperative and thus the TRPM8-null mice somehow compensated over the course of
development for the loss of the primary molecular cold sensor.
While these two explanations are not necessarily mutually exclusive, it remained
unknown whether the residual cold sensing resided within TRPM8-expressing neurons
or whether there is a second population of cold-responsive sensory neurons. To test
this, we engineered a mouse to express simian diphtheria toxin receptor in TRPM8-
expressing cells, which gives us the ability to eliminate TRPM8-expressing neurons with
46
a timecourse dependent on when we chose to give the animals diphtheria toxin
(“ablated” animals; see Chapter 2). This allows us to examine the role of these cells
independent of developmental compensation.
When we tested ablated animals with the two-temperature choice assay, we
found results mostly similar to M8KO mice. When the test plate was set to 30°C,
ablated mice (n=12) spent 51.0±2.8% of the time on the test plate, which did not change
significantly as the plate temperature decreased to 20°C (57.7±3.3% at 25°C and
56.1±5.2% at 20°C; Figure 3.2). Where the two lines diverged was at 15°C, as the
ablated mice spent significantly more time on the test plate at 44.3±4.9% (p=0.000652).
Similarly, ablated mice crossed at rates of 109.6±8.6% at 25°C and 109.9±15.8% at 20°C,
while at 15°C their rate was 94.3±10.3% (Figure 3.3). Again, the ablated mice were
significantly different from M8KO mice when the test plate was set to 15°C
(p=0.003122).
These data support the first explanation presented above—that TRPM8 does not
account for all aspects of cold sensing—but that the cellular location of such
mechanisms resides within the TRPM8-expressing cells themselves. A previous study
looking at thresholds of cold-sensing cells in vitro suggested that cells can be
categorized as low or high threshold, with the former group responding to smaller
decreases in temperature and relying heavily on TRPM8 for signal generation, and the
latter group responding only to large decreases in temperature and relies less on TRPM8
for signal generation (Madrid et al. 2009). Our data fit into this model in that the low
threshold population would be completely impaired at 25°C and 20°C in the M8KO mice,
47
but the high threshold population would only be mildly impaired in this line. Since all
cells expressing any amount of TRPM8 are killed in the ablated line, both populations
are impaired in these mice.

Evaporative cooling
We also tested wildtype, M8KO, and ablated mice using the acetone evaporation
assay, for which we have developed a scoring system to quantify changes in quality of
responses to acetone, as described in the Chapter 2. Using this scoring system, we have
found that while wildtype mice (n=8) exhibit an average score of 2.1±0.1, which
corresponds to jumping and/or paw shaking, and M8KO mice (n=8) exhibit average
scores of only 1.1±0.1, which corresponds to brief startle or sniff, behaviors which could
conceivably be attributed to mechanical or olfactory stimulation (Figure 3.5). Similarly,
ablated mice (n=12) exhibit scores of 1.2±0.1, which is not significantly different from
M8KO mice.
48
We also tested wildtype mice before and after administration of the TRPM8
antagonist PBMC. At baseline, wildtype mice (n=17) exhibited an acetone score of
2.2±0.1 (Figure 3.6). PBMC given at a dose of 10mg/kg significantly decreased this score
(n=9) to 1.8±0.1 (p=0.002244). Doubling the dose to 20mg/kg further decreased the
score (n=9) to 1.4±0.1 (p=<.0001 vs. baseline; p=0.026358 vs. 10mg/kg). These values
were not quite equivalent to M8KO mice, but we were unable to further increase the
dose due to dramatic effects on thermoregulation (see Chapter 6).
Figure 3.5: Evaporative cooling

Wildtype mice (n=8) exhibit an average acetone score of 2.1±0.1, while M8KO (n=8) and
ablated mice (n=12) exhibit average acetone scores of 1.1±0.1 and 1.2±0.1, respectively.
Student’s t-test vs. WT: ***p<0.001
49
Conclusion
The results from the temperature preference, cold avoidance, and evaporative
acetone assays agree with previously published results concluding that TPRM8 confers
innocuous cold sensing in vivo. It is interesting to note that our tests of temperature
preference and avoidance at 15°C yielded different results in M8KO and ablated mice,
while acetone evaporation, which is estimated to reach 16°C—only one degree
warmer—yielded identical results between the two lines. The data from the first two
assays also suggest that 15°C may be an important transition temperature wherein the
activation of TRPM8 neurons switches from the low-threshold cool sensors to the high-
threshold cold sensors.
Figure 3.6: Evaporative cooling with PBMC

Wildtype mice exhibit an average baseline acetone score of 2.2±0.1 (n=17), and a 10mg/kg
dose of PBMC reduces that score to 1.8±0.1 (n=9), while a 20mg/kg dose reduces it to 1.4±0.1
(n=9). Both of these doses were significant reductions, and were significantly different from
each other (p=0.026358). Student’s t-test vs. WT: **p<0.01, ***p<0.001
50
It is also interesting to note that treatment with PBMC resulted in graded
decreases in acetone responses. The acetone scores achieved by both of these doses
flanked the average score TRPM8-heterozygous mice (1.6±0.1; see Appendix 1),
suggesting partial blockade of TPRM8 on the whole-animal scale. We would speculate
that further increasing the dose to 30mg/kg would lead to global TRPM8 blockade and
acetone scores equivalent to M8KO mice. However, as the thermoregulatory side
effects of this drug are profound (Chapter 6), it is impossible to test this at present.
Future studies investigating the cellular location of TRPM8-based thermoregulation and
its overlap with peripheral cold sensing, and the development of TRPM8 antagonists
which could target the latter while sparing the former would further enlighten this issue.
On the whole, however, it is clear, from these data and others’, that TRPM8 confers
innocuous cold sensing in vivo.

51
CHAPTER FOUR: Noxious Cold Sensing
Introduction
While the role of TRPM8 in innocuous cold sensing is uncontested, its role in
noxious cold sensing is still debated (Daniels and McKemy 2007). The reports from the
three different groups on the generation of TPRM8-null mice showed different results in
similar behavioral assays of noxious cold. For example, in the cold plate assay, one
group found no differences between M8KO mice and wildtype mice on a -1°C plate,
while another observed an increase in response latency in M8KO mice on a 0°C plate,
while the third group reported no differences between M8KO and control mice on cold
plates set to 10°C, 0°C, -5°C, or -10°C (Bautista et al. 2007; Colburn et al. 2007; Ajay
Dhaka et al. 2007). Two groups found that icilin-enhanced paw flinching on a cold plate
as well as wet dog shaking was dependent on TRPM8, while the third group did not test
behavioral responses to chemical agonists. In the two-temperature choice assay, results
were mixed in that one group found that M8KO mice showed no preference for a room
temperature plate over one set at 5°C, while another group found that at temperature
comparisons of 30°C vs. 10°C or 5°C, M8KO mice preferred the 30°C plate only slightly
less than wildtype mice. To affirm that 5°C is indeed a noxious stimulus, the Colburn
group challenged mice with a temperature comparison of 45°C vs. 5°C, finding that
while M8KO mice showed only a slight preference for the 5°C side, control animals
preferred to brave the heat of 45°C over the chill of 5°C.
Despite these mixed results, it seems likely that TRPM8 is involved in noxious
cold sensing due to its in vitro properties as well as the insulating properties of the skin
52
in which the terminals of TRPM8-expressing neurons are embedded. First, the
temperature activation threshold of the channel was originally reported as ~25°C when
heterologously expressed, with a current saturation temperature of ~8°C (McKemy,
Neuhausser, and Julius 2002). This means that while the channel can pass current at
warmer temperatures, cation conductance increases with colder temperatures.
Furthermore, in vivo measurements of temperature thresholds of TRPM8-expressing
neurons show a range of activation thresholds ranging from ~29-2°C for C-fibers and
~12-4°C for Aδ-fibers, encompassing both the innocuous and noxious cold temperature
ranges (Bautista et al. 2007; Madrid et al. 2009). Finally, measurements of
intracutaneous skin temperature of humans upon cooling with topically applied
thermodes indicates that even with a thermode set to -5°C, it takes the skin 40 seconds
to cool to only ~17°C, yet the threshold of verbal reporting of cold-induced pain is
reached in ten seconds, when the skin is ~25°C, within the activation range of TRPM8
(Morin and Bushnell 1998).
With these data in mind, we set out to determine whether TRPM8 is involved in
detection of noxious cold in vivo. We used neural activation as well as a variety of
behavioral assays to examine this question, finding that, in agreement with the
reasoning outlined above, TRPM8 is indeed involved in noxious cold detection. The
variability seen in previous studies of behavior in M8KO mice is likely due to
irregularities in assays and quantification schemes, highlighting the need for more
rigorous testing paradigms (Story et. al 2003).

53
Icilin-induced wet dog shaking
The one behavioral assay that seems to not be disputed in the literature as being
dependent on TRPM8 is the icilin-induced wet dog shaking assay, therefore we
determined whether we could replicate this finding using both M8KO and ablated mice.
After i.p. injection of 50mg/kg icilin, wildtype animals (n=8) exhibited 18.8±3.9 shakes
over twenty minutes (Figure 4.1). M8KO mice (n=8), in contrast, only exhibited 0.9±0.5
shakes over the same time period, indicating that this is a TRPM8-dependent
phenomenon. Similarly, ablated animals (n=12) exhibited 2.5±0.8 shakes, which was not
significantly different from M8KO mice (p=0.098965).

Figure 4.1: Wet dog shakes

Administration of 50mg/kg icilin i.p. results in wet dog shaking behavior. Wildtype mice (n=8)
exhibited 18.8±3.9 shakes over twenty minutes, while M8KO mice (n=8) exhibited 0.9±0.5
shakes and ablated mice exhibited 2.5±0.8 shakes. Student’s t-test vs. WT: **p<0.01.

54
Icilin-induced paw flinching
Since injection of icilin into the intraperitoneal cavity stimulates sensory neurons
of the nodose and visceral ganglia, we next tested whether stimulation of peripheral
sensory ganglia with icilin is TRPM8-dependent. To measure this, we quantified the
number of nocifensive responses (paw flinches) as well as the length of time the mice
spent attending to the paw over twenty minutes following intraplantar injection of 24µg
of icilin, finding that wildtype mice (n=6) displayed 40.8±8.1 flinches and spent
122.3±26.4s attending to the injected paw (Figure 4.2). M8KO mice (n=5), however,
displayed only 5.6±1.9 flinches and spent only 20.7±12.8s attending to the paw. This
confirms the wet-dog shaking data suggesting that icilin responses are TRPM8-
dependent and suggests that peripheral stimulation with this amount of icilin is
sufficient to elicit nocifensive responses.
55

Figure 4.2: Icilin-induced nocifensive responses

Injection of 24µg of icilin into the hindpaw leads to nocifensive flinching. Wildtype mice (n=6)
displayed 40.8±8.1 flinches (A) and spent 122.3±26.4s attending to the injected paw (B), while
M8KO mice (n=5) displayed only 5.6±1.9 flinches and spent only 20.7±12.8s attending to the
paw. Student’s t-test vs. WT: **p<0.01.
56
Neural activation
Stimulation of the paw with either intense temperatures or thermomimetic
compounds leads to expression of the transcription factor c-fos (Fos) in neurons of the
spinal cord dorsal horn, the cells with which primary sensory neurons make their first
synapse (Hunt, Pini, and Evan 1987; Coggeshall 2005). Here we tested whether the
TRPM8 activators cold (~0°C), menthol, and icilin could induce Fos activation in the
spinal cord dorsal horn, and whether this activation is TPRM8-dependent.
First, we characterized the amount of Fos activation by these stimuli in wildtype
mice in comparison to capsaicin and mustard oil (which stimulate TRPV1- and TRPA1-
expressing afferents, respectively; Figure 4.3). We found that stimulation with 0°C (n=4)
resulted in 43.3±2.0 Fos-positive nuclei in the dorsal horn ipsilateral to the stimulation
site, while only 17.0±3.0 neurons expressed Fos on the contralateral side (Figure 4.4).
Stimulation with menthol (n=7) led to 47.0±1.5 Fos-positive nuclei on the ipsilateral side,
and only 14.6±2.0 on the contralateral side. Stimulation with icilin (n=12) resulted in
47.1±2.6 Fos-positive cells ipsilateral, and 10.9±1.2 contralateral. Capsaicin injections
(n=8) led to 50.5±4.5 Fos-positive neurons ipsilateral and 11.4±1.2 contralateral, while
mustard oil injections (n=4) led to 51.8±1.7 Fos-positive neurons ipsilateral and 9.3±2.4
neurons contralateral. Injections with vehicle (n=4) led to Fos expression in 8.5±1.8
neurons ipsilateral and 4.5±1.3 contralateral, indicating that injections alone can induce
a modest amount of Fos expression, but the majority of the Fos expression seen with
the different chemicals is specific. Taken together, these data indicate that a significant
57

Figure 4.3: Fos induction with cold and chemicals

Stimulation of the hindpaw with cold (0°C) or chemicals such as menthol, icilin, capsaicin, or
mustard oil lead to the induction of Fos expression in the ipsilateral spinal cord dorsal horn,
while the contralateral side expresses very little Fos. Scale bar=50µm.
58
amount of Fos induction in the spinal cord dorsal horn occurs with peripheral
stimulation with TRPM8 agonists, comparable to stimulation with TRPV1 or TRPA1
agonists.
Next, we determined whether the Fos expression induced by cold, menthol, and
icilin was TPRM8-dependent. We repeated the stimulations with M8KO mice, finding
that for 0°C (n=3), 18.3±0.3 neurons expressed Fos on the ipsilateral side, and only
Figure 4.4: Quantification of Fos induction with cold and chemicals

Different stimuli lead to induction of Fos in the ipsilateral spinal cord dorsal horn of wildtype
mice. The number of neurons expressing Fos with cold (n=4) was 43.3±2.0 ipsilateral (ipsi.)
and 17.0±3.0 contralateral (contra.). Menthol (n=7) led to Fos induction in 47.0±1.5 ipsi. and
14.6±2.0 contra. Icilin (n=12) induced 47.1±2.6 neurons to express Fos on the ipsi. side and
10.9±1.2 on the contra. side. Capsaicin (n=8) induced Fos expression in 50.5±4.5 neurons ipsi.
and 11.4±1.2 contra. Mustard oil (n=4) induced Fos in 51.8±1.7 cells ipsi. and 9.3±2.4 cell
contra. Vehicle injections (n=4) did not induce significant amounts of Fos expression,
however 8.5±1.8 neurons showed Fos expression ipsi. and 4.5±1.3 contra., indicating that
some baseline Fos expression is present in the mice. Student’s t-test vs. contra.: ***p<0.001,
n.s. p>0.05.
59

Figure 4.5: TRPM8-dependent Fos expression

A) In M8KO mice, cold (n=3) induced Fos expression in 18.3±0.3 neurons ipsi. and 8.0±1.0
contra. Menthol (n=3) induced Fos expression in 14.3±1.2 cells ipsi. and 8.0±1.2 cells contra.
Icilin (n=5) induced expression in 9.8±2.1 cells ipsi. and 4.6±1.7 cells contra. Although the
ipsilateral induction of Fos was significantly higher than contralateral for cold and menthol,
these values were dramatically reduced when compared to ipsilateral values from wildtype
mice (B). Capsaicin injections induced a normal amount of Fos in M8KO mice (47.7±1.3 ipsi.).
Student’s t-test vs. contra in A, vs. WT in B.: *p<0.05, **p<0.01, ***p<0.001, n.s. p>0.05.
60
8.0±1.0 contralateral (Figure 4.5A). Menthol (n=3) induced Fos expression in 14.3±1.2
cells ipsilateral and 8.0±1.2 cells contralateral, while icilin (n=5) induced expression in
9.8±2.1 cells ipsilateral and 4.6±1.7 cells contralateral. While the ipsilateral induction of
Fos was significant for both cold and menthol as compared to contralateral values, Fos
induction by all three stimuli was dramatically reduced in M8KO mice as compared to
wildtype animals (Figure 4.5B). This was specific to TRPM8 agonists, however, since
injection of capsaicin led to similar levels of ipsilateral Fos expression in both mouse
lines. Together these data suggest that propagation of the neural signal from the
periphery to the central nervous system in response to noxious cold and cold-mimetic
chemicals depends on the expression of TRPM8 in peripheral sensory neurons.

Temperature preference and cold avoidance
Next, we tested wildtype (n=16), M8KO (n=8), and ablated (n=12) animals using
the two-temperature choice assay with the home plate set to 30°C and the test plate set
to either 10°C or 5°C. When the test plate was 10°C, wildtype mice spent only 2.2±2.2%
of the time on the test plate, and did not spend any time on the test plate when it was
set to 5°C (Figure 4.6). M8KO mice spent 9.4±3.8% of the time on the test plate when it
was set to 10°C, which did not reach statistical significance (p=0.088894), however they
61
did spend significantly more time on the 5°C plate at 1.4±0.3%. In contrast, ablated
mice spent 46.1±6.2% of the time on the test plate when it was set to 10°C, and
24.6±4.4% on the test plate when it was set to 5°C, both of which were significantly
different from M8KO mice (p<0.001).
Quantifying the data in terms of cold avoidance yielded similar results. At 10°C,
wildtype mice crossed only 1.0±1.0% of the crossing rate when both plates were set at
30°C, and did not cross at all when the test plate was set to 5°C (Figure 4.7). M8KO mice
Figure 4.6: Temperature preference

When the test plate was 10°C, wildtype mice (n=16) spent only 2.2±2.2% of the time on the test
plate, and did not spend any time on the test plate when it was set to 5°C. M8KO mice (n=8)
spent 9.4±3.8% of the time on the test plate when it was set to 10°C, which did not reach
statistical significance (p=0.088894), however they did spend significantly more time on the 5°C
plate at 1.4±0.3%. In contrast, ablated mice (n=12) spent 46.1±6.2% of the time on the test
plate when it was set to 10°C, and 24.6±4.4% on the test plate when it was set to 5°C, both of
which were significantly different from M8KO mice (p<0.001).Student’s t-test vs. WT: **p<0.01,
***p<0.001, n.s. p>0.05.
62
crossed at a significantly higher rate than wildtype mice at both temperatures,
18.0±5.4% at 10°C and 13.6±3.2% at 5°C. Once again, ablated mice were significantly
different from M8KO mice at both temperatures (p<0.001), with crossing rates of
86.5±13.0% at 10°C and 62.9±9.2% at 5°C. Data can also be visualized using heatmaps,
with exploration behavior with the test plate set to 45°C (noxious heat) for comparison
(Figure 4.8). Together these results support the conclusion that TRPM8 is involved in
noxious cold avoidance, although the TRPM8-expressing cells themselves generate a
signal sufficient to effect some measure of cold avoidance in the absence of TRPM8.
Figure 4.7: Noxious cold avoidance

When the test plate was set to 10°C , wildtype mice (n=16) crossed at 1.0±1.0% of the crossing
rate when both plates were set at 30°C, and did not cross at all when the test plate was set to
5°C. M8KO mice (n=8) crossed at a rate of 18.0±5.4% at 10°C and 13.6±3.2% at 5°C. Ablated
mice (n=12) crossed at a rate of of 86.5±13.0% at 10°C and 62.9±9.2% at 5°C, which were both
significantly different from M8KO mice (p<0.001). Error bars omitted for clarity. Student’s t-
test vs. WT: *p<0.05, **p<0.01, ***p<0.001.
63
Cold plate
The final assay of noxious cold we tested the mice with was the 0°C cold plate
assay. There is a great amount of variation in the literature regarding how this test
should be conducted, which has led to a widely varying range of results (Story and
Gereau 2006; Bautista et al. 2007; Colburn et al. 2007; Ajay Dhaka et al. 2007).
Therefore we used a long (five minute) cutoff time before removing the animals from
Figure 4.8: MouseChaser heatmaps of noxious temperatures

Visualization of mouse exploration behavior in the two-temperature choice assay with the
home plate (left) set to 30 °C and the test plate set to either 5°C or 45°C. Note the differences
in wildtype (n=16), M8KO (n=8), and ablated (n=12) mice at 5°C, as well as the differences
between noxious cold and noxious heat. Yelow dotted line indicates the plate divider.

64
the plate and quantified all behaviors observed using wildtype mice. Most studies
report hindpaw flinching or lifting and escape (jumping) behaviors as a measure of
noxious cold responses, however these behaviors were rarely observed for purebred
C57Bl/6 mice previously acclimated to the testing chamber with the plate turned off
(data not shown). However, we did observe two reliable, robust behaviors exhibited by
every wildtype mouse tested: front paw flinching or shivering-like behavior followed by
licking of the front paws (Figure 4.9). Wildtype mice (n=8) exhibited the front paw
flinching with a latency of 6.8±0.9s and front paw licking with a latency of 15.6±2.5s.
These behaviors were TRPM8-dependent, as M8KO mice (n=10) exhibited latencies to
flinch of 58.0±7.9s and to lick of 72.9±15.4s. Ablated mice (n=10) were similar to M8KO
mice in that the latency to flinch was 69.8±15.0s and the latency to lick was 98.5±17.1s.
It is interesting to note that these response latencies approach or exceed the 60 second
cutoff time researchers typically employ so as to avoid any confounding effects of tissue
damage.
65
Conclusion
These studies of neural activation and behavioral responses to TRPM8 agonists
show that TRPM8 is clearly involved in the detection of noxious cold in vivo. One
interesting observation from the studies of ablated mice is that, while M8KO-mice
display varying degrees of deficiencies in different behavioral assays as compared to
wildtype mice, these deficiencies can be enhanced with the removal of the cells which
normally express the channel. It has been suggested that a cold receptor is a type of
neuron, not simply a protein, and these results may support this idea, although it is clear
that the type of neuron in question probably expresses TRPM8 (Reid 2005). Future
studies repeating the induction of Fos in the spinal cord dorsal horn using ablated mice
Figure 4.9: Cold plate

Wildtype mice (n=8) displayed stereotyped front paw flinching and licking behaviors in
response to a 0°C cold plate, with response latencies of 6.8±0.9s for flinching and 15.6±2.5s for
licking. M8KO mice (n=10) exhibited latencies to flinch of 58.0±7.9s and to lick of 72.9±15.4s.
Ablated mice (n=10) exhibited an average latency to flinch of 69.8±15.0s and latency to lick of
98.5±17.1s. Student’s t-test vs. WT: *p<0.05, **p<0.01.
66
may explain the small but significant amount of Fos induced with 0°C in M8KO mice
(Figure 4.5A). Additionally, as innocuous and noxious cold sensing is thought to result
from the activation of different populations of cells with different temperature
thresholds, future studies into the identification of different neurochemical markers of
these populations would be useful. These markers could then be used to genetically
target and ablated specific subsets of cold sensing neurons, allowing us to investigate
the role of TRPM8 in noxious cold sensing in the context of normal innocuous sensing.
Additionally, these markers could be used as potential drug targets to help alleviate cold
hypersensitivity that is a common symptom of chronic pain patients, a subject discussed
in the following chapter.

67
CHAPTER FIVE: Cold Hypersensitivity with Injury
Introduction
Cold hypersensitivity encompasses both allodynia and hyperalgesia. Allodynia is
the abnormal sensation of pain in response to a normally innocuous stimulus, while
hyperalgesia is an exaggerated noxious response in response to a normally noxious
stimulus (Figure 5.1). In addition to normal cold sensitivity discussed in Chapters 3 and
4, TRPM8 has also been implicated in the cold hypersensitivity that is a distressing
symptom of inflammatory and neuropathic pain conditions (Colburn et al. 2007).
Inflammatory pain can be experimentally induced via the injection of complete Freund’s
adjuvant (CFA) into the plantar surface of the mouse hindpaw. Using this model, it was
shown that M8KO mice do not develop cold hypersensitivity as assessed with the
acetone evaporation test as compared to wildtype mice (Colburn et al. 2007).
Additionally, M8KO mice also failed to exhibit hypersensitivity to the acetone test in the
chronic constriction injury (CCI) model of neuropathic pain. Given that TRPM8 is
suggested to be involved in these two types of chronic pain, we asked whether TRPM8-
expressing cells are the cellular location of symptoms of cold hypersensitivity and
whether the TRPM8 antagonist PBMC could be used to treat such symptoms using both
CFA and CCI pain models.
68

TRPM8-expressing cells and inflammatory injury
First, to confirm that TRPM8 is indeed required for inflammatory cold
hypersensitivity we used unilateral intraplantar injection CFA into the hindpaws of
wildtype and M8KO mice. Wildtype mice (n=10) exhibited an increased acetone score
of 3.2±0.1 ipsilateral to the injection two days post-injection as compared to an
ipsilateral score of 2.0±0.1 baseline (Figure 5.2). M8KO mice (n=6) exhibited a baseline
acetone score of 1.0±0.1, which significantly increased to 1.8±0.1 two days after injury,
although responses at both time points were significantly lower than controls. No
changes in contralateral scores were observed for any mice (data not shown).
Figure 5.1: Allodynia and hyperalgesia

Stimulus intensity
P
a
i
n

S
c
a
l
e
Innocuous Noxious
Pain threshold
A
B

Injury shifts responses from normal (dotted curve) to sensitized (solid curve). Innocuous stimuli
which normally evoked responses below the pain threshold (grey dashed line) now evoke pain
(arrow A), which is allodynia. Noxious stimuli which normally evoked pain now evoke a higher
pain response (arrow B), which is hyperalgesia.

69
Interestingly, ablated mice (n=10) also showed an increase in acetone score from
1.1±0.1 at baseline to 2.0±0.2 two days after injection, but, again, both responses
before and after injury were significantly lower than controls. No differences between
the three groups were observed in the development of mechanical hypersensitivity,
with both control and ablated mice (n=6 each group) showing a similar decrease in paw
withdrawal threshold (Figure 5.3). These data indicate that while native TRPM8-
expressing cells are partially responsible for the cold hypersensitivity seen with
inflammatory injury, another population of cells, independent of TRPM8, is also
involved.
Figure 5.2: Cold hypersensitivity with CFA

Wildtype mice (n=10) exhibited acetone scores of 2.0±0.1 at baseline which increased to
3.2±0.1 two days post-injury. M8KO mice (n=6) exhibited acetone scores of 1.0±0.1 at baseline
which increased to 1.8±0.1 post-injury. Ablated mice (n=10) exhibited acetone scores of
1.1±0.1 at baseline which increased to 2.0±0.2 after injury. Both M8KO and ablated mice
exhibited acetone scores which were significantly (p<0.001) reduced compared to wildtype
mice both before and after injury. Student’s t-test baseline vs. post-injury: **p<0.01,
***p<0.001.

70
PBMC attenuates inflammatory cold hypersensitivity
When 10mg/kg PBMC was administered to wildtype mice (n=8) on day two post-injury, we
observed a response score of 2.5±0.2, which was significantly lower than the vehicle control
group (Figure 5.4). The effect of PBMC wore off within 24 hours, when acetone scores in the
PBMC-treated group increased to 3.0±0.1, which was not significantly different from the 2.7±0.4
score of the vehicle control group. This suggests that acute inhibition of TRPM8 using this drug
can diminish cold hypersensitivity from CFA-induced inflammation.

TRPM8-expressing cells and neuropathic injury
Similarly, we tested wildtype, M8KO, and ablated animals for cold
hypersensitivity with the CCI model of neuropathic pain. Acetone scores peak at day
Figure 5.3: Mechanical hypersensitivity with CFA

Wildtype mice (n=6) exhibited paw withdrawal thresholds of 7.0±0.2g at baseline which
decreased to 2.7±0.2g two days post-injury. M8KO mice (n=6) exhibited paw withdrawal
thresholds of 6.8±0.2g at baseline which decreased to 2.8±0.2g post-injury. Ablated mice (n=6)
exhibited paw withdrawal thresholds of 6.9±0.3g at baseline which decreased to 3.1±0.2g after
injury. Student’s t-test baseline vs. post-injury: ***p<0.001.

71
seven using this model, therefore we tested the mice at this timepoint (Knowlton et al.
2011). Although the overall peak responses of wildtype mice were larger as compared
to CFA, again both wildtype and M8KO animals exhibited an increase in acetone scores
as compared to baseline (Figure 5.5). Control animals (n=8) had ipsilateral acetone
scores of 1.8±0.1 at baseline which increased to 4.0±0.2 after injury, while M8KO mice
(n=8) had ipsilateral acetone scores of 0.9±0.1 at baseline which increased to 1.5±0.2
after injury. Again, although the increase in M8KO scores was significant (p= 0.013864),
the peak scores of M8KO mice at day seven were lower than the scores of wildtype mice
at baseline. Again, no differences in contralateral scores (data not shown) or
Figure 5.4: PBMC treatment of CFA-induced cold hypersensitivity

Wildtype mice treated with vehicle (n=8) exhibited acetone scores of 3.5±0.3 at two days post
injury, while wildtype mice treated with 10mg/kg PBMC (n=8) exhibited acetone scores of
2.5±0.2. Both groups were equivalent 24 hours later, with the vehicle group exhibiting scores
of 2.7±0.4 and the PBMC-treated group exhibiting scores of 3.0±0.1. Student’s t-test vs.
vehicle: *p<0.05.

72
mechanical hypersensitivity (Figure 5.6) were observed between the different
genotypes.
When we tested ablated mice for cold hypersensitivity using this injury model,
we observed an interesting result. At baseline, these animals (n=8) had ipsilateral
acetone scores of 1.2±0.1 that increased to 2.5±0.2 after injury, which was a significantly
higher increase than that seen in M8KO mice (p= 0.000153; Figure 5.5). In the ablated
mice, cells which normally express TRPM8—and thus confer normal cold sensitivity as
evidenced by the lower baseline score compared to wildtype mice—are killed before
the induction of injury, thus the cold hypersensitivity observed cannot be assigned to
Figure 5.5: Cold hypersensitivity with CCI

Wildtype mice (n=8) exhibited acetone scores of 1.8±0.1 at baseline which increased to 4.0±0.2
seven days post-injury. M8KO mice (n=8) exhibited acetone scores of 0.9±0.1 at baseline which
increased to 1.5±0.2 post-injury. Ablated mice (n=8) exhibited acetone scores of 1.2±0.1 at
baseline which increased to 2.5±0.2 after injury. Both M8KO and ablated mice exhibited
acetone scores which were significantly (p<0.01) reduced compared to wildtype mice both
before and after injury, and M8KO and ablated mice differed significantly from each other
(p<0.05). Student’s t-test baseline vs. post-injury: *p<0.05, ***p<0.001.

73
this population of cells. This suggests that this symptom is due to the activation of a
non-TRPM8 population of cells. However, since this cold hypersensitivity is not
observed in M8KO-mice, this suggests that the phenomenon is indeed TRPM8-
dependent. Taken together, these results suggest ectopic expression of TRPM8 in this
model of neuropathic injury that leads to cold hypersensitivity.
However, when we screened for ectopic expression of TRPM8 in ganglia
collected from ablated mice ipsilateral to the injury site using qPCR, we were unable to
detect any TRPM8 transcripts (data not shown). One caveat to this experiment is that
samples were collected at day 14 post-injury, thus any ectopic expression may have
been extinguished by this point. Thus it may be that there is indeed ectopic expression
Figure 5.6: Mechanical hypersensitivity with CCI

Wildtype mice (n=8) exhibited paw withdrawal thresholds of 6.7±0.2g at baseline which
decreased to 3.9±0.2g seven days post-injury. M8KO mice (n=8) exhibited paw withdrawal
thresholds of 7.0±0.2g at baseline which decreased to 4.4±0.2g post-injury. Ablated mice (n=8)
exhibited paw withdrawal thresholds of 6.7±0.2g at baseline which decreased to 4.3±0.3g after
injury. Student’s t-test baseline vs. post-injury: ***p<0.001.
74
of the channel, however it is dynamic and occurs at an earlier timepoint than our
samples were taken at.

PBMC attenuates neuropathic cold hypersensitivity
We next tested whether treatment with PBMC could alleviate symptoms of cold
hypersensitivity in wildtype mice with neuropathic injury. When 10mg/kg PBMC was
administered to wildtype mice (n=4) on day seven post-injury, the behavioral response
scores dropped to 3.0±0.1, a significant decrease when compared to vehicle-treated
animals who scored 4.0±0.2 (Figure 5.7). As with the CFA model, the drug wore off
within 24 hours, when the PBMC-treated animals’ scores returned to 4.1±0.2, identical
Figure 5.7: PBMC treatment of CCI-induced cold hypersensitivity

Wildtype mice treated with vehicle (n=4) exhibited acetone scores of 4.0±0.2 at seven days
post injury, while wildtype mice treated with 10mg/kg PBMC (n=4) exhibited acetone scores of
3.0±0.1. Both groups were equivalent 24 hours later, with the vehicle group exhibiting scores
of 4.1±0.2 and the PBMC-treated group exhibiting scores of 4.1±0.2. Student’s t-test vs.
vehicle: **p<0.01.

75
to the vehicle treated animals’ scores. Thus PBMC is effective in diminishing symptoms
of cold hypersensitivity in the CCI model of neuropathic pain.

Conclusion
It would be greatly beneficial to chronic pain patients to have a drug which could
control symptoms of cold hypersensitivity; however we first must identify which
molecular targets and cells are involved in these symptoms. These and previous
experiments clearly establish that TRPM8 is involved in cold hypersensitivity with both
inflammatory and neuropathic injury (Colburn et al. 2007). PBMC is an effective drug at
reducing these symptoms in both injury models, although it does not completely
attenuate responses. We could not test higher doses due to the significant effects on
thermoregulation (see Chapter 6) which would likely complicate interpretation of these
results. However, given that the aim of a good symptom-controlling drug would be to
reduce the hypersensitivity to cold without abolishing normal thermosensation (e.g.
numbness), this may not be a completely undesirable effect. Reformulation of the drug,
if possible, may yield a compound that specifically targets sensory afferents without
having the strong thermoregulatory effect observed here. Such a drug may bring much-
needed relief to both chronic pain and chemotherapy patients experiencing symptoms
of cold hypersensitivity.
The experiments using ablated mice reveal that native TRPM8-expressing cold
sensing cells are only partially responsible for responses to evaporative cooling in both
injury models. Since there were no differences between M8KO and ablated mice,
76
another population which is TRPM8-independent is also involved in CFA-induced cold
hypersensitivity. One possible molecular candidate is the TRPA1 ion channel, however
experiments investigating the identity of this second population should be done in the
context of TRPM8-null mice (O Caspani and Heppenstall 2009; Knowlton et al. 2010).
In the case of CCI-induced neuropathic injury, it was quite interesting to observe
a difference between M8KO and ablated mice. Previous groups have investigated the
possibility that TRPM8 dysregulation is a mechanism of cold hypersensitivity after
neuropathic injury, albeit with drastically differing results (Ombretta Caspani et al. 2009;
Su et al. 2011). Since all native TRPM8 expression is eradicated prior to injury in the
ablated mice, and they exhibit cold hypersensitivity after injury which is not present in
M8KO mice, it is quite possible that this cold hypersensitivity involves transient ectopic
expression. Future investigations into the timecourse of this expression should
enlighten this issue.
77
CHAPTER SIX: Thermoregulation
Introduction
TRPM8 has recently been reported to be involved in thermoregulation, a role
that is not entirely unexpected given that other temperature sensitive ion channels,
particularly TRPV1, have also been implicated in regulating body temperature (Ruskin,
Anand, and LaHoste 2007; Tajino et al. 2007; Gavva 2008; Masamoto, Kawabata, and
Fushiki 2009; Tajino et al. 2011). For example, several studies have shown that TRPV1-
null mice display attenuated fever responses, and administration of a TRPV1 antagonist
induces thermogenesis in rats and humans (Iida et al. 2005; Montell and Caterina 2007;
Gavva 2008; Gavva et al. 2008). Our recent study examining the action of PBMC in vitro
data shows that the drug is a profoundly potent TRPM8 antagonist with sub-nanomolar
affinity (Knowlton et al. 2011). Therefore we determined if this compound is equally
effective in blocking channel function in vivo.

M8KO mice are deficient in icilin-induced hyperthermia.
The potent TRPM8 agonist icilin is well known to produce an intense behavioral
response in rodents that is manifested as shivering, “wet-dog” shaking and also results
in an increase in core body temperature in rats (Wei 1981; Ding et al. 2008; Tajino et al.
2011). To date, all known in vivo effects of icilin are dependent on TRPM8, yet genetic
evidence demonstrating that icilin-induced hyperthermia is TRPM8 dependent is yet to
be established (Ajay Dhaka et al. 2007; Knowlton et al. 2010).
78

Therefore we first examined the role of TRPM8 activation in thermoregulatory
responses by subcutaneously injecting 10mg/kg icilin into wildtype and M8KO mice
implanted with thermal telemeters. Consistent with data in rats, we observed a
pronounced hyperthermic effect of 1.6°C on average in wildtype mice (n=4), which
resolved within 90 minutes (Figure 6.1) (Ding et al. 2008). However, this hyperthermic
response was absent in M8KO mice (n=4), with only a small injection-related artifact

observed that was similar to vehicle injections (Figure 6.1). When we administered
1mg/kg capsaicin subcutaneously (s.c.) to wildtype and M8KO mice we found a
Figure 6.1: Icilin-induced hyperthermia

Injection of 10mg/kg icilin resulted in an average increase in core body temperature of 1.6°C in
WT mice (n=4) which resolved within 75 minutes as measured by thermal telemetry. This
hyperthermic response to icilin was not present in M8KO animals (n=4), which only exhibited a
mild (<0.5°C) and transient (<30 minutes) increase, similar to that observed with vehicle (see
Figure 6.3). Error bars omitted for clarity. Students’ t-test vs. WT: *p<0.05.
79
profound and transient hypothermic effect of around 4°C that was similar in both
genotypes, indicating that the M8KO mice were still able to mount a chemically-induced
thermoregulatory response (Figure 6.2). Injection of the DMSO/saline (DS) vehicle s.c.
induced only a brief increase in body temperature of around 0.5°C which peaked within
30 minutes post-injection in both genotypes (Figure 6.3).

PBMC induces hypothermia in a dose- and TRPM8-dependent manner
We next determined if antagonizing TRPM8 alters thermoregulatory responses
in a likewise, yet reversed, manner. We found that subcutaneous injections of the
required vehicle for PBMC (10% Solutol/20%PEG-200/saline; SPS) resulted in intense
grooming and scratching at the site of injection in both wildtype and M8KO mice. Since
Figure 6.2: Capsaicin-induced hypothermia

Injection of 1mg/kg capsaicin resulted in a robust hypothermic response (~4°C drop) in both
wildtype and M8KO mice (n=4 each group) which lasted approximately 75 minutes. Error bars
omitted for clarity.
80
stress is known to influence thermoregulation, we therefore switched to intraperitoneal
injections of solutions warmed to body temperature (~37°C) immediately before
injection and administered as far away from the telemeter implantation site as possible
(Nomoto et al. 2004). This approach resulted in no obvious adverse effects associated
with intraperitoneal vehicle injections (Figure 6.4).

Figure 6.3: DS vehicle control

Injection of vehicle (20% DMSO/80% saline; DS) resulted in no change in core body
temperature in either genotype (n=4 each group) beyond a small spike in body temperature
within 30 minutes of injection. Error bars omitted for clarity.
81

We tested a range of PBMC doses (2, 10, 20mg/kg), finding no effect with
2mg/kg (identical to vehicle, data not shown) and a small, but significant drop in core
body temperature in wildtype mice (n=4) with 10mg/kg, which peaked at 0.8°C below
baseline by two hours post-injection (Figure 6.5). Strikingly, at 20mg/kg, we observed a
dramatic and severe hypothermic effect in wildtype mice (n=4) of more than 6°C, with a
drop in core body temperature to below 30°C in one instance (Figure 6.6). A drop in
Figure 6.4: SPS vehicle control

Intraperitoneal injection of warmed 10% Solutol/20% PEG-200/saline (SPS) vehicle resulted in
no changes in core body temperature besides the injection spike in either wildtype or M8KO
mice (n=4 each group). Error bars omitted for clarity.
82
core body temperature of more than two degrees lasted at least four hours on average.
Importantly, M8KO mice (n=4 each dose) showed no fluctuations in core
temperature besides the transient injection artifact at all doses (Figures 6.4-6). These
data show that blockade of TRPM8 activity at high PBMC doses significantly alters
thermoregulation, providing pharmacological evidence that, like TRPV1, TRPM8 is
involved in the maintenance of core body temperature (Gavva 2008).

Figure 6.5: 10mg/kg PBMC-induced hypothermia

Injection of warmed 10mg/kg PBMC resulted in a subtle hypothermic effect (<1°C drop) in
wildtype mice which resolved within three hours of injection, whereas M8KO mice remained
unaffected. (n=4 each group). Error bars omitted for clarity. Student’s t-test vs. WT: *p<0.05.

*

83

Conclusion
Emerging evidence suggests that TRPM8 plays a role in thermoregulation, both
through the stimulation of skin afferents with chemical agonists or with cooling (Ruskin,
Anand, and LaHoste 2007; Tajino et al. 2007; Ding et al. 2008; Tajino et al. 2011). Here,
we have confirmed that icilin, a chemical TRPM8 agonist more potent than menthol, can
also induce an increase in body temperature and the effect of icilin is TRPM8-
dependent, despite reports that icilin can also activate TRPA1 in vitro (Wei and Seid
1983; Story et al. 2003; Ajay Dhaka et al. 2007; Ding et al. 2008; Knowlton et al. 2010;
Knowlton and McKemy 2011). Even though M8KO mice do not respond to icilin, these
animals retain the ability to mount a chemically-induced thermoregulatory response as
Figure 6.6: 20mg/kg PBMC-induced hypothermia

Injection of warmed 20mg/kg PBMC resulted in a profound hypothermic response (>6°C drop
within 45 minutes) in wildtype animals, which did not occur in M8KO mice (n=4 each group).
Error bars omitted for clarity. Student’s t-test vs. WT: *p<0.05.

*

84
we observed an identical effect in both wildtype and M8KO mice in response to the
TRPV1 agonist capsaicin. Therefore it appears that TRPM8-expressing afferents have
the ability to affect thermoregulatory responses, although the exact neurological
mechanism remains to be explored.
Due to this evidence and recent reports of TRPV1 antagonists having undesired
thermoregulatory effects, we were concerned that a TRPM8 antagonist would also
affect thermoregulation (Gavva 2008; Gavva et al. 2008). Indeed, when we
administered PBMC at the highest dose of 20mg/kg, we observed a profound
hypothermic effect, with one mouse reaching body temperatures below the
temperature range of the telemeter (<30°C), a core temperature classified as deep
hypothermia in humans (Polderman 2009). The pharmacokinetics of PBMC are as yet
unknown, yet the hypothermic effect observed here lasted around four hours on
average, and in thermoregulatory and behavioral experiments the effects were gone by
less than one day after administration. Interestingly, halving the dose (10mg/kg) almost
completely abolished the hypothermic response, with core body temperatures dropping
less than one degree—a surprising change in effect for such a small reduction in dose.
Indeed, while this drop in core temperature was significantly different from vehicle-
injected control animals, it was within the normal range of circadian changes in body
temperature in these mice (data not shown).
These experiments suggest that PBMC may not be a suitable drug for TRPM8
antagonism for analgesic purposes if high doses are required since it would likely lead to
hypothermic side effects. However, in the case of trauma it may in fact be desirable to
85
induce hypothermia to combat inflammation, therefore this drug may be useful for this
purpose (Polderman 2009). Nonetheless, these experiments with TRPM8 agonism and
antagonism clearly demonstrate that TRPM8-expressing cells are involved in
thermoregulation.
86
CHAPTER SEVEN: Cooling-Induced Analgesia
Introduction
Analgesia, or anti-nociception, can be achieved through a variety of
mechanisms—with varying degrees of effectiveness—and can be roughly divided into
intrinsic and extrinsic mechanisms (Gebhart 1983). The intrinsic mechanisms are what
we consider ‘top-down’ mechanisms, or signals originating from the high-level areas of
the nervous system (i.e. the brain) to inhibit or block pain signals. These intrinsic
mechanisms commonly involve tonic descending inhibition of peripheral nociceptive
signals, and include both the serotonergic system and the endorphin or opiate system,
both of which are activated in stress-induced analgesia (Jolas and Aghajanian 1997; C.
Stein and Lang 2009).
It is often when this intrinsic system fails to adequately relieve pain that we turn
to extrinsic mechanisms. Extrinsic activation of the intrinsic mechanisms, such as
injections of morphine to activate the opiate system and oral administration of selective
serotonin reuptake inhibitor drugs (SSRIs) for the serotonergic system, can lead to
effective analgesia, but this comes at a cost. Since these intrinsic systems involve
receptors widely expressed throughout not only the nervous system but the entire
body, the side effects of these drugs are often undesirable, such as addiction, tolerance,
chronic itching, and death by overdose in the case of opiates, and effects on mood in
the case of serotonin-targeting drugs (C. Stein and Zo 2009). Other extrinsic approaches
to pain relief and control involve the pharmacological efforts to silence the peripheral
neurons generating the initial pain signals, but once again come with deleterious side
87
effects. For example, anesthetic drugs such as lidocaine shut down sodium channels in
sensory neurons, but do not specifically target nociceptors, resulting in general
numbness and motor deficits (Roberson et al. 2011). Recent efforts with the generation
of analgesic drugs targeting the heat- and capsaicin-sensitive TRPV1 channel have also
encountered problems with pain specificity as studies with humans resulted in
unexpected spikes in body temperature (Gavva et al. 2008; Wong and Gavva 2009). And
finally, non-steroidal anti-inflammatory drugs (NSAIDs) target prostaglandin-induced
inflammation and activation of nociceptors, but have limited efficacy and often irritate
the gastrointestinal system when taken over extended periods of time (McCormack
1994).
Thus we turn to the other form of extrinsic analgesia, that involving the
activation of sensory processes which serve to mask or inhibit nociceptive signals
leading to the sensation of pain (Melzack and Wall 1965). The “Gate Theory” of pain
postulates that activation of certain peripheral neurons can activate interneurons in the
spinal cord which can either directly or indirectly inhibit the incoming signals of
nociceptive neurons (see Figure 7.1). The effective sensory modalities leading to this
form of analgesia include vibration (such as rubbing), warmth, and cooling (Bini et al.
1984; Hayes and Katayama 1986; Leknes et al. 2008; Wang et al. 2012). As these stimuli
are innocuous in nature, there is less risk of deleterious side effects, however the
problem of efficacy remains. Here, we will focus on cooling-induced analgesia and the
involvement of TRPM8 and TRPM8-expressing neurons.
88

Cooling-induced analgesia
The use of local cooling, such as the application of an ice pack to a muscle cramp,
burn, or mechanical injury, has been around for centuries (Evans 1981). Indeed, in the
19
th
and 20
th
centuries, the use of refrigerant sprays was indicated for a variety of
ailments from muscle spasms, migraine, menstrual cramps, to arthritis pain, although
such sprays can lead to problems with tissue freezing and toxicity (Ellis 1961; Evans
1981). The profound cooling induced by such sprays is thought to work through a
mechanism of “counter-irritation” wherein the irritation of intense cold masks the
Figure 7.1: Gate theory of pain for extrinsic analgesia by sensory activation

Based on the Gate Theory of Pain (Melzack and Wall 1965), pain signals traveling from
peripheral nociceptors to the spinal cord projection neurons can be modulated both by
interneurons and other sensory neurons in the periphery. In the case of mechanical
intervention to relieve pain (e.g. rubbing your toe after stubbing it on a table leg), analgesia
is achieved through the activation of inhibitory interneurons by mechanoreceptors, which
then in turn dampen the activation of the pain projection neurons. Furthermore, in the
activation of their own modality-specific projection neurons, mechanoreceptors may also
directly inhibit the pain pathway projection neurons. Thus the activation of the
mechanoreceptors serves to mask and dampen pain signals.
89
feeling of pain (Ellis 1961). However, such profound cooling of nerves leads to complete
conduction block of all information (i.e. numbness and motor dysfunction), thus a more
practical use of cold for analgesia involves mild cooling (Evans 1981).
In a seminal paper, Proudfoot and colleagues investigated the contribution of
TRPM8 activation to cooling-induced analgesia in rats (Proudfoot et al. 2006).
Activation of cold-sensing neurons in the skin with icilin, menthol, or cool (~17°C)
temperatures resulted in reduced behavioral responses of mechanical hypersensitivity
due to neuropathic and inflammatory injury, as well as TRPA1-agonist induced
hypersensitivity. Furthermore, intrathecal administration of antisense oligonucleotides
against TRPM8 (which lowered TRPM8 expression in sensory ganglia) abolished this
effect. Interestingly, the use of temperatures below 16°C did not lead to analgesia,
suggesting that the effect was due to the activation of cold-sensitive cells tuned to
innocuously cool temperatures (see Chapter 3). Cooling can also inhibit the first two
phases of acute pain behaviors resulting from intraplantar injection of formalin, and this
analgesic effect is abolished in TRPM8-null mice (Ajay Dhaka et al. 2007).

Cooling-induced analgesia is dependent on TRPM8 and TRPM8-expressing cells
To confirm that cooling-induced analgesia is TRPM8-dependent in mice, we
tested our ablated TRPM8
DTR-GFP
transgenic animals as well as TRPM8-null mice after the
CCI model of neuropathic pain (Chapter 5). In light of our findings that the two lines
display different levels of cold hypersensitivity after neuropathic injury (M8KO mice
showed none while ablated mice did show some hypersensitivity, albeit less than
90
wildtype mice; see Chapter 5), we wanted to test whether they were equally deficient in
cooling-induced analgesia.
To confirm that we could replicate the cooling-induced analgesia in mice that
Proudfoot et al. reported in rats, we first tested wildtype animals (Proudfoot et al.
2006). Testing occurred ten days after injury with the electronic von Frey assay. At this
timepoint, the animals were mechanically hypersensitive, with paw withdrawal latencies
of 7.7±0.5g on the contralateral side and 3.8±0.3g on the ipsilateral side to the injury
(baseline; Figure 7.2). After five minutes of paw cooling in a 17°C shallow water bath,
contralateral thresholds did not change significantly from baseline, nor did they show
significant changes through the 60 minutes following the cooling. On the other hand,
ipsilateral paw withdrawal thresholds increased to 7.2±0.3g, and remained not
significantly different from contralateral values for twenty minutes following the
cooling, at which point it gradually reverted to baseline values. This pattern of behavior
suggests that five minutes of mild cooling produces relief from mechanical
hypersensitivity that lasts for twenty minutes.
91
Next, we set out to confirm that this cooling-induced analgesia was indeed
TRPM8 dependent using M8KO mice. Similar to wildtype mice, M8KO mice (n=8)
exhibited mechanical hypersensitivity at baseline, with a mean contralateral paw
withdrawal threshold of 7.3±0.1g and a mean ipsilateral paw withdrawal threshold of
4.0±0.2g (Figure 7.3). Dissimilar to wildtype mice, however, the M8KO mice showed no
changes in ipsilateral paw withdrawal thresholds after cooling throughout the entire 60
minute testing period. This indicates that the cooling-induced analgesia seen in
wildtype mice is indeed TRPM8-dependent.
Figure 7.2: Cooling-induced analgesia in wildtype mice

Wildtype mice (n=8) exhibit mechanical hypersensitivity after CCI-induced neuropathic
injury, with baseline paw withdrawal thresholds of 7.7±0.5g on the contralateral side and
3.8±0.3g on the ipsilateral side. Immediately after five minutes of cooling at 17°C, the
ipsilateral side thresholds increased 7.2±0.3g, which was not significantly different from
contralateral thresholds. Analgesia was seen for twenty minutes following cooling, and then
ipsilateral threshold levels gradually decreased to baseline levels. Students t-test vs.
contralateral: #p>0.05.
92

Ablated mice lacking TRPM8-expressing neurons before neuropathic injury show
some degree of cold hypersensitivity, which we interpret to mean that there is some
dynamic expression of TRPM8 over the course of injury (Chapter 5). Therefore to
confirm the cellular location of TRPM8-based cooling analgesia, we also examined
ablated mice. As with wildtype and M8KO mice, ablated mice also showed mechanical
hypersensitivity after CCI at baseline, with a mean contralateral paw withdrawal
threshold of 7.5±0.3g and ipsilateral threshold of 4.2±0.3g (Figure 7.4). Similar to M8KO
mice, these animals exhibit no changes in ipsilateral paw withdrawal thresholds over the
Figure 7.3: Lack of cooling-induced analgesia in M8KO mice

M8KO mice (n=8) exhibit mechanical hypersensitivity after CCI-induced neuropathic injury,
with baseline paw withdrawal thresholds of 7.3±0.1g on the contralateral side and 4.0±0.2g
on the ipsilateral side. Immediately after five minutes of cooling at 17°C, the ipsilateral side
thresholds did not change (4.1±0.2g), which was significantly different from contralateral
thresholds. No differences between baseline and any of the post-cooling timepoints were
observed.
93
course of the experiment after cooling. This leads us to conclude that native TRPM8-
expressing cells are required for cooling-induced analgesia.

Conclusion
Taken together, these data show that TRPM8 is required for cooling-induced
analgesia against mechanical hypersensitivity induced by neuropathic injury. Moreover,
the cellular address of this process is native TRPM8-expressing cells, presumably those
tuned to temperatures within the innocuous range. Identification of the precise subset
of TRPM8-expressing cells is open for future investigation and may lead to the
identification of drug targets for the specific activation of this population for purposes of
Figure 7.4: Lack of cooling-induced analgesia in ablated mice

Ablated mice (n=8) exhibit mechanical hypersensitivity after CCI-induced neuropathic injury,
with baseline paw withdrawal thresholds of 7.5±0.3g on the contralateral side and 4.2±0.3g
on the ipsilateral side. Immediately after five minutes of cooling at 17°C, the ipsilateral side
thresholds did not change (4.3±0.3g), which was significantly different from contralateral
thresholds. No differences between baseline and any of the post-cooling timepoints were
observed.
94
controlling pain. Such a treatment has the potential to provide chronic pain patients
with much-needed relief with few side effects provided that therapies do not alter other
TRPM8-based physiological processes such as body temperature (Chapter 6).
95
CHAPTER EIGHT: Antipruritis
Introduction
Pain and itch (pruritis) are thought to be processed by the nervous system in
similar ways, although they are diametrically opposed: pain (often from scratching)
alleviates itch, while blocking pain—with drugs such as opiates or preventing
glutamatergic neurotransmission in primary nociceptors—leads to chronic itch
(Davidson and Giesler 2010; Lagerström et al. 2010; Yang Liu et al. 2010). Similar to the
concepts of allodynia and hyperalgesia (Chapter 5), alloknesis and hyperknesis result
from sensitization of itch pathways (Figure 8.1). Also similar to pain, itch can be divided
into acute and chronic itch, the latter of which can be described as inflammatory or
Figure 8.1: Alloknesis and hyperknesis
Stimulus intensity
Itch Scale
Sub-threshold Suprathreshold
Itch threshold
A
B

Disease, infection, or injury shifts responses from normal (dotted curve) to sensitized (solid
curve). Subthreshold stimuli which normally evoked responses below the itch threshold (e.g.
the feel of a wool sweater; grey dashed line) now evoke itch (arrow A), which is alloknesis.
Suprathreshold pruritic stimuli which normally evoked itch now evoke a higher itch response
(arrow B), which is hyperknesis.
96
neuropathic, as well as psychogenic (Yosipovitch, Carstens, and McGlone 2007; T. Patel
and Yosipovitch 2010; K. N. Patel and Dong 2010).
The molecular mechanisms of pruritus are currently being deciphered, yet thus
far both TRPV1 and TRPA1—nociceptive TRP channels—have been implicated in the
processing of both histaminergic and nonhistaminergic itch (B. M. Kim et al. 2004;
Imamachi et al. 2009; Wilson et al. 2011). Indeed, it appears that a number of
pruritogens of both the histaminergic and nonhistaminergic pathways rely on TRPV1
and/or TRPA1, including histamine, 48/80, cowhage, PAR activators, chloroquine, and
the endogenous pruritogen BAM8-22, with the exception of serotonin (K. N. Patel and
Dong 2010; Akiyama, Carstens, and Carstens 2010; Klein, Carstens, and Carstens 2011).
These findings are puzzling since capsaicin and heat have widely been used to treat itch,
indicating that the neural logic of itch is complex (Hercogova 2005; Yosipovitch,
Carstens, and McGlone 2007; Davidson and Giesler 2010).
The balance of itch and pain signals has been proposed to account for the same
population of neurons mediating two distinct sensations, a phenomenon called
‘specificity theory’ (K. N. Patel and Dong 2010). However, this presents a tricky situation
when designing specific antipruritic therapies since a potential side effect of activating
pain over itch is the undesirable generation of chronic pain. Current methods of
antipruritis are poorly effective largely because they rely heavily on antihistamines,
which only target a subset of the known pruritic pathways (Yosipovitch, Carstens, and
McGlone 2007). Besides antihistamines, current methods include anesthesia, which
leads to undesirable numbing, or drugs aimed at either blocking the affected pathway or
97
eradicating the causative parasite or organism (Hercogova 2005). However, these last
two strategies imply that the cause or pathway of the pathological itch are known and
can be treated pharmacologically, which is unfortunately not normally the case
(Hercogova 2005; Yosipovitch, Carstens, and McGlone 2007; Davidson and Giesler 2010;
K. N. Patel and Dong 2010).
The final method for antipruritis involves counterirritation, as mentioned above
with capsaicin and in chapter 7 for control of pain. Menthol and cooling have been used
for centuries for this purpose, although the mechanism has not yet been explained.
Since menthol and cooling are both activators of TRPM8, we investigated whether
TRPM8 affects itch.

Development of spontaneous skin lesions in ablated mice
The first sign that TRPM8-expressing neurons may influence sensations of itch
came in the form of spontaneous lesions developing in TRPM8
DTR-GFP
mice after
administration of DTx, lesions reminiscent of three other studies on itch (Figure 8.2)
(Lagerström et al. 2010; Yang Liu et al. 2010; Ross et al. 2010). Ablated mice (n=23)
developed lesions of the ears, eyes, whisker pads, neck, back, and tail, often exhibiting
more than one lesion at a time. Occasionally lesions were severe enough to warrant
euthanization to prevent suffering, but more often the lesions healed. The latency to
lesion development was as little as eight days following the first injection of toxin and
the latest observed first lesion occurred 42 days after injection, with the 50% latency to
lesion occurring on approximately day ten post-injection. Lesions developed
98
sporadically, and no heightened levels of baseline scratching were observed when
animals were observed for one hour (data not shown). Every ablated mouse used in
these studies developed at least one skin lesion, while no control or M8KO mice used in
any of these studies were observed to develop skin lesions. Care was taken to avoid
lesioned areas when performing behavioral assays.

Figure 8.2: Spontaneous lesions in ablated mice
Whisker pad Ear and eye Multiple Severe lesion

Ablated mice developed spontaneous skin lesions of the ears, eyes, whisker pads, neck, back,
and tail, often many at one time. A Kaplan-Meier survival analysis (n=23) reveals that the 50%
time for the development lesions is approximately ten days following the first injection of DTx.
99
Induced itch
Since two previous studies involving silencing glutamatergic neurotransmission
in peripheral nociceptors and one study on the removal of itch-processing interneurons
in the spinal cord reported induced hyperknesis in addition to spontaneous skin lesion
development, we next tested our ablated mice with a variety of pruritic compounds
(Lagerström et al. 2010; Yang Liu et al. 2010; Ross et al. 2010). Intradermal (i.d.)
injection of chemicals into the cheek allows for differential quantification of pain and
itch behaviors since itch involves scratching with the hindpaw and pain involves wiping
with the front paw (Shimada and LaMotte 2008).
100

Figure 8.3: Histamine hyperknesis

Intradermal injection of 100µg histamine in 10µl into the cheek resulted in 42.1±4.9 scratches
over 40 minutes in wildtype mice (n=8) and 60.1±6.6 scratches in ablated mice (n=8). Student’s
t-test vs. WT: *p<0.05.
Figure 8.4: Chloroquine hyperknesis

Intradermal injection of 200µg chloroquine in 10µl into the cheek resulted in 60.3±7.4 scratches
over 40 minutes in wildtype mice (n=7) and 108.6±9.4 scratches in ablated mice (n=7).
Student’s t-test vs. WT: **p<0.01.
101
In comparison to wildtype mice, ablated mice exhibited hyperknesis and
alloknesis to chemicals activating both histaminergic and nonhistaminergic pathways.
Intradermal injection of 100µg histamine in 10µl into the cheek resulted in 42.1±4.9
scratches over 40 minutes in wildtype mice (n=8) and 60.1±6.6 scratches in ablated mice
(n=8; Figure 8.3).
In addition to the histaminergic pathway, ablated mice also showed sensitization
with chloroquine and the endogenous pruritogen BAM8-22, both of which are
histamine-independent and mediated through members of the Mrgpr family of GPCRs
and TRPA1 (Q. Liu et al. 2009; Wilson et al. 2011). Intradermal injection of 200µg
chloroquine in 10µl into the cheek resulted in 60.3±7.4 scratches over 40 minutes in
wildtype mice (n=7) and 108.6±9.4 scratches in ablated mice (n=7; Figure 8.4). Similarly,
Figure 8.5: BAM8-22 alloknesis

Intradermal injection of 60µg BAM8-22 in 10µl into the cheek resulted in 14.4±1.7 scratches
over 40 minutes in wildtype mice (n=7) and 26.6±2.7 scratches in ablated mice (n=7). Student’s
t-test vs. WT: **p<0.01.
102
ablated mice displayed alloknesis with a low dose of BAM8-22, displaying 26.6±2.7
scratches over 40 minutes (n=7), while wildtype mice only displayed 14.4±1.7 scratches
(n=7; Figure 8.5).
The final pruritic compound tested was serotonin, which induces itch via binding
to a serotonin receptor, likely types 1-3, although precisely which receptor is unclear
(Akiyama, Carstens, and Carstens 2010; Jeffry, Kim, and Chen 2011). When injected into
the cheek of wildtype animals, 30µg serotonin induces wildtype mice (n=14) to scratch
42.1±6.8 times in 40 minutes (Figure 8.6). Interestingly, serotonin failed to elicit
significant scratching in ablated mice (n=12), with only 4.8±1.4 scratches observed over
40 minutes. Examination of our microarray data reveals ~1.3-fold loss of type 1 and
Figure 8.6: Loss of serotonin responses

Intradermal injection of 30µg serotonin in 10µl into the cheek results in 42.1±6.8 scratches over
40 minutes in wildtype mice (n=14) and 4.8±1.4 scratches in ablated mice (n=12). Student’s t-
test vs. WT: ***p<0.001.
103
type 3 serotonin receptors in ablated mice, but no change in the levels of type 2
receptors (data not shown).

Itch hypersensitivity is TRPM8-dependent
We next repeated the injections of pruritogens in M8KO mice to determine if the
effects we observed with ablated mice were dependent on TRPM8 or on other
molecules expressed within TRPM8 cells. Injections with histamine revealed that,
Figure 8.7: TRPM8-dependent alloknesis and hyperknesis

Intradermal injection of A)100µg histamine, B) 100µg 48/80, C) 200µg chloroquine and D) 60µg
BAM8-22 elicited to 64.4±8.6 (n=9), 91.8±7.8 (n=6), 103.5±14.2 (n=6), and 29.3±8.7 (n=8)
scratches in M8KO mice, respectively. Student’s t-test vs. WT: *p<0.05, **p<0.01,***p<0.001,
n.s. p>0.05.
104
similar to ablated mice, M8KO mice (n=9) scratched 64.4±8.6 times over 40 minutes
(Figure 8.7). Additionally, injection of compound 48/80, which degranulates mast cells
so that they release histamine, resulted in hyperknesis in M8KO mice (n=6), with
91.8±7.8 scratches. Non-histaminergic scratching with chloroquine was also TRPM8-
dependent with 103.5±14.2 scratches in M8KO mice (n=6). BAM8-22 results were
inconclusive, with 33.0±9.1 scratches in M8KO mice (n=7; p= 0.081009).
Serotonin-induced scratching was also TRPM8-dependent, but an interesting
effect was observed. While ablated mice showed practically no scratching after
injections of serotonin, M8KO mice (n=12) displayed an intermediate level of scratching,
with 17.3±2.4 scratches over 40 minutes (Figure 8.8).

Figure 8.8: Partially TRPM8-dependent serotonin effect

Intradermal injection of 30µg serotonin in M8KO mice (n=12) resulted in 17.3±2.4 scratches
over 40 minutes, which was significantly different from both WT and ablated mice (p=0.000403
vs. ablated). Student’s t-test vs. WT: **p<0.01,***p<0.001.
105
Conclusion
The observation that ablated mice spontaneously develop lesions, along with the
experimental results showing alloknesis and hyperknesis in TRPM8-deificient animals
suggests that TRPM8 and TRPM8-expressing neurons have an inhibitory effect on itch.
Cold-sensitive fibers are known to exhibit basal action potential firing regardless of
stimulation, the rate of which is reduced in M8KO mice (Bautista et al. 2007). It may be
that this basal firing is a mechanism for suppressing itch. In M8KO mice, the reduced
firing rate may be enough to suppress spontaneous itch (and thus M8KO mice do not
develop spontaneous lesions), but not enough to suppress itch in response to injected
pruritogens. Ablated mice, on the other hand, lack these cells altogether, thus both
spontaneous and induced itch are disinhibited.
The experiments with serotonin are particularly interesting given it was the only
pruritogen which resulted in decreased scratching as compared to wildtype mice. The
two serotonin receptors we found to be affected in our microarray set were types 1 and
3. Type 1 serotonin receptors are inhibitory GPCRs, while the type 3 receptor is an
excitatory ion channel. Previous experiments with a type 3 serotonin receptor knockout
mouse found no changes in induced itch, instead finding reduced responses in pain
assays (Zeitz et al. 2002). This presents the intriguing possibility that serotonin induces
itch via inhibition of TRPM8-expressing afferents, effectively a mechanism of
disinhibition. Future experiments using receptor subtype-specific agonists and
antagonists will provide insight into this issue.

106
CHAPTER NINE: Conclusion
In the studies described in this dissertation, we have conducted a number of
behavioral analyses of the effects of manipulating the cold-sensitive ion channel TRPM8
in mice. We have found six types of behavior which are to some degree TRPM8-
dependent, falling into two general categories of active and inhibitory. The active
behaviors were innocuous cold sensation, noxious cold sensation, cold hypersensitivity
with injury, and thermoregulation, and the inhibitory behaviors were analgesia and
antiprurisis. Together these studies lead to a better appreciation of the in vivo roles of
TRPM8 and TRPM8-expressing neurons beyond mere innocuous cold sensation (Daniels
and McKemy 2007).
In Chapter 3, we discussed experiments on innocuous cold sensation using
M8KO and ablated animals as well as pharmacological inhibition of TRPM8 with PBMC.
Using the two-temperature choice assay, we found that both M8KO and ablated mice
showed no preference for a 30°C plate over one cooled down to 20°C, nor did they
avoid the cooler temperatures. However, when the cooler plate was set at 15°C, M8KO
mice showed some preference for the warmer plate and avoidance of the cooler plate,
albeit at values significantly different from wildtype mice. Interestingly, ablated mice
did not show this behavior, remaining unaffected by the 15°C plate, indicating that
either developmental compensation in the M8KO mice or signals generated by other
molecules within TRPM8-expressing neurons was responsible for this difference.
In the evaporative cooling assay, both ablated and M8KO mice showed equal
impairment in responses to acetone evaporation, estimated to reach ~16°C.
107
Additionally, inhibiting the channel with systemic administration of PBMC reduced
responses in this assay in a dose-dependent manner, although at the highest dose
tested (20mg/kg), profound thermoregulatory effects were seen. Together these data
confirm previous findings that TRPM8 confers the in vivo ability to sense innocuous cold
temperatures.
Chapter 4 discussed the controversial role of TRPM8 in noxious cold sensing,
which was debated following the publication of the three original reports on M8KO mice
(Bautista et al. 2007; Colburn et al. 2007; Dhaka et al. 2007). Icilin induced both wet dog
shaking when administered i.p. and paw flinching with intraplantar injections, both of
which behaviors were TRPM8-dependent. Furthermore, activation of postsynaptic cells
within the spinal cord dorsal horn with icilin, menthol, or 0°C was also TRPM8-
dependent, indicating that without TRPM8, neural activity is not sent to the spinal cord.
Using the two-temperature choice assay, this time with the test plate set to
temperatures within the noxious cold range, we found that both M8KO and ablated
mice show some preference for the warmer home plate and avoidance of the colder
plate, although both genotypes show some impairment compared to wildtype mice. As
with a test plate setting of 15°C, M8KO and ablated mice exhibited significantly different
patterns of behavior, again indicating either developmental compensation in M8KO
mice or the contribution of additional molecules with TRPM8-expressing neurons.
Our final test of noxious cold was the cold plate assay. In a departure from
previously published studies, we found that purebred C57Bl/6 mice (the background
strain onto which all of our genetically modified mice are crossed) rarely displayed
108
hindpaw flinching and jumping behaviors which are normally measured for this assay.
Instead, all of the animals tested displayed two behaviors which we call front paw
flinching and front paw licking. Comparing the latencies of these two behaviors upon
placement on the cold plate, revealed that M8KO and ablated mice showed dramatic
impairments. Together these data suggest, through measures of behavioral responses
to cold and cold-mimetic chemicals, as well as through examination of neural activity
signatures, that normal expression of TRPM8 is necessary for the detection of noxious
cold.
The hypersensitivity to cold that is a symptom of chronic pain conditions was
discussed in Chapter 5. Using a model of induced inflammatory injury we found that
while M8KO and ablated animals develop less severe symptoms of cold allodynia as
assessed with the evaporative cooling assay, the symptoms were still significantly higher
than before injury. This would suggest that this symptom has both TRPM8-dependent
and -independent components. To test whether symptoms can be treated via TRPM8
antagonism, we tested wildtype mice with systemic administration of 10mg/kg PBMC,
finding that the drug can indeed reduce symptoms of cold allodynia. Again, we were
unable to reach full relief of hypersensitivity due to the limitations of thermoregulatory
side effects at higher doses.
In a model of neuropathic pain, we found a similar result in the evaporative
cooling assay in that both M8KO and ablated mice developed symptoms of cold
allodynia which were less severe than those exhibited by wildtype mice. However, in
this injury model we found that ablated mice displayed significantly higher levels of cold
109
hypersensitivity than M8KO mice. Since ablated mice lack TRPM8-expressing neurons
but not the TRPM8 gene, as M8KO mice do, we interpret the data to suggest that
expression of the TRPM8 gene in cells other than those which normally express TRPM8
occurs, although we could not detect any TRPM8 expression in DRGs taken ipsilateral to
the injury site at day 14 post-injury. Therefore it may be that any potential ectopic
expression of TRPM8 occurs in a dynamic manner and is undetectable two weeks after
injury.
Also similar to inflammatory injury, treatment of wildtype animals with
neuropathic injury with PBMC led to reductions in symptoms of col allodynia. Thus,
regardless of when any putative ectopic expression of TRPM8 occurs, it is clear that
impairment of TRPM8 leads to the development of less severe symptoms of cold
allodynia compared to normal mice in both inflammatory and neuropathic injury
conditions.
Chapter 6 explored the influence of TRPM8 activation and inhibition on
thermoregulation. Activation of the channel with 10mg/kg icilin led to a modest
hyperthermic response which was absent in M8KO mice. Conversely, inhibition of
TRPM8 with PBMC produced hypothermic responses which were dose- and TRPM8-
dependent. At 10mg/kg, PBMC resulted in a very mild hypothermia of less than one
degree in wildtype mice, whereas doubling the dose to 20mg/kg led to a profound drop
in core body temperature, in one instance reaching temperatures below 30°C. At both
doses the core body temperatures of M8KO mice were unaffected, indicating that
signals from TRPM8 are necessary for the proper maintenance of body temperature.
110
Chapter 7 marks the transition from active effects to inhibitory effects with a
discussion on the role of TRPM8 in cooling-induced analgesia. Wildtype mice exhibit a
reduction in symptoms of mechanical allodynia that occur with neuropathic injury for
twenty minutes following five minutes of mild cooling. The analgesic effect was absent
in both M8KO and ablated mice, indicating that this phenomenon depends on the
channel itself, not simply on activity from TRPM8-expressing cells. Interestingly,
responses on the uninjured side did not vary significantly with cooling, indicating that
the activation of TRPM8 specifically inhibits signals responsible for mechanical allodynia
after injury.
To our knowledge, Chapter 8 outlines the first evidence that TRPM8 affects itch.
Ablated mice developed skin lesions shortly after ablation, similar to three previous
reports on lesions produced by sensitization of itch signaling pathways (Lagerström et
al. 2010; Yang Liu et al. 2010; Ross et al. 2010). We investigated whether our mice were
hypersensitive to induced itch as well, finding that they indeed exhibited heightened
scratching behaviors in response to compounds activating both histaminergic and
nonhistaminergic itch pathways as compared to wildtype mice. The only exception was
serotonin, which induced very little scratching in ablated mice as compared to wildtype
mice, an intriguing finding given that two types of serotonin receptors (type 1 and type
3) were decreased in our microarray data comparing gene transcripts from wildtype
mice to those from ablated animals.
The heightened itch responses seen in ablated animals could either be due to
TRPM8 itself or to other molecules expressing within the TRPM8-expressing population.
111
Therefore we also test M8KO mice with injections of pruritogens. Remarkably, M8KO
mice also exhibited heightened scratching responses as compared to wildtype mice,
indicating that TRPM8 is required for normal responses. Since the ablated mice lack
TRPM8-expressing cells altogether, it is unlikely that pruritogens directly activate these
cells to generate itch signals. Rather, it is more likely that activation of TRPM8 serves as
a brake on itch signaling, thus when the channel or the cells are removed, heightened
itch results. Indeed, menthol is a common drug used for itchy skin conditions, which
likely serves to increase the rate of TRPM8-based inhibitory signaling to counteract the
itch.
The one exception to this model is serotonin, which, as mentioned above, fails to
elicit significant amounts of scratching in ablated mice. When we repeated these
injections in M8KO mice, we found a level of scratching intermediate to both wildtype
and ablated mice. The two serotonin receptor types implicated in the microarray data
set are interesting in that one (type 1) is an inhibitory GPCR, while the other (type 3) is
an excitatory ion channel. Previous experiments with a type 3-knockout mouse showed
that serotonin-induced itch is preserved, while pain-related behaviors are impaired in
mice lacking this receptor (Zeitz et al. 2002). Since activation of the type 1 receptor
leads to inhibition of neural signals, this may suggest that the effect of TRPM8 on itch is
inhibitory, and that inhibition of TRPM8 cells through the type 1 receptor may
effectively disinhibit the itch circuit.
Future studies on this work will provide us with a better understanding of the
myriad functions of TRPM8 in vivo. These results and many other studies suggest that
112
the population of TRPM8-expressing sensory neurons can be divided into
subpopulations that may affect different behaviors. Identifying molecular markers of
these subpopulations and using genetic manipulations to specifically target them will
further our understanding of this fascinating cell population, and may provide us with
new pathways to target chronic conditions such as pain and itch.
113
REFERENCES
Abe, J, H Hosokawa, M Okazawa, M Kandachi, Y Sawada, K Yamanaka, K Matsumura,
and S Kobayashi. 2005. “TRPM8 protein localization in trigeminal ganglion and taste
papillae.” Brain Res Mol Brain Res 136 (1-2): 91-98.
Abe, J, H Hosokawa, Y Sawada, K Matsumura, and S Kobayashi. 2006. “Ca
2+
-dependent
PKC activation mediates menthol-induced desensitization of transient receptor
potential M8.” Neurosci Lett 397 (1-2): 140-144.
van Aken, A F, M Atiba-Davies, W Marcotti, R J Goodyear, J E Bryant, G P Richardson, K
Noben-Trauth, and C J Kros. 2008. “TRPML3 mutations cause impaired mechano-
electrical transduction and depolarization by an inward-rectifier cation current in
auditory hair cells of variant-waddler mice.” J Physiol 586 (Pt 22): 5403-5418.
Akiyama, T, M Iodi Carstens, and E Carstens. 2010. “Facial injections of pruritogens and
algogens excite partly overlapping populations of primary and second-order
trigeminal neurons in mice.” Journal of neurophysiology 104 (5) (November): 2442-
50.
Andersson, D A, H W Chase, and S Bevan. 2004. “TRPM8 activation by menthol, icilin,
and cold is differentially modulated by intracellular pH.” J Neurosci 24 (23): 5364-
5369.
Andersson, D A, M Nash, and S Bevan. 2007. “Modulation of the cold-activated channel
TRPM8 by lysophospholipids and polyunsaturated fatty acids.” J Neurosci 27 (12):
3347-3355.
Axelsson, H E, J K Minde, A Sonesson, G Toolanen, E D Hogestatt, and P M Zygmunt.
2009. “Transient receptor potential vanilloid 1, vanilloid 2 and melastatin 8
immunoreactive nerve fibers in human skin from individuals with and without
Norrbottnian congenital insensitivity to pain.” Neuroscience 162 (4): 1322-1332.
Babes, A, D Zorzon, and G Reid. 2004. “Two populations of cold-sensitive neurons in rat
dorsal root ganglia and their modulation by nerve growth factor.” Eur J Neurosci 20
(9): 2276-2282.
Bai, V U, S Murthy, K Chinnakannu, F Muhletaler, S Tejwani, E R Barrack, S H Kim, M
Menon, and G P Veer Reddy. 2010. “Androgen regulated TRPM8 expression: a
potential mRNA marker for metastatic prostate cancer detection in body fluids.” Int
J Oncol 36 (2): 443-450.
114
Bandell, M, A E Dubin, M J Petrus, A Orth, J Mathur, S W Hwang, and A Patapoutian.
2006. “High-throughput random mutagenesis screen reveals TRPM8 residues
specifically required for activation by menthol.” Nat Neurosci 9 (4): 493-500.
Bandell, M, G M Story, S W Hwang, V Viswanath, S R Eid, M J Petrus, T J Earley, and A
Patapoutian. 2004. “Noxious cold ion channel TRPA1 is activated by pungent
compounds and bradykinin.” Neuron 41 (6): 849-857.
Bang, S, and S W Hwang. 2009. “Polymodal ligand sensitivity of TRPA1 and its modes of
interactions.” J Gen Physiol 133 (3): 257-262.
Bautista, D M, S E Jordt, T Nikai, P R Tsuruda, A J Read, J Poblete, E N Yamoah, A I
Basbaum, and D Julius. 2006. “TRPA1 mediates the inflammatory actions of
environmental irritants and proalgesic agents.” Cell 124 (6): 1269-1282.
Bautista, D M, P Movahed, A Hinman, H E Axelsson, O Sterner, E D Hogestatt, D Julius, S
E Jordt, and P M Zygmunt. 2005. “Pungent products from garlic activate the
sensory ion channel TRPA1.” Proc Natl Acad Sci U S A 102 (34): 12248-12252.
Bautista, D M, J Siemens, J M Glazer, P R Tsuruda, A I Basbaum, C L Stucky, S E Jordt, and
D Julius. 2007. “The menthol receptor TRPM8 is the principal detector of
environmental cold.” Nature 448 (7150): 204-208.
Bavencoffe, A, D Gkika, A Kondratskyi, B Beck, A S Borowiec, G Bidaux, J Busserolles, et
al. 2010. “The transient receptor potential channel TRPM8 is inhibited via the
alpha2A adrenoreceptor signaling pathway.” J Biol Chem.
Bechi, G, P Scalmani, E Schiavon, R Rusconi, S Franceschetti, and M Mantegazza. 2012.
“Pure haploinsufficiency for Dravet syndrome Na(V)1.1 (SCN1A) sodium channel
truncating mutations.” Epilepsia 53 (1): 87-100.
Beck, B, G Bidaux, A Bavencoffe, L Lemonnier, S Thebault, Y Shuba, G Barrit, R Skryma,
and N Prevarskaya. 2007. “Prospects for prostate cancer imaging and therapy using
high-affinity TRPM8 activators.” Cell Calcium 41 (3): 285-294.
Behrendt, H J, T Germann, C Gillen, H Hatt, and R Jostock. 2004. “Characterization of the
mouse cold-menthol receptor TRPM8 and vanilloid receptor type-1 VR1 using a
fluorometric imaging plate reader (FLIPR) assay.” Br J Pharmacol 141 (4): 737-745.
Benedikt, J, J Teisinger, L Vyklicky, and V Vlachova. 2007. “Ethanol inhibits cold-menthol
receptor TRPM8 by modulating its interaction with membrane phosphatidylinositol
4,5-bisphosphate.” J Neurochem 100 (1): 211-224.
115
Bessou, P, and E R Perl. 1969. “Response of cutaneous sensory units with unmyelinated
fibers to noxious stimuli.” Journal of neurophysiology 32 (6) (November): 1025-43.
Bidaux, G, M Flourakis, S Thebault, A Zholos, B Beck, D Gkika, M Roudbaraki, et al. 2007.
“Prostate cell differentiation status determines transient receptor potential
melastatin member 8 channel subcellular localization and function.” J Clin Invest
117 (6): 1647-1657.
Bini, G, G Cruccu, K E Hagbarth, W Schady, and E Torebjörk. 1984. “Analgesic effect of
vibration and cooling on pain induced by intraneural electrical stimulation.” Pain 18
(3) (March): 239-48.
De Blas, G A, A Darszon, A Y Ocampo, C J Serrano, L E Castellano, E O Hernandez-
Gonzalez, M Chirinos, F Larrea, C Beltran, and C L Trevino. 2009. “TRPM8, a
versatile channel in human sperm.” PLoS ONE 4 (6): e6095.
Bodding, M, U Wissenbach, and V Flockerzi. 2007. “Characterisation of TRPM8 as a
pharmacophore receptor.” Cell Calcium.
Brauchi, S, P Orio, and R Latorre. 2004. “Clues to understanding cold sensation:
thermodynamics and electrophysiological analysis of the cold receptor TRPM8.”
Proc Natl Acad Sci U S A 101 (43): 15494-15499.
Brauchi, S, G Orta, C Mascayano, M Salazar, N Raddatz, H Urbina, E Rosenmann, F
Gonzalez-Nilo, and R Latorre. 2007. “Dissection of the components for PIP2
activation and thermosensation in TRP channels.” Proc Natl Acad Sci U S A 104 (24):
10246-10251.
Brauchi, S, G Orta, M Salazar, E Rosenmann, and R Latorre. 2006. “A hot-sensing cold
receptor: C-terminal domain determines thermosensation in transient receptor
potential channels.” J Neurosci 26 (18): 4835-4840.
Cahusac, P M, and R Noyce. 2007. “A pharmacological study of slowly adapting
mechanoreceptors responsive to cold thermal stimulation.” Neuroscience 148 (2):
489-500.
del Camino, Donato, Sarah Murphy, Melissa Heiry, Lee B Barrett, Taryn J Earley, Colby a
Cook, Matt J Petrus, et al. 2010. “TRPA1 contributes to cold hypersensitivity.” J
Neurosci 30 (45): 15165-15174.
Canetta, S E, E Luca, E Pertot, L W Role, and D A Talmage. 2011. “Type III Nrg1 back
signaling enhances functional TRPV1 along sensory axons contributing to basal and
inflammatory thermal pain sensation.” PLoS ONE 6 (9): e25108.
116
Carr, R W, S Pianova, D D McKemy, and J A Brock. 2009. “Action potential initiation in
the peripheral terminals of cold-sensitive neurones innervating the guinea-pig
cornea.” J Physiol 587 (Pt 6): 1249-1264.
Caspani, O, and P A Heppenstall. 2009. “TRPA1 and cold transduction: an unresolved
issue?” J Gen Physiol 133 (3): 245-249.
Caspani, Ombretta, Sandra Zurborg, Dominika Labuz, and Paul a Heppenstall. 2009. “The
contribution of TRPM8 and TRPA1 channels to cold allodynia and neuropathic
pain.” PloS one 4 (10) (January): e7383.
Chen, J, S K Joshi, S DiDomenico, R J Perner, J P Mikusa, D M Gauvin, J A Segreti, et al.
2011. “Selective blockade of TRPA1 channel attenuates pathological pain without
altering noxious cold sensation or body temperature regulation.” Pain 152 (5):
1165-1172.
Chen, Z, O Ishizuka, T Imamura, N Aizawa, Y Kurizaki, Y Igawa, O Nishizawa, and K E
Andersson. 2009. “Stimulation of skin menthol receptors stimulates detrusor
activity in conscious rats.” Neurourol Urodyn.
Chuang, H H, W M Neuhausser, and D Julius. 2004. “The super-cooling agent icilin
reveals a mechanism of coincidence detection by a temperature-sensitive TRP
channel.” Neuron 43 (6): 859-869.
Coggeshall, Richard E. 2005. “Fos, nociception and the dorsal horn.” Progress in
Neurobiology 77 (5): 299-352.
Colburn, R W, M L Lubin, D J Stone Jr., Y Wang, D Lawrence, M R D’Andrea, M R Brandt,
Y Liu, C M Flores, and N Qin. 2007. “Attenuated cold sensitivity in TRPM8 null
mice.” Neuron 54 (3): 379-386.
da Costa, Diogo Santos M, Flavia Carla Meotti, Edinéia Lemos Andrade, Paulo César Leal,
Emerson Marcelo Motta, and João B Calixto. 2010. “The involvement of the
transient receptor potential A1 (TRPA1) in the maintenance of mechanical and cold
hyperalgesia in persistent inflammation.” Pain 148 (3): 431-437.
Daniels, R L, and D D McKemy. 2007. “Mice left out in the cold: commentary on the
phenotype of TRPM8-nulls.” Mol Pain 3: 23.
Daniels, R L, Y Takashima, and D D McKemy. 2009. “Activity of the Neuronal Cold Sensor
TRPM8 Is Regulated by Phospholipase C via the Phospholipid Phosphoinositol 4,5-
Bisphosphate.” J Biol Chem 284 (3): 1570-1582.
117
Davidson, Steve, and Glenn J Giesler. 2010. “The multiple pathways for itch and their
interactions with pain.” Trends in neurosciences 33 (12) (December): 550-8.
Davis, K D, and G E Pope. 2002. “Noxious cold evokes multiple sensations with distinct
time courses.” Pain 98 (1-2): 179-185.
Dhaka, A, T J Earley, J Watson, and A Patapoutian. 2008. “Visualizing cold spots: TRPM8-
expressing sensory neurons and their projections.” J Neurosci 28 (3): 566-575.
Dhaka, Ajay, Amber N Murray, Jayanti Mathur, Taryn J Earley, Matt J Petrus, and Ardem
Patapoutian. 2007. “TRPM8 is required for cold sensation in mice.” Neuron 54 (3)
(May 3): 371-378.
Ding, Z, T Gomez, J L Werkheiser, A Cowan, and S M Rawls. 2008. “Icilin induces a
hyperthermia in rats that is dependent on nitric oxide production and NMDA
receptor activation.” Eur J Pharmacol 578 (2-3): 201-208.
Doerner, J F, G Gisselmann, H Hatt, and C H Wetzel. 2007. “Transient receptor potential
channel A1 is directly gated by calcium ions.” J Biol Chem 282 (18): 13180-13189.
Dragoni, I, E Guida, and P McIntyre. 2006. “The cold and menthol receptor TRPM8
contains a functionally important double cysteine motif.” J Biol Chem 281 (49):
37353-37360.
Du, J, X Yang, L Zhang, and Y M Zeng. 2009. “Expression of TRPM8 in the distal
cerebrospinal fluid-contacting neurons in the brain mesencephalon of rats.”
Cerebrospinal Fluid Res 6: 3.
Du, S, I Araki, H Kobayashi, H Zakoji, N Sawada, and M Takeda. 2008. “Differential
expression profile of cold (TRPA1) and cool (TRPM8) receptors in human urogenital
organs.” Urology 72 (2): 450-455.
Duarte, D B, J H Duan, G D Nicol, M R Vasko, and C M Hingtgen. 2011. “Reduced
expression of SynGAP, a neuronal GTPase-activating protein, enhances capsaicin-
induced peripheral sensitization.” J Neurophysiol 106 (1): 309-318.
Eid, S R, and D N Cortright. 2009. “Transient receptor potential channels on sensory
nerves.” Handb Exp Pharmacol (194): 261-281.
Ellis, M. 1961. “The treatment of pain by ethyl chloride and other cooling sprays.”
Practitioner 187: 367-70.
118
Erler, I, D M Al-Ansary, U Wissenbach, T F Wagner, V Flockerzi, and B A Niemeyer. 2006.
“Trafficking and assembly of the cold-sensitive TRPM8 channel.” J Biol Chem 281
(50): 38396-38404.
Evans, D. 1981. “Cryoanalgesia.” Anaesthesia 36: 1003-1013.
Everaerts, W, T Gevaert, B Nilius, and D De Ridder. 2008. “On the origin of bladder
sensing: Tr(i)ps in urology.” Neurourol Urodyn 27 (4): 264-273.
Fajardo, O, V Meseguer, C Belmonte, and F Viana. 2008. “TRPA1 channels mediate cold
temperature sensing in mammalian vagal sensory neurons: pharmacological and
genetic evidence.” J Neurosci 28 (31): 7863-7875.
Frederick, J, M E Buck, D J Matson, and D N Cortright. 2007. “Increased TRPA1, TRPM8,
and TRPV2 expression in dorsal root ganglia by nerve injury.” Biochem Biophys Res
Commun 358 (4): 1058-1064.
Garrison, Sheldon R, Alexander Dietrich, and Cheryl L Stucky. 2012. “TRPC1 contributes
to light-touch sensation and mechanical responses in low-threshold cutaneous
sensory neurons.” Journal of neurophysiology 107 (3) (February): 913-22.
Gauchan, P, T Andoh, A Kato, and Y Kuraishi. 2009. “Involvement of increased
expression of transient receptor potential melastatin 8 in oxaliplatin-induced cold
allodynia in mice.” Neurosci Lett 458 (2): 93-95.
Gavva, Narender R. 2008. “Body-temperature maintenance as the predominant function
of the vanilloid receptor TRPV1.” Trends Pharmacol Sci 29 (11): 550-557.
Gavva, Narender R, James J S Treanor, Andras Garami, Liang Fang, Sekhar Surapaneni,
Anna Akrami, Francisco Alvarez, et al. 2008. “Pharmacological blockade of the
vanilloid receptor TRPV1 elicits marked hyperthermia in humans.” Pain 136 (1-2)
(May): 202-10.
Gebhart, G. F. 1983. “Recent Developments in the Neurochemical Bases of Pain and
Analgesia.” NIDA Monograph 45: 19-35.
van Genderen, M M, M M Bijveld, Y B Claassen, R J Florijn, J N Pearring, F M Meire, M A
McCall, et al. 2009. “Mutations in TRPM1 are a common cause of complete
congenital stationary night blindness.” Am J Hum Genet 85 (5): 730-736.
Gentry, C, N Stoakley, D A Andersson, and S Bevan. 2010. “The roles of iPLA2, TRPM8
and TRPA1 in chemically induced cold hypersensitivity.” Mol Pain 6: 4.
119
Giesbrecht, G G, and M Younes. 1995. “Exercise- and cold-induced asthma.” Can J Appl
Physiol 20 (3): 300-314.
Hayashi, T, T Kondo, M Ishimatsu, S Yamada, K Nakamura, K Matsuoka, and T Akasu.
2009. “Expression of the TRPM8-immunoreactivity in dorsal root ganglion neurons
innervating the rat urinary bladder.” Neurosci Res 65 (3): 245-251.
Hayes, R L, and Y Katayama. 1986. “Range of environmental stimuli producing
nociceptive suppression: implications for neural mechanisms.” Annals of the New
York Academy of Sciences 467 (January): 1-13.
Hercogova, Jana. 2005. “Topical anti-itch therapy.” Dermatologic therapy 18: 341-343.
Hjerling-Leffler, J, M Alqatari, P Ernfors, and M Koltzenburg. 2007. “Emergence of
functional sensory subtypes as defined by transient receptor potential channel
expression.” J Neurosci 27 (10): 2435-2443.
Hondoh, A, Y Ishida, S Ugawa, T Ueda, Y Shibata, T Yamada, M Shikano, S Murakami, and
S Shimada. 2010. “Distinct expression of cold receptors (TRPM8 and TRPA1) in the
rat nodose-petrosal ganglion complex.” Brain Res 1319: 60-69.
Hui, K, Y Guo, and Z P Feng. 2005. “Biophysical properties of menthol-activated cold
receptor TRPM8 channels.” Biochem Biophys Res Commun 333 (2): 374-382.
Hunt, S P, A Pini, and G Evan. 1987. “Induction of c-fos-like protein in spinal cord
neurons following sensory stimulation.” Nature 328 (6131): 632-634.
Iida, Tohko, Isao Shimizu, Michele L Nealen, Ashley Campbell, and Michael Caterina.
2005. “Attenuated fever response in mice lacking TRPV1.” Neuroscience letters 378
(1) (April 11): 28-33.
Imamachi, Noritaka, Goon Ho Park, Hyosang Lee, David J Anderson, Melvin I Simon,
Allan I Basbaum, and Sang-Kyou Han. 2009. “TRPV1-expressing primary afferents
generate behavioral responses to pruritogens via multiple mechanisms.”
Proceedings of the National Academy of Sciences of the United States of America
106 (27) (July 7): 11330-5.
Jeffry, Joseph, Seungil Kim, and Zhou-Feng Chen. 2011. “Itch signaling in the nervous
system.” Physiology (Bethesda, Md.) 26 (4) (August): 286-92.
Johnson, C D, D Melanaphy, A Purse, S A Stokesberry, P Dickson, and A V Zholos. 2009.
“Transient receptor potential melastatin 8 channel involvement in the regulation of
vascular tone.” Am J Physiol Heart Circ Physiol 296 (6): H1868-77.
120
Jolas, Thierry, and George K Aghajanian. 1997. “NEUROTENSIN AND THE SEROTONERGIC
SYSTEM.” Progress in Neurobiology 52: 455-468.
Jordt, S E, D M Bautista, H H Chuang, D D McKemy, P M Zygmunt, E D Hogestatt, I D
Meng, and D Julius. 2004. “Mustard oils and cannabinoids excite sensory nerve
fibres through the TRP channel ANKTM1.” Nature 427 (6971): 260-265.
Jung, Steffen, Derya Unutmaz, Phillip Wong, Gen-Ichiro Sano, Kenia De los Santos, Tim
Sparwasser, Shengji Wu, et al. 2002. “In vivo depletion of CD11c+ dendritic cells
abrogates priming of CD8+ T cells by exogenous cell-associated antigens.”
Immunity 17 (2) (August): 211-20.
Karashima, Y, K Talavera, W Everaerts, A Janssens, K Y Kwan, R Vennekens, B Nilius, and
T Voets. 2009. “TRPA1 acts as a cold sensor in vitro and in vivo.” Proc Natl Acad Sci
U S A 106 (4): 1273-1278.
Katsura, H, K Obata, T Mizushima, H Yamanaka, K Kobayashi, Y Dai, T Fukuoka, A
Tokunaga, M Sakagami, and K Noguchi. 2006. “Antisense knock down of TRPA1, but
not TRPM8, alleviates cold hyperalgesia after spinal nerve ligation in rats.” Exp
Neurol 200 (1): 112-123.
Kim, A Y, Z Tang, Q Liu, K N Patel, D Maag, Y Geng, and X Dong. 2008. “Pirt, a
phosphoinositide-binding protein, functions as a regulatory subunit of TRPV1.” Cell
133 (3): 475-485.
Kim, Byung Moon, Sang Hee Lee, Won Sik Shim, and Uhtaek Oh. 2004. “Histamine-
induced Ca(2+) influx via the PLA(2)/lipoxygenase/TRPV1 pathway in rat sensory
neurons.” Neuroscience letters 361 (1-3) (May 6): 159-62.
Klein, Amanda, Mirela Iodi Carstens, and E Carstens. 2011. “Facial injections of
pruritogens or algogens elicit distinct behavior responses in rats and excite
overlapping populations of primary sensory and trigeminal subnucleus caudalis
neurons.” Journal of neurophysiology 106 (3) (September): 1078-88.
Knowlton, Wendy M, Richard L Daniels, Radhika Palkar, Daniel D McCoy, and David D
McKemy. 2011. “Pharmacological blockade of TRPM8 ion channels alters cold and
cold pain responses in mice.” PLoS ONE 6 (9) (January): e25894.
Knowlton, Wendy M, and David D McKemy. 2011. “TRPM8: from cold to cancer,
peppermint to pain.” Current pharmaceutical biotechnology 12 (1) (January 1): 68-
77.
121
Knowlton, Wendy M., A Bifolck-Fisher, D M Bautista, and D D McKemy. 2010. “TRPM8,
but not TRPA1, is required for neural and behavioral responses to acute noxious
cold temperatures and cold-mimetics in vivo.” Pain 150 (2): 340-350.
Kobayashi, K, T Fukuoka, K Obata, H Yamanaka, Y Dai, A Tokunaga, and K Noguchi. 2005.
“Distinct expression of TRPM8, TRPA1, and TRPV1 mRNAs in rat primary afferent
neurons with adelta/c-fibers and colocalization with trk receptors.” J Comp Neurol
493 (4): 596-606.
Kozyreva, T V, Eia Tkachenko, T A Potapova, A G Romashchenko, and M I Voevoda.
2011. “[Relationship of single-nucleotide polymorphism rs11562975 in thermo-
sensitive ion channel TRPM8 gene with human sensitivity to cold and menthol].”
Fiziol Cheloveka 37 (2): 71-76.
Kuhn, F J, G Knop, and A Luckhoff. 2007. “The transmembrane segment S6 determines
cation versus anion selectivity of TRPM2 and TRPM8.” J Biol Chem 282 (38): 27598-
27609.
Kuhn, F J, C Kuhn, and A Luckhoff. 2009. “Inhibition of TRPM8 by icilin distinct from
desensitization induced by menthol and menthol derivatives.” J Biol Chem 284 (7):
4102-4111.
Kwan, K Y, and D P Corey. 2009. “Burning cold: involvement of TRPA1 in noxious cold
sensation.” J Gen Physiol 133 (3): 251-256.
Kwan, K Y, J M Glazer, D P Corey, F L Rice, and C L Stucky. 2009. “TRPA1 modulates
mechanotransduction in cutaneous sensory neurons.” J Neurosci 29 (15): 4808-
4819.
Kwan, Kelvin Y, Andrew J Allchorne, Melissa A Vollrath, Adam P Christensen, Duan-Sun
Zhang, Clifford J Woolf, and David P Corey. 2006. “TRPA1 Contributes to Cold,
Mechanical, and Chemical Nociception but Is Not Essential for Hair-Cell
Transduction.” Neuron 50 (2): 277-289.
Lagerström, Malin C, Katarzyna Rogoz, Bjarke Abrahamsen, Emma Persson, Björn
Reinius, Karin Nordenankar, Caroline Olund, et al. 2010. “VGLUT2-dependent
sensory neurons in the TRPV1 population regulate pain and itch.” Neuron 68 (3)
(November 4): 529-42.
Lashinger, E S, M S Steiginga, J P Hieble, L A Leon, S D Gardner, R Nagilla, E A Davenport,
B E Hoffman, N J Laping, and X Su. 2008. “AMTB, a TRPM8 channel blocker:
evidence in rats for activity in overactive bladder and painful bladder syndrome.”
Am J Physiol Renal Physiol 295 (3): F803-10.
122
Lee, K P, J Y Jun, I Y Chang, S H Suh, I So, and K W Kim. 2005. “TRPC4 is an essential
component of the nonselective cation channel activated by muscarinic stimulation
in mouse visceral smooth muscle cells.” Mol Cells 20 (3): 435-441.
Leknes, Siri, Jonathan C W Brooks, Katja Wiech, and Irene Tracey. 2008. “Pain relief as
an opponent process: a psychophysical investigation.” The European journal of
neuroscience 28 (4) (August): 794-801.
Li, Q, X Wang, Z Yang, B Wang, and S Li. 2009. “Menthol induces cell death via the
TRPM8 channel in the human bladder cancer cell line T24.” Oncology 77 (6): 335-
341.
Liu, B, and F Qin. 2005. “Functional control of cold- and menthol-sensitive TRPM8 ion
channels by phosphatidylinositol 4,5-bisphosphate.” J Neurosci 25 (7): 1674-1681.
Liu, Qin, Zongxiang Tang, Lenka Surdenikova, Seungil Kim, Kush N Patel, Andrew Kim, Fei
Ru, et al. 2009. “Sensory Neuron-Specific GPCR Mrgprs Are Itch Receptors
Mediating Chloroquine-Induced Pruritis.” Cell 139 (7): 1353-1365.
Liu, Y, M L Lubin, T L Reitz, Y Wang, R W Colburn, C M Flores, and N Qin. 2006.
“Molecular identification and functional characterization of a temperature-
sensitive transient receptor potential channel (TRPM8) from canine.” Eur J
Pharmacol 530 (1-2): 23-32.
Liu, Yang, Omar Abdel Samad, Ling Zhang, Bo Duan, Qingchun Tong, Claudia Lopes, Ru-
Rong Ji, Bradford B Lowell, and Qiufu Ma. 2010. “VGLUT2-dependent glutamate
release from nociceptors is required to sense pain and suppress itch.” Neuron 68
(3) (November 4): 543-56.
Luo, W, S R Wickramasinghe, J M Savitt, J W Griffin, T M Dawson, and D D Ginty. 2007.
“A Hierarchical NGF Signaling Cascade Controls Ret-Dependent and Ret-
Independent Events during Development of Nonpeptidergic DRG Neurons.” Neuron
54 (5): 739-754.
Ma, S, G G, V E Ak, D Jf, and H H. 2008. “Menthol derivative WS-12 selectively activates
transient receptor potential melastatin-8 (TRPM8) ion channels.” Pak J Pharm Sci
21 (4): 370-378.
Macpherson, L J, A E Dubin, M J Evans, F Marr, P G Schultz, B F Cravatt, and A
Patapoutian. 2007. “Noxious compounds activate TRPA1 ion channels through
covalent modification of cysteines.” Nature 445 (7127): 541-545.
123
Madrid, R, T Donovan-Rodriguez, V Meseguer, M C Acosta, C Belmonte, and F Viana.
2006. “Contribution of TRPM8 channels to cold transduction in primary sensory
neurons and peripheral nerve terminals.” J Neurosci 26 (48): 12512-12525.
Madrid, R, E de la Pena, T Donovan-Rodriguez, C Belmonte, and F Viana. 2009. “Variable
threshold of trigeminal cold-thermosensitive neurons is determined by a balance
between TRPM8 and Kv1 potassium channels.” J Neurosci 29 (10): 3120-3131.
Mahieu, F, A Janssens, M Gees, K Talavera, B Nilius, and T Voets. 2010. “Modulation of
the cold-activated cation channel TRPM8 by surface charge screening.” J Physiol
588 (Pt 2): 315-324.
Malkia, A, R Madrid, V Meseguer, E de la Pena, M Valero, C Belmonte, and F Viana.
2007. “Bidirectional shifts of TRPM8 channel gating by temperature and chemical
agents modulate the cold sensitivity of mammalian thermoreceptors.” J Physiol 581
(Pt 1): 155-174.
Mandadi, S, S T Nakanishi, Y Takashima, A Dhaka, A Patapoutian, D D McKemy, and P J
Whelan. 2009. “Locomotor networks are targets of modulation by sensory
transient receptor potential vanilloid 1 and transient receptor potential melastatin
8 channels.” Neuroscience 162 (4): 1377-1397.
Masamoto, Yukiko, Fuminori Kawabata, and Tohru Fushiki. 2009. “Intragastric
administration of TRPV1, TRPV3, TRPM8, and TRPA1 agonists modulates autonomic
thermoregulation in different manners in mice.” Bioscience, biotechnology, and
biochemistry 73 (5) (May): 1021-7.
Matta, Jose A, and Gerard P Ahern. 2007. “Voltage is a partial activator of rat
thermosensitive TRP channels.” J Physiol 585 (2): 469-482.
McCormack, Keith. 1994. “Non-steroidal anti-inflammatory drugs and spinal nociceptive
processing.” Pain 59: 9-43.
McCoy, D D, Wendy M. Knowlton, and D D McKemy. 2011. “Scraping through the ice:
uncovering the role of TRPM8 in cold transduction.” Am J Physiol Regul Integr
Comp Physiol 300 (6): R1278-87.
McKemy, David D. 2005. “How cold is it? TRPM8 and TRPA1 in the molecular logic of
cold sensation.” Mol Pain 1 (January): 16.
McKemy, David D, W M Neuhausser, and D Julius. 2002. “Identification of a cold
receptor reveals a general role for TRP channels in thermosensation.” Nature 416
(6876): 52-8.
124
Melzack, Ronald, and Patrick D. Wall. 1965. “Pain Mechanisms : A New Theory.”
Science1 150 (3699): 971-979.
Meseguer, V, Y Karashima, K Talavera, D D’Hoedt, T Donovan-Rodriguez, F Viana, B
Nilius, and T Voets. 2008. “Transient receptor potential channels in sensory
neurons are targets of the antimycotic agent clotrimazole.” J Neurosci 28 (3): 576-
586.
Montell, Craig, and Michael J Caterina. 2007. “Thermoregulation: channels that are cool
to the core.” Current biology : CB 17 (20) (October 23): R885-7.
Morin, C, and M C Bushnell. 1998. “Temporal and qualitative properties of cold pain and
heat pain: a psychophysical study.” Pain 74 (1): 67-73.
Mukerji, G, J Waters, I P Chessell, C Bountra, S K Agarwal, and P Anand. 2006. “Pain
during ice water test distinguishes clinical bladder hypersensitivity from
overactivity disorders.” BMC Urol 6: 31.
Myers, B R, Y M Sigal, and D Julius. 2009. “Evolution of thermal response properties in a
cold-activated TRP channel.” PLoS ONE 4 (5): e5741.
Nassini, R, M Gees, S Harrison, G De Siena, S Materazzi, N Moretto, P Failli, et al. 2011.
“Oxaliplatin elicits mechanical and cold allodynia in rodents via TRPA1 receptor
stimulation.” Pain 152 (7): 1621-1631.
Nomoto, Shigeki, Masaaki Shibata, Masami Iriki, and Walter Riedel. 2004. “Role of
afferent pathways of heat and cold in body temperature regulation.” International
journal of biometeorology 49 (2) (November): 67-85.
Obata, K, H Katsura, T Mizushima, H Yamanaka, K Kobayashi, Y Dai, T Fukuoka, A
Tokunaga, M Tominaga, and K Noguchi. 2005. “TRPA1 induced in sensory neurons
contributes to cold hyperalgesia after inflammation and nerve injury.” J Clin Invest
115 (9): 2393-2401.
Park, C K, M S Kim, Z Fang, H Y Li, S J Jung, S Y Choi, S J Lee, K Park, J S Kim, and S B Oh.
2006. “Functional expression of thermo-transient receptor potential channels in
dental primary afferent neurons: Implication for tooth pain.” J Biol Chem.
Parra, A, R Madrid, D Echevarria, S del Olmo, C Morenilla-Palao, M C Acosta, J Gallar, A
Dhaka, F Viana, and C Belmonte. 2010. “Ocular surface wetness is regulated by
TRPM8-dependent cold thermoreceptors of the cornea.” Nat Med 16 (12): 1396-
1399.
125
Patel, Kush N, and Xinzhong Dong. 2010. “An itch to be scratched.” Neuron 68 (3)
(November 4): 334-9.
Patel, Tejesh, and Gil Yosipovitch. 2010. “Therapy of pruritus.” Expert opinion on
pharmacotherapy 11 (10) (July): 1673-82.
Pedretti, A, C Marconi, I Bettinelli, and G Vistoli. 2009. “Comparative modeling of the
quaternary structure for the human TRPM8 channel and analysis of its binding
features.” Biochim Biophys Acta 1788 (5): 973-982.
Peier, A M, A Moqrich, A C Hergarden, A J Reeve, D A Andersson, G M Story, T J Earley,
et al. 2002. “A TRP channel that senses cold stimuli and menthol.” Cell 108 (5): 705-
15.
De Petrocellis, L, K Starowicz, A S Moriello, M Vivese, P Orlando, and V Di Marzo. 2007.
“Regulation of transient receptor potential channels of melastatin type 8 (TRPM8):
effect of cAMP, cannabinoid CB(1) receptors and endovanilloids.” Exp Cell Res 313
(9): 1911-1920.
Petrus, M, A M Peier, M Bandell, S W Hwang, T Huynh, N Olney, T Jegla, and A
Patapoutian. 2007. “A role of TRPA1 in mechanical hyperalgesia is revealed by
pharmacological inhibition.” Mol Pain 3: 40.
Phelps, C B, and R Gaudet. 2007. “The role of the N terminus and transmembrane
domain of TRPM8 in channel localization and tetramerization.” J Biol Chem 282
(50): 36474-36480.
Polderman, Kees H. 2009. “Mechanisms of action, physiological effects, and
complications of hypothermia.” Critical care medicine 37 (7 Suppl) (July): S186-202.
Premkumar, L S, M Raisinghani, S C Pingle, C Long, and F Pimentel. 2005.
“Downregulation of transient receptor potential melastatin 8 by protein kinase C-
mediated dephosphorylation.” J Neurosci 25 (49): 11322-11329.
Prevarskaya, N, L Zhang, and G Barritt. 2007. “TRP channels in cancer.” Biochim Biophys
Acta 1772 (8): 937-946.
Proudfoot, C J, E M Garry, D F Cottrell, R Rosie, H Anderson, D C Robertson, S M
Fleetwood-Walker, and R Mitchell. 2006. “Analgesia Mediated by the TRPM8 Cold
Receptor in Chronic Neuropathic Pain.” Curr Biol 16 (16): 1591-1605.
Reid, Gordon. 2005. “ThermoTRP channels and cold sensing: what are they really up
to?” Pflügers Archiv : European journal of physiology 451 (1) (October): 250-63.
126
Roberson, D P, Alexander M Binshtok, F Blasl, B P Bean, and C J Woolf. 2011. “Targeting
of sodium channel blockers into nociceptors to produce long-duration analgesia: a
systematic study and review.” British journal of pharmacology 164 (1) (September):
48-58.
Rohacs, T, C M Lopes, I Michailidis, and D E Logothetis. 2005. “PI(4,5)P2 regulates the
activation and desensitization of TRPM8 channels through the TRP domain.” Nat
Neurosci 8 (5): 626-634.
Ross, Sarah E, Alan R Mardinly, Alejandra E McCord, Jonathan Zurawski, Sonia Cohen,
Cynthia Jung, Linda Hu, et al. 2010. “Loss of inhibitory interneurons in the dorsal
spinal cord and elevated itch in Bhlhb5 mutant mice.” Neuron 65 (6) (March 25):
886-98.
Ruskin, David N, Rene Anand, and Gerald J LaHoste. 2007. “Menthol and nicotine
oppositely modulate body temperature in the rat.” European journal of
pharmacology 559 (2-3) (March 22): 161-4.
Sabnis, A S, M Shadid, G S Yost, and C A Reilly. 2008. “Human lung epithelial cells
express a functional cold-sensing TRPM8 variant.” Am J Respir Cell Mol Biol 39 (4):
466-474.
Saito, S, and R Shingai. 2006. “Evolution of thermoTRP ion channel homologs in
vertebrates.” Physiol Genomics 27 (3): 219-230.
Sawada, Y, H Hosokawa, A Hori, K Matsumura, and S Kobayashi. 2007. “Cold sensitivity
of recombinant TRPA1 channels.” Brain Res 1160: 39-46.
Seebacher, F, and S A Murray. 2007. “Transient receptor potential ion channels control
thermoregulatory behaviour in reptiles.” PLoS ONE 2: e281.
Shimada, Steven G, and Robert H LaMotte. 2008. “Behavioral differentiation between
itch and pain in mouse.” Pain 139 (3) (October 31): 681-7.
Slominski, A. 2008. “Cooling skin cancer: menthol inhibits melanoma growth. Focus on
‘TRPM8 activation suppresses cellular viability in human melanoma’.” Am J Physiol
Cell Physiol 295 (2): C293-5.
Stein, Christoph, and Leonie Julia Lang. 2009. “Peripheral mechanisms of opioid
analgesia.” Current opinion in pharmacology 9 (1) (February): 3-8.
Stein, Christoph, and Christian Zo. 2009. “Opioids and Sensory Nerves.” Handbook of
experimental pharmacology 194: 495-518.
127
Stein, R J, S Santos, J Nagatomi, Y Hayashi, B S Minnery, M Xavier, A S Patel, et al. 2004.
“Cool (TRPM8) and hot (TRPV1) receptors in the bladder and male genital tract.” J
Urol 172 (3): 1175-1178.
Story, G M, and R W 4th Gereau. 2006. “Numbing the senses: role of TRPA1 in
mechanical and cold sensation.” Neuron 50 (2): 177-180.
Story, G M, A M Peier, A J Reeve, S R Eid, J Mosbacher, T R Hricik, T J Earley, et al. 2003.
“ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold
temperatures.” Cell 112 (6): 819-829.
Stucky, C L, A E Dubin, N A Jeske, S A Malin, D D McKemy, and G M Story. 2009. “Roles of
transient receptor potential channels in pain.” Brain Res Rev 60 (1): 2-23.
Su, Lin, Chao Wang, Yong-Hao Yu, Yong-Ying Ren, Ke-Liang Xie, and Guo-Lin Wang. 2011.
“Role of TRPM8 in dorsal root ganglion in nerve injury-induced chronic pain.” BMC
neuroscience 12 (January): 120.
Suzuki, S C, H Furue, K Koga, N Jiang, M Nohmi, Y Shimazaki, Y Katoh-Fukui, M
Yokoyama, M Yoshimura, and M Takeichi. 2007. “Cadherin-8 is required for the first
relay synapses to receive functional inputs from primary sensory afferents for cold
sensation.” J Neurosci 27 (13): 3466-3476.
Tajino, Koji, Hiroshi Hosokawa, Shingo Maegawa, Kiyoshi Matsumura, Ajay Dhaka, and
Shigeo Kobayashi. 2011. “Cooling-sensitive TRPM8 is thermostat of skin
temperature against cooling.” PloS one 6 (3) (January): e17504.
Tajino, Koji, Kiyoshi Matsumura, Kaori Kosada, Tetsuro Shibakusa, Kazuo Inoue, Tohru
Fushiki, Hiroshi Hosokawa, and Shigeo Kobayashi. 2007. “Application of menthol to
the skin of whole trunk in mice induces autonomic and behavioral heat-gain
responses.” American journal of physiology. Regulatory, integrative and
comparative physiology 293 (5) (November): R2128-35.
Takashima, Y, R L Daniels, Wendy Knowlton, J Teng, E R Liman, and D D McKemy. 2007.
“Diversity in the neural circuitry of cold sensing revealed by genetic axonal labeling
of transient receptor potential melastatin 8 neurons.” J Neurosci 27 (51): 14147-
14157.
Takashima, Y, L Ma, and D D McKemy. 2010. “The development of peripheral cold
neural circuits based on TRPM8 expression.” Neuroscience 169 (2): 828-842.
Thut, P D, D Wrigley, and M S Gold. 2003. “Cold transduction in rat trigeminal ganglia
neurons in vitro.” Neuroscience 119 (4): 1071-1083.
128
Tsavaler, L, M H Shapero, S Morkowski, and R Laus. 2001. “Trp-p8, a novel prostate-
specific gene, is up-regulated in prostate cancer and other malignancies and shares
high homology with transient receptor potential calcium channel proteins.” Cancer
Res 61 (9): 3760-3769.
Tsukimi, Y, K Mizuyachi, T Yamasaki, T Niki, and F Hayashi. 2005. “Cold response of the
bladder in guinea pig: involvement of transient receptor potential channel,
TRPM8.” Urology 65 (2): 406-410.
Tsuruda, P R, D Julius, and D L Minor Jr. 2006. “Coiled coils direct assembly of a cold-
activated TRP channel.” Neuron 51 (2): 201-212.
Vanden Abeele, F, A Zholos, G Bidaux, Y Shuba, S Thebault, B Beck, M Flourakis, Y
Panchin, R Skryma, and N Prevarskaya. 2006. “Ca2+-independent phospholipase
A2-dependent gating of TRPM8 by lysophospholipids.” J Biol Chem 281 (52): 40174-
40182.
Voets, T, G Droogmans, U Wissenbach, A Janssens, V Flockerzi, and B Nilius. 2004. “The
principle of temperature-dependent gating in cold- and heat-sensitive TRP
channels.” Nature 430 (7001): 748-754.
Voets, T, G Owsianik, A Janssens, K Talavera, and B Nilius. 2007. “TRPM8 voltage sensor
mutants reveal a mechanism for integrating thermal and chemical stimuli.” Nat
Chem Biol 3 (3): 174-182.
Wang, Sen, Jongseok Lee, Jin Y Ro, and Man-Kyo Chung. 2012. “Warmth suppresses and
desensitizes damage-sensing ion channel TRPA1.” Molecular pain 8 (1) (March 29):
22.
Wei, E T. 1981. “Pharmacological aspects of shaking behavior produced by TRH, AG-3-5,
and morphine withdrawal.” Fed Proc 40 (5): 1491-1496.
Wei, E T, and D A Seid. 1983. “AG-3-5: a chemical producing sensations of cold.” J Pharm
Pharmacol 35 (2): 110-112.
Weil, A, S E Moore, N J Waite, A Randall, and M J Gunthorpe. 2005. “Conservation of
functional and pharmacological properties in the distantly related temperature
sensors TRVP1 and TRPM8.” Mol Pharmacol 68 (2): 518-527.
Wilson, Sarah R, Kristin A Gerhold, Amber Bifolck-Fisher, Qin Liu, Kush N Patel, Xinzhong
Dong, and Diana M Bautista. 2011. “TRPA1 is required for histamine-independent,
Mas-related G protein-coupled receptor-mediated itch.” Nature neuroscience 14
(5) (May): 595-602.
129
Wong, Gilbert Y, and Narender R Gavva. 2009. “Therapeutic potential of vanilloid
receptor TRPV1 agonists and antagonists as analgesics: Recent advances and
setbacks.” Brain research reviews 60 (1) (April): 267-77.
Xing, Hong, Jennifer Ling, Meng Chen, and Jianguo G Gu. 2006. “Chemical and cold
sensitivity of two distinct populations of TRPM8-expressing somatosensory
neurons.” Journal of neurophysiology 95 (2) (February): 1221-30.
Xing, Hong, Jennifer X Ling, Meng Chen, Richard D Johnson, Makoto Tominaga, Cong-Yi
Wang, and Jianguo Gu. 2008. “TRPM8 mechanism of autonomic nerve response to
cold in respiratory airway.” Molecular pain 4 (January): 22.
Yamamoto, Yoshio, Taku Hatakeyama, and Kazuyuki Taniguchi. 2009.
“Immunohistochemical colocalization of TREK-1, TREK-2 and TRAAK with TRP
channels in the trigeminal ganglion cells.” Neuroscience letters 454 (2) (April 24):
129-33.
Yamamura, Hisao, Shinya Ugawa, Takashi Ueda, Akimichi Morita, and Shoichi Shimada.
2008. “TRPM8 activation suppresses cellular viability in human melanoma.”
American journal of physiology. Cell physiology 295 (2) (August): C296-301.

Yang, Xiao-Ru, Mo-Jun Lin, Lionel S McIntosh, and James S K Sham. 2006. “Functional
expression of transient receptor potential melastatin- and vanilloid-related
channels in pulmonary arterial and aortic smooth muscle.” American journal of
physiology. Lung cellular and molecular physiology 290 (6) (June): L1267-76.
Yosipovitch, Gil, Earl Carstens, and Francis McGlone. 2007. “Chronic itch and chronic
pain: Analogous mechanisms.” Pain 131 (1-2) (September): 4-7.
Zakharian, Eleonora, Baskaran Thyagarajan, Robert J French, Evgeny Pavlov, and Tibor
Rohacs. 2009. “Inorganic polyphosphate modulates TRPM8 channels.” PloS one 4
(4) (January): e5404.
Zeitz, Karla P, Nicolas Guy, Annika B Malmberg, Sahera Dirajlal, William J Martin, Linda
Sun, Douglas W Bonhaus, Cheryl L Stucky, David Julius, and Allan I Basbaum. 2002.
“The 5-HT3 subtype of serotonin receptor contributes to nociceptive processing via
a novel subset of myelinated and unmyelinated nociceptors.” J Neuroscience 22 (3)
(February 1): 1010-9.
Zhang, Lei, and Gregory John Barritt. 2004. “Evidence that TRPM8 is an androgen-
dependent Ca2+ channel required for the survival of prostate cancer cells.” Cancer
research 64 (22) (November 15): 8365-73.
130
Zhang, Lei, Sarahlouise Jones, Kate Brody, Marcello Costa, and Simon J H Brookes. 2004.
“Thermosensitive transient receptor potential channels in vagal afferent neurons of
the mouse.” American journal of physiology. Gastrointestinal and liver physiology
286 (6) (June): G983-91.
Zhao, Huan, Leslie K Sprunger, and Steven M Simasko. 2010. “Expression of transient
receptor potential channels and two-pore potassium channels in subtypes of vagal
afferent neurons in rat.” American journal of physiology. Gastrointestinal and liver
physiology 298 (2) (February): G212-21.
Zurborg, Sandra, Brian Yurgionas, Julia A Jira, Ombretta Caspani, and Paul a Heppenstall.
2007. “Direct activation of the ion channel TRPA1 by Ca2+.” Nat Neurosci 10 (3)
(March): 277-279.

131
APPENDIX A: TRPM8 Haploinsufficiency
Introduction
Proper cold sensing is vital to seeking out environments which allow an organism to
maintain its optimal body temperature at the lowest energy cost. In mammals, cold
temperatures are primarily detected by the transient receptor potential melastatin 8
(TRPM8) channel, which is expressed in a small population of sensory neurons. This
channel is a member of the large TRP channel family, which includes the heat- and
capsaicin-sensitive channel TRPV1. In addition to detecting cold, TRPM8 is also
activated by menthol and the synthetic cold-mimetic chemical icilin. As discussed in
previous chapters, TRPM8-null mice are largely insensitive to cold and display
deficiencies in a number of behavioral tests of cold sensation, and increasing evidence
shows that TRPM8-expressing neurons play a role in the detection of both innocuously
cool temperatures as well as cold temperatures that may be painful or harmful to an
animal.
Haploinsufficiency occurs when decreasing the number of functional alleles of a
gene leads to alterations in phenotype. There is evidence that haploinsufficiency may
occur in animals heterozygous for other TRP channels, including TRPC4, TRPML3, and
TRPM1 (Lee et al. 2005; van Aken et al. 2008; van Genderen et al. 2009). In the two
original reports on TRPA1-null mice, heterozygous mice exhibited intermediate
phenotypes in response to TRPA1-specific agonists as compared to wildtype and full
knockout mice (Bautista et al. 2006; Kelvin Y Kwan et al. 2006). Haploinsufficiency can
be observed for sodium channel Na
v
1.1 in humans, leading to severe myoclonic epilepsy
132
of infancy or Dravet syndrome (Bechi et al. 2012). In sensory neurons, mice
heterozygous for either Nrg1 or SynGAP have been shown to exhibit abnormal sensory
signaling in TRPV1-expressing neurons (Canetta et al. 2011; Duarte et al. 2011). Finally,
one group has recently reported alterations in cold sensitivity in humans with mutations
in the TRPM8 gene (Kozyreva et al. 2011).
Here, we investigate whether TRPM8-heterozygous (M8-het) mice exhibit
haploinsufficiency. We find that M8-hets also exhibit a variety of phenotypic effects
from complete deficiency (equivalent to M8KO mice) to intermediate effects in
response to cooling of the skin. Interestingly, this haploinsufficiency is restricted to
somatosensory neurons innervating the skin, as stimulation of visceral neurons leads to
a normal phenotype in M8-het mice. Furthermore, peripheral ganglia from
heterozygous mice (as compared to wildtype mice) express less TRPM8 transcript as
assessed by qPCR. These findings suggest that the TRPM8 haploinsufficiency can affect
sensory experience of cold.
133

Figure A.1: Dorsal horn Fos expression and TRPM8 heterozygosity

A) M8-het mice show 28±2.0 and 28±1.9 activated neurons in the ipsilateral spinal cord
with menthol and icilin paw injections, respectively, which is intermediate to WT and M8KO
levels. WT mice showed 47±1.5 nuclei with menthol or 47±2.6 nuclei with icilin, while
M8KO mice showed 14±1.2 (menthol) and 12±1.5 (icilin) activated neurons. B) No significant
differences were observed in the number of ipsilateral dorsal horn nuclei expressing Fos
between WT, M8-het, and M8KO mice after unilateral stimulation with capsaicin. Students’
t-test vs. WT: **p<0.01, n.s. p>0.05.
134
Decreased Fos expression in M8-het mice
As discussed in Chapter 4, injection of either menthol or icilin into the hindpaw
of an anesthetized wildtype mouse led to the activation of 47±1.5 or 47±2.6 neurons,
respectively, in the ipsilateral dorsal horn spinal cord sections as marked by expression
of the transcription factor Fos (Figure A.1A). M8KO mice showed 14±1.2 activated
neurons with menthol and 12±1.5 activated neurons with icilin. M8-het mice showed
28±2.0 and 28±1.9 activated neurons with menthol and icilin, respectively, values which
were significantly intermediate to both wildtype and TRPM8-KO values (p<0.01).
Control injections with capsaicin yielded similar values of activated neurons in each
genotype (51±4.5 for wildtype, 49±3.7 for TRPM8-heterozygous, and 48±1.3 for TRPM8-
KO mice; Figure A.1B).

Sensory behaviors
Mild evaporative cooling induced by the application of a drop of acetone onto
the mouse hindpaw leads to behaviors which can be reliably scored according to
severity and type (Knowlton et al. 2011). Using this numerical scoring system, wildtype
mice (n=8) give an acetone score of 2.1±0.1, while M8KO mice (n=8) give a response of
1.1±0.1 (Figure A.2A). TRPM8-het mice (n=8) tested with this assay give a response
score of 1.6±0.1, which is significantly different from both wildtype and M8KO mice
(p<0.01), indicating that responses to innocuous cold may depend on the number of
functioning copies of TRPM8.
135

Figure A.2: Evaporative cooling and cold plate and TRPM8 heterozygosity

A) M8-het mice show intermediate responses to innocuous cooling by acetone evaporation
as compared to WT and M8KO mice. WT mice score on average 2.1±0.1, M8KO mice score
1.1±0.1, and M8-het mice score 1.6±0.1. B) M8-het mice also show both intermediate and
completely impaired responses to noxious cold (0°C). Wildtype mice exhibit a 6.8±0.9s
latency to flinching and 15.6±2.5s latency to licking of the front paws after being placed on
the cold plate, while M8-het mice take 24.4±7.2s to flinch and 54.5±16.3s to lick, both of
which are significantly longer than WT. M8KO mice take 58±8s to flinch and 73±15.4s to
flinch, only the latter of which is not significantly different from M8-het mice. Students’ t-
test: n.s. p>0.05, *p<0.05, **p<0.01.
136
While evaporative cooling is considered a mild or cool stimulus, we also tested
the animals with a more severe/colder stimulus using the cold plate assay. Placing
wildtype mice (n=8) on a 0°C cold plate led to stereotyped responses of flinching or
shaking of the front paws with a latency of 6.8±0.9s, followed by licking of the front
paws with a latency of 15.6±2.5s (Figure A.2B). M8KO mice (n=10) displayed a severely
impaired response, with flinching latency of 58±8s and a licking latency of 73±15.4s.
M8-het mice (n=15) displayed a flinching latency of 24.4±7.2s, which is significantly
different from both wildtype and M8KO mice (p<0.05 and p<0.01, respectively).
Interestingly, their flinching latency is 54.5±16.3s, which is significantly different from
wildtype (p<0.05), but not from M8KO mice. This may suggest that the flinching
behavior, since it occurs before the licking, is more indicative of thermal sensation
(cold), while the licking behavior is related to thermal nociception (cold pain), however
more studies need to be done in order to accept this categorization. Regardless, it is
clear that M8-het mice display deficiencies in noxious cold behaviors.
We also tested M8-het mice in the two-temperature choice assay with the test
plate set to 30°C, 20°C, 15°C, and 45°C. At 30 °C, M8-het mice spent the same
proportion of the trial time on the test plate (40.2±6.2%; n=11) compared to WT mice
(46.7±5.9%; n=16) and M8KO mice (55.2±5.0%; n=8; Figure A.3A). When the test plate
was set to 20°C, M8-het mice (n=11) spent 25.1±10% of the trial on the colder plate,
which was significantly higher than WT mice (2.3±0.8%; n=16). M8KO mice (n=8) spent
a similar proportion of time on the colder plate as the M8-het mice, spending 45.9±7.5%
137

Figure A.3: Temperature preference and avoidance and TRPM8 heterozygosity

A) M8-het mice spend significantly more time on the cooler test plate when it is set to 20°C
than do WT mice, and significantly less time than M8KO mice when the plate is set to 15°C.
No differences are observed between the three lines when the test plate is set to 30°C or
45°C. B) The rates of plate crossing were normalized so that the rate when the test plate
was set to 30°C was set to 100%. As the test plate is cooled, M8-het mice cross at a rate
that is significantly different from both WT and M8KO mice at 20°C and significantly
different only from M8KO mice when the plate is 15°C. No differences in crossing rate were
seen between all three lines when the test plate was set to 45°C. Error bars omitted for
clarity. Students’ t-tests: vs. WT: *p<0.05, n.s. p>0.05; vs. M8KO @p<0.05 @@p<0.01.
138
of the trial on the 20°C plate. At 15°C, this relationship flipped, in that M8-het mice and
WT mice were not significantly different (7.4±5.1%, n=11 and 1.5±0.7%, n=16,
respectively), while M8KO mice spent significantly more time on the test plate than
both types of mice (20.3±3.1%, n=8). As a control, we tested all three lines of mice
when the test plate was set to 45°C, and found no significant differences between the
lines (WT 14.3±4.3%, n=16; M8-het 10.9±3.4% n=7; M8KO 7.0±1.0%, n=8).
Similarly, we examined the normalized crossing rate to measure cold avoidance
(Knowlton et al. 2011; McCoy, Knowlton, and McKemy 2011). Crossing rates are
determined by comparing the number of crosses at various temperatures to the number
of crosses when both plates were set to 30°C, thus the 30°C condition corresponds to a
crossing rate of 100% (Figure A.3B). When the test plate was lowered to 20°C, wildtype
mice (n= 16) reduced their crossing rate to 10.4±2.9%, while M8-het mice (n=11)
crossed 28.5±10.7%, and M8KO mice (n=7) crossed 112.7±26.6% of neutral conditions.
At this temperature, M8-het mice crossed at a significantly higher rate than wildtype
mice and a significantly lower rate than M8KO mice. When the plate was further cooled
to 15°C, wildtype mice crossed at a rate of 6.5±3.3%, M8-het mice at a rate of
12.8±10.1%, and M8KO mice at a rate of 49.7±7.7%. At this lower temperature, M8-het
mice were no longer different from wildtype mice, yet remained significantly different
from M8KO mice (p=0.003584).
139
Another assay to test TRPM8-dependent chemically-induced behavior is the wet
dog shaking assay, wherein icilin is injected intraperitoneally into the mouse and the
number of “wet dog shakes” measured (Wei and Seid 1983; Colburn et al. 2007; Ajay
Dhaka et al. 2007). Over the first twenty minutes following injection, wildtype mice
(n=8) exhibited 19±3.4 shakes, while M8KO mice (n=8) exhibited only 1±0.5 shakes
(Figure A.4). In a departure from the previous behavioral assays described above, as
well as from one previous study (Colburn et al. 2007), M8-het mice (n=8) exhibited
20±5.8 shakes, a value similar to wildtype animals. The main difference between this
assay and the previous assays listed is that this involves the stimulation of visceral
neurons in either enteric or nodose ganglia, as opposed to peripheral sensory neurons
in the trigeminal or dorsal root ganglia. Previous studies have indicated that the
Figure A.4: Icilin-induced wet dog shakes and TRPM8 heterozygosity

Intraperitoneal injections of 50mg/kg icilin resulted in 19±3.4 shakes over twenty minutes in
wildtype mice, while M8KO mice only exhibited 1±0.5 shakes. M8-het mice shook as often
as WT mice, with 20±5.8 shakes. Students’ t-test vs. WT: n.s. p>0.05.

140
peripheral and visceral sensory systems may differ in their mechanisms of cold sensing,
so the behavioral differences may be due to different TRPM8 expression schemes in
different types of ganglia (Fajardo et al. 2008).

Quantitative assessment of TRPM8 expression in M8-het sensory ganglia
The staining and behavioral data outlined above strongly suggest that cold-
sensitive TRPM8-expressing primary sensory neurons are impaired, albeit indirectly,
since both measures are the result of polysynaptic activity. To test the primary sensory
neurons directly, we harvested trigeminal ganglia from both wildtype (n=3) and M8-het
(n=3) mice and purified RNA transcripts. Using three different primer sets, one which
targets 5’ exons 3-7 (“M8 Probe 1”), one which targets middle exons 10-14 (immediately
before the null mutation was introduced; “M8”), and one which targets 3’ exons 23-27
(“M8 ISHPrim2”), we found marked reductions in TRPM8 expression in M8-het mice
(Figure A.5). All transcript levels were normalized to the mean levels expressed by
wildtype mice, with 100±16.3% for the 5’ primer set, 100±24.5% for the middle primer
set, and 100±9.1% for the 3’ primer set. In contrast, M8-het mice only expressed
21±3.7% of the 5’ primer set, 21.1±6.2% of the middle primer set, and 15.6±6.4% of the
3’ primer set. It is interesting to note that, although the truncated transcript can be
detected in M8KO homozygous mice (Bautista et al. 2007), there is not a detectable
imbalance with these primer sets in the proportion of 5’ transcripts to 3’ transcripts.
141
This would suggest that the truncated transcript is efficiently degraded and this this is
truly a case of haploinsufficiency as opposed to a dominant-negative effect of the
mutant transcript.

Conclusion
TRPM8 has been implicated in both sensing innocuous cool temperatures as well
as noxious responses to cold temperatures and cold-mimetic chemicals, as discussed in
Chapters 3 and 4 (Bautista et al. 2007; Colburn et al. 2007; Daniels and McKemy 2007;
Figure A.5: Decreased TRPM8 expression in M8-het ganglia

TRPM8 transcript levels were probed using qPCR on TG samples from WT (n=3) and M8-het
(n=3) mice. The mean levels of each transcript in WT mice were set to 100% and levels
detected in M8-het samples normalized to these. Using a primer set directed at 5’ exons 3-7
(M8 Probe 1), WT mice expressed 100±16.3% while M8-het mice expressed 21±3.7%. Using a
primer set directed at the middle exons 10-14 of the transcript (M8), which is immediately
upstream of the mutation site, WT mice expressed 100±24.5% while M8-het mice expressed
21.1±6.2%. Using a primer set directed at the 3’ exons 23-27, WT mice expressed 100±9.1%
while M8-het mice expressed 15.6±6.4%. Students’ t-test vs. WT: *p<0.05, **p<0.01.
142
Ajay Dhaka et al. 2007; Knowlton et al. 2010). The finding that mice lacking one
functional allele of TPRM8 display behavioral deficits in assays of both innocuous cool
and noxious cold sensing further reinforces this.
The observation of a differential effect in behaviors to noxious cold (cold plate
test) is particularly intriguing in light of reports on the division of cold-responsive
sensory neurons into two groups: low- and high-threshold (Thut, Wrigley, and Gold
2003; Madrid et al. 2009). One possible method of interpreting the cold plate data is
that the two behaviors represent the activity of these two subpopulations of cold-
sensing neurons, one of which is only partially impaired and the other is fully impaired.
Indeed, in the current model of TRPM8-based cold signaling, the temperature threshold
of the cell is set by the relative currents of TRPM8-based excitation and K
v
1 potassium
channel-based brake inhibition (Madrid et al. 2009; McCoy, Knowlton, and McKemy
2011). The high threshold (HT) population would have a higher TRPM8 to K
v
1 ratio than
the low threshold (LT) population, thus the HT population would begin responding to
warmer temperatures (i.e. smaller decrease from body temperature), while the LT
population would need a much stronger stimulus (i.e. lower temperature) to generate
enough excitatory current to generate neural signals. The data presented here reveal
that activity of cold-sensing cells is affected by the number of functional copies of the
TRPM8 gene, which suggests that one or both of the subpopulations could be
compromised. Future studies investigating the level of TRPM8 expression at the single-
cell level, and on the distribution of temperature thresholds in cells cultured from M8-
het mice will shed more light on this intriguing model of TRPM8-based cold sensing.
143
In addition to previous reports finding that icilin-induced behaviors are TRPM8-
dependent, here we see that injection of icilin into the paw resulted in an intermediate
number of neurons expressing Fos in the spinal cord dorsal horn. Thus it is interesting
to note that stimulation of visceral sensory neurons with icilin did not lead to behavioral
deficits in M8-het mice while M8KO mice were essentially devoid of responses in the
wet dog shaking assay. Clearly, there must be a difference between the i.p. injections
and the plantar injections of the drug. Differences between nodose and peripheral
sensory ganglia have been previously noted in regard to the putative cold-sensing
channel TRPA1 (see Appendix 2), thus it may be that the logic of sensory processing in
the viscera and in the skin is slightly different in regards to TRPM8 (Fajardo et al. 2008).
Additionally, the dose of icilin given i.p. (50mg/kg) is slightly lower than that used in a
previously published study reporting a difference (Colburn et al. 2007). Future
experiments testing different doses, as well as further exploration of the responsiveness
of visceral neurons to cold will clarify this issue.
Finally, it is interesting to note a roughly 80% reduction in TRPM8 transcript
levels in M8-het mice as compared to wildtype mice. This suggests that the regulation
of the transcript is a complex process, and further studies into this phenomenon might
be informative for other processes where TRPM8 transcription may be important (i.e.
with injury; see Chapter 5).
144
APPENDIX B: TRPA1 and Noxious Cold Sensing
Introduction
There remains a significant amount of debate within the field of
thermosensation regarding the identity of the sensor for noxious cold, as discussed in
Chapter 4. The most promising candidate, besides TRPM8, has traditionally been
another TRP channel, TRPA1. TRPA1 is expressed exclusively within a subset of
peptidergic TRPV1-expressing population of neurons, and not in TRPM8-expressing
neurons, which presented an appealing model for the burning feeling experienced at
extremely cold temperatures (Davis and Pope 2002; Story et al. 2003). Initial cloning
studies showed that TRPA1 (then termed ANKTM1) could be activated by cooling below
17°C using heterologous expression systems in calcium-imaging experiments (Story et al.
2003). Interestingly, this initial study also found that TRPA1-mediated current
inactivated quickly with prolonged cooling and was desensitized to multiple bouts of
cooling, puzzling findings in light of sustained TRPM8 activity with cooling, even in light
of adaptation (McKemy, Neuhausser, and Julius 2002).
From these initial reports, a myriad of studies have grown on either side of the
debate of whether TRPA1 is really a noxious cold sensor (McKemy 2005). The matter
was complicated by the finding that TRPA1 operates as a receptor-operated channel (as
in conjunction with the bradykinin receptor, a GPCR) and as a receptor for plant-derived
and environmental irritants (many of them electrophilic compounds) such as
isothiocyanates and acrolein, the pungent ingredients in wasabi and tear gas,
respectively (Bandell et al. 2004; Jordt et al. 2004; Bautista et al. 2005; Bang and Hwang
145
2009). Indeed, one group reported an overlap of responses to mustard oil and cold in
cultured DRG cells, and thus concluded that the cold current must be carried by TRPA1,
although two other groups searched for and found no such overlap (Babes, Zorzon, and
Reid 2004; Bandell et al. 2004; Jordt et al. 2004). Even with the generation of TRPA1-
null mice, the matter was still not resolved, with some groups finding no deficits in
behavioral assays of acute noxious cold sensation, and others significant deficits,
although not complete loss, in similar assays (Bautista et al. 2006; Kelvin Y Kwan et al.
2006; Bautista et al. 2007; Karashima et al. 2009; K Y Kwan et al. 2009; Gentry et al.
2010). Studies with TRPA1-specific antagonists administered to intact animals has led to
similarly conflicting results (Petrus et al. 2007; J. Chen et al. 2011).
146
The preponderance of replicable evidence supporting the claim that TRPA1 is a
sensor of noxious cold comes from heterologous expression studies, wherein the cDNA
coding for TRPA1 is transfected into typically human embryonic kidney (HEK), Chinese
hamster ovary (CHO), or frog oocyte cells (Story et al. 2003; Sawada et al. 2007;
Karashima et al. 2009; del Camino et al. 2010). It is important to note that when we
transfect HEK cells with either the rat or mouse TRPA1 orthologue (but not other TRP
channels; data not shown), we routinely observe high levels of baseline calcium in the
Figure B.1: Increased baseline calcium with TRPA1 expression

HEK cells transfected with the rat TRPA1 construct (solid line) have significantly higher
levels of baseline calcium prior to stimulation at 100 seconds with 20µM mustard oil
(vertical line) as compared to cells transfected with empty pcDNA3 vector (dashed line).
At 96 seconds, immediately prior to the addition of mustard oil, control cells (n=20)
exhibited calcium ratios of 0.50±0.01, while TRPA1-transfected cells (n=21) exhibited
ratios of 0.73±0.03 (Student’s t-test p<0.0001).

147
cells as compared to cells transfected with the empty pcDNA3 vector, even before
application of an agonist (Figure B.1). This would suggest that heterologous expression
of the TRPA1 cDNA on its own does not fully recapitulate physiological expression, and
that perhaps the channel is affected by cofactors or binding proteins which are not
present in these cell types. This observation is important in light of two reports stating
that cold induces a buildup of calcium within the cell independent of TRPA1, and that
internal calcium can activate TRPA1 via an EF-hand domain in the N-terminus (Doerner
et al. 2007; Zurborg et al. 2007). As general enzymatic processes can be slowed by
cellular cooling, activation of the channel by internal calcium buildup is and attractive
explanation to the finding that cold activation of TRPA1 only occurs with slow cooling
(over 100 seconds or more) and not with brief (≤20s) cooling, as the shorter timeframe
would not be sufficient for a buildup of calcium within the cell (Story and Gereau 2006;
Sawada et al. 2007; O Caspani and Heppenstall 2009; K Y Kwan and Corey 2009).
Furthermore, many TRPA1 agonists covalently modify cysteine residues in the channel,
leading to prolonged activation (Macpherson et al. 2007). Thus TRPA1 could act to
amplify calcium signals of cellular stress or damage induced by cooling (Bang and Hwang
2009).
A recent study proposed an interesting resolution to this issue by suggesting that
physiologically-relevant currents carried by TRPA1 can be augmented, but not induced,
by cold (del Camino et al. 2010). That is, if TRPA1 is pre-activated by another agonist
(and there is a very large list of agonists), the channel can be potentiated by cold and
thus contribute to noxious cold sensation (Bang and Hwang 2009). The most
148
straightforward example of this would be in the condition of inflammatory injury, where
the damaged tissue generates a number of inflammatory chemicals, many of which
activate TRPA1 (Bang and Hwang 2009). A common symptom with injury is cold
allodynia (Chapter 5), in which normally innocuous cold temperatures feel painful or
uncomfortable. Since TRPA1 in this condition would be pre-activated by the
inflammatory compounds, cold could potentiate these responses. Indeed, using TRPA1-
specific blockers and TRPA1-null mice, a role for TRPA1 in cold allodynia has been shown
in a number of models of inflammatory and neuropathic injury, as well as for the cold
allodynia experienced with the chemotherapeutic drug oxaliplatin (Obata et al. 2005;
Katsura et al. 2006; Frederick et al. 2007; da Costa et al. 2010; J. Chen et al. 2011;
Nassini et al. 2011).
Despite this possible reconciliation and a study suggesting that TRPM8 could
feasibly account for both low- and high-threshold cold-responsive sensory neurons,
skeptics persist (K Y Kwan and Corey 2009; Madrid et al. 2009). As discussed in Chapter
4, the consensus after the initial reports on TRPM8-null mice was that it was not
involved in noxious cold sensing, that this was instead the realm of TRPA1. Given that
studies of TRPM8- and TRPA1-single-null mice were inconclusive, we set out to
determine whether the results of these studies were confounded by compensation by
the other receptor (Bautista et al. 2006). That is, if we removed both genes, would we
see complete loss of cold responses?

149
TRPM8/TRPA1 Double-Null Mice
In collaboration with Dr. Diana Bautista of the University of California at
Berkeley, we set out to compare the responses of wildtype, TRPM8-null (M8KO), TRPA1-
null (A1KO), and double-null (DKO) mice in assays of cold preference and avoidance,
flinching responses to cold mimetics, and neural activation with both cold mimetics and
cold. We hypothesized that if the channels were compensating for each other in the
single-null mutants, then the DKO mice should exhibit different behaviors than either of
the single-mutant lines.
The first behavioral assay was carried out by the Bautista lab. In a previous study
on TRPM8-nulls, Dr. Bautista had published data comparing wildtype mice and TRPA1-
null mice on the two-temperature choice assay, finding no differences between the two
groups at any temperature tested, even as low as 5°C (Bautista et al. 2007). This was
surprising since TRPM8-null mice show some, albeit not completely normal, preference
for the home plate when the test plate is set to either 10°C or 5°C, and was presumed to
be explained by activity of TRPA1 in cold nociceptors (Figure B.2A). When we removed
TRPM8 to see if this channel was masking a role for TRPA1 in the lowest two
temperatures tested, we found no significant differences between the DKO and M8KO
mice (Figure B.2A). That is, testing the effect of ablating TRPA1 expression in the
context of no TRPM8 did not produce a phenotype beyond what was observed for
M8KO mice. This finding was corroborated by quantifying the results in terms of cold
avoidance rather than preference. Again, M8KO and DKO mice displayed similar deficits

150

Figure B.2: DKO temperature preference and cold avoidance

A) M8KO mice show no temperature preference as low as 15°C, and at 10°C and 5 °C show
some preference for the warmer plate, albeit significantly less than wildtype mice.
Additional removal of TRPA1 (DKO mice) does not lead to any significant differences from
M8KO mice, indicating that TRPA1 is not responsible for the preference of M8KO mice at the
two lowest temperatures tested. B) Similarly, there is no difference in the crossing rates of
M8KO and DKO mice over all temperatures tested, and the crossing rates of both groups
were significantly different from the rate of wildtypes over the temperature range 20-5°C.
Students’ t-test vs. WT *p<0.05, **p<0.01, ***p<0.001.
151
in cold avoidance as compared to wildtype mice (Figure B.2B), indicating that TRPA1
does not participate in either temperature preference or cold avoidance.
TRPA1 has been reported to be activated by the cold-mimetic TRPM8 agonist
icilin in heterologous expression systems (Story et al. 2003), so next we used this
chemical in a paw flinching assay to see, as with mustard oil-induced flinching (Bautista
et al. 2006), if removing TRPA1 led to additional behavioral deficits beyond M8KO mice.
We injected 24µg of icilin into the hindpaw of wildtype, A1KO, M8KO, and DKO mice and
quantified both the number of flinches and the time spent attending to the injected paw
during the twenty-minute observation period. Under this stimulation paradigm, whole
animal behaviors (i.e. “wet dog” shakes) were rarely observed and responses were
largely limited to the injected hindpaw. We found that for paw flinching, A1KO
responses were similar to wildtype responses (mean 40.8±5.8 and 40.8±8.1 behaviors,
respectively; n=6 each group), and that M8KO responses were significantly decreased
from both of these genotypes (5.6±1.9; n=5; p<0.01 compared to wildtype; Figure B.3A).
Furthermore, DKO mice showed deficits which were not significantly different
from M8KO mice (9.6±1.9; n=8; p>0.05 vs. M8KO), indicating that TRPA1 is not
necessary for icilin-induced nocifensive behaviors. Similarly, the amount of time spent
attending to the injected paw followed the same pattern, with wildtype and A1KO mice
spending a similar amount of time (169.5±25.0 and 212.5±45.3 seconds, respectively),
and M8KO and DKO mice spending a similar amount of time which was significantly
lower than wildtype mice (27.2±11.7 seconds and 38.0±9.5 seconds, respectively;
p<0.01 vs. wildtype; Figure B.3B).
152

Figure B.3: Icilin-induced flinching and TRPA1

A) A1KO mice show normal icilin-induced flinching as compared to wildtype mice. Removal
of TRPA1 in the background of no TRPM8 expression (DKO) does not lead to additional
deficits beyond M8KO mice. B) Similarly, the same pattern holds for the amount of time
during the 20 minute observation period. There is no significant difference between A1KO
and wildtype mice, nor do M8KO mice and DKO mice differ, which are both significantly
impaired when compared to wildtypes. Students’ t-test vs. WT: *p<0.05, **p<0.01,
***p<0.001, n.s. p>0.05.
153
Since behavioral responses involve the input of a complex neural circuit involving
peripheral sensory neurons, central processing, and final motor output and may be
influenced by a number of other sensory and internal circuits, we decided to simplify
our investigation and examine neural activation at the level of the first sensory synapse
in the circuit within the spinal cord. The transcription factor c-fos (Fos) has been
successfully used as a marker in the spinal cord of recent, strong stimulation of sensory
neurons (Hunt, Pini, and Evan 1987). As discussed in Chapter 4, intraplantar menthol or
icilin, as well as repeated stimulation of the hindpaw of an anesthetized mouse with 0°C
can induce robust Fos expression in spinal cord dorsal horn neurons at levels
comparable to capsaicin and mustard oil.
We repeated our cold stimulus paradigm and the intraplantar menthol and icilin
injections in A1KO mice (see Chapter 4), finding the average number of Fos-positive
nuclei observed in A1KO animals stimulated with cold, menthol, or icilin was 43.3±1.3
ipsilateral and 15.5±3.5 contralateral (n=4), 40.3±1.0 ipsilateral and 19.0±4.1
contralateral (n=4), or 45.3±2.6 ipsilateral and 16.7±1.5 contralateral (n=3), respectively
(all p<0.001 ipsi. vs. contra.; Figure B.4A). The number of Fos-positive nuclei induced by
cold, menthol, or icilin was indistinguishable from that observed in wildtype animals
(p>0.05). In contrast, mustard oil-evoked Fos expression was significantly reduced in
A1KO mice versus wildtype controls (average 16.0±0.6 ipsilateral nuclei, n=3, p<0.01;
Figure B.4A).
154

Figure B.4: Dorsal horn Fos expression and TRPA1

A) A1KO mice show normal levels of ipsilateral Fos expression in the spinal cord dorsal horn
after unilateral paw stimulation with 0°C cold, menthol (MEN), or icilin as compared to
wildtype mice. A1KO mice do, however, show reduced Fos levels with stimulation with
mustard oil (MO). B) No significant differences were observed in the number of ipsilateral
dorsal horn nuclei expressing Fos between M8KO mice and DKO mice after unilateral
stimulation with either 0°C cold, menthol, or icilin. Students’ t-test vs. WT: n.s. p>0.05,
***p<0.001.
155
The lack of deficits in cold-evoked Fos expression and cold behaviors observed in
A1KO mice may again result from compensation and activity of functional TRPM8
channels, thus we examined cold-evoked Fos expression in DKO mice (Bautista et al.
2006). Stimulation with cold, menthol, or icilin resulted in mean levels of Fos-positive
nuclei as follows: 21.5±1.0 ipsilateral nuclei and 13.0±2.1 contralateral nuclei for cold
stimulation (n=4), 13.5±1.9 ipsilateral nuclei and 13.5±0.6 contralateral nuclei for
menthol stimulation (n=4), and 15.2±1.4 ipsilateral nuclei and 12.3±1.8 contralateral
nuclei for icilin stimulation (n=6; Figure B.4B). In only the cold stimulation group did we
find a significantly higher number of labeled nuclei (p<0.05) on the ipsilateral side
compared to the contralateral side, as was observed before in cold-stimulated M8KO
mice. Moreover, no additional deficits were observed in the DKO animals.

Conclusion
These data show that, at the level of gene expression induced by neural activity
and behavioral responses, TRPM8 is required for acute sensing of cold and cold
mimetics and further supports the findings that TRPA1 serves no role in acute cold
detection in vivo under normal, non-injury conditions. It is interesting to note that, as
was seen in Chapter 5, M8KO and TRPM8-ablated mice still show an increase in cold
hypersensitivity in both inflammatory and neuropathic pain conditions, and that TRPA1
has been proposed to mediate a certain portion of cold hypersensitivity in these injury
models. Future experiments investigating the effect of either selective pharmacological
blockade and/or genetic disruption of both of these channels on the development and
156
symptomology of these models would be interesting and could provide insight in the
molecular mechanisms of pathology and potential symptom relief.

Asset Metadata

Creator Knowlton, Wendy Michelle (author)

Core Title The roles of TRPM8 in cold sensation: the six sides of TRPM8

Contributor Electronically uploaded by the author (provenance)

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Neuroscience

Publication Date 06/01/2013

Defense Date 05/02/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag behavior,cold,neurobiology,OAI-PMH Harvest,TRPM8

Language English

Advisor McKemy, David (committee chair), Bottjer, Sarah W. (committee member), Butler, Samantha (committee member), Ko, Chien-Ping (committee member), Maxson, Robert E., Jr. (committee member)

Creator Email wendy.knowlton@gmail.com,wknowlto@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-43771

Unique identifier UC11289388

Identifier usctheses-c3-43771 (legacy record id)

Legacy Identifier etd-KnowltonWe-874.pdf

Dmrecord 43771

Document Type Dissertation

Rights Knowlton, Wendy Michelle

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract Our ability to detect temperatures is conferred by thermally-activated sensory neurons of the peripheral nervous system which innervate the skin, with cell bodies located in the dorsal root and trigeminal ganglia. These thermosensitive neurons express thermally sensitive ion channels of the transient receptor potential (TRP) family, including TRPV1, which responds to hot temperatures as well as the ‘hot’ ingredient in chili peppers, capsaicin, and TRPM8, which responds to cold temperatures and the ‘cool’ ingredient in mint, menthol. Beyond thermosensation, these channels have been implicated in a variety of functions including pain, thermoregulation, pain relief, and itch. ❧ Here we examine the roles of the cold-sensitive channel TRPM8 in sensory behavioral responses. We use three main approaches in this study: the disruption of the Trpm8 gene in mice, the ablation of TRPM8-expressing cells in adulthood using BAC transgenesis and the simian diphtheria toxin receptor transgene, and administration of the potent TRPM8 antagonist PBMC. ❧ We found six behaviors are affected by TRPM8 expression using one or more of these methods. These behaviors fall into two general categories: active, or behaviors where activation of TRPM8-expressing cold-sensitive neurons generates to behavioral responses, and inhibitory, where the activation of TRPM8 cells leads to the inhibition of behavioral responses. Active behaviors include responses to innocuous cool, noxious cold, thermoregulation, and cold hypersensitivity after injury. Inhibitory behaviors include inhibition of pain by mild cooling and the inhibition of itch. ❧ The data presented here expand the previously established role of TRPM8 as an in vivo sensor of innocuous cold into a channel involved in a variety of behaviors, including some which are intuitively opposed (i.e. pain and analgesia). Furthermore, we have found a previously unappreciated role for TRPM8 in the inhibition of itch, a finding which will surely create new avenues of drug development for the treatment of chronic itch conditions. Altogether, these studies suggest that the population of sensory neurons expressing TRPM8 is quite diverse, and that understanding this diversity may someday enable us to pharmacologically manipulate specific subsets, leading to novel treatments for various sensory-related pathologies as well as a better understanding of the logic of thermosensation and pain.