TRPV3 IS A POLYMODAL RECEPTOR

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in the Graduate School of the Ohio State University

By

Craig K. Colton, B.S.

The Ohio State University 2006

Dissertation Committee: Approved by Dr. Michael X Zhu, Adviser

Dr. Tsonwin Hai Adviser Dr. John Oberdick Ohio State Biochemistry Graduate Program Dr. Chien-liang Glenn Lin

Dr. Thomas Kasulis

ABSTRACT

The vanilloid family of transient receptor potential (TRPV) channels contains four thermosensitive members (TRPV1-4), each having a distinguishable temperature threshold (Th). In addition to thermal stimuli, TRPV1, or capsaicin receptor, has been shown to respond to acidic pH, exogenous ligands (capsaicin and resiniferatoxin), and endogenous substances such as anandamide and lipooxygenase products. Furthermore,

TRPV1 channel activity is modulated by G-protein coupled receptors (GPCRs) such as bradykinin (Bk) and purinergic (P2Y) receptors. The polymodal nature of TRPV1 has resulted in the identification of its role in thermal sensation, pain, asthma, urinary incontinence, muscle cramps, and migraine headaches. TRPV3 is the newest member of the thermosensitive TRPV channels. Although TRPV3 has been shown to respond to warm temperatures, it was reported to be unresponsive to low pH and chemical activators of other TRPV family members. Our data shows that TRPV3 is activated by the boron containing compound, 2APB. In addition, we report here that TRPV3 responds to pH 5.6 when overexpressed in HEK293 cells, and that this response is sensitive to PKA. Also,

TRPV3 channel activity is modulated by GPCRs such as bradykinin (B2R), histamine

(H1R), and purinergic receptors. We also show that the activation of TRPV3 by the B2R is dependant on phospholipase C, arachidonic acid (AA), and depletion of phosphatidylinositol 4,5-bisphosphate from the plasma membrane. Interestingly, we

ii show that several other lipid, or lipid-like compounds, modulate TRPV3 channel activity.

We also present evidence that TRPV3 is capable of forming heterotetramers with

TRPV1. The similarities in the properties of TRPV3 and TRPV1, as well as the ability of

TRPV3 to form heterotetramers with TRPV1, suggests that TRPV3 may be involved in many of the same physiological processes as TRPV1. We therefore conclude that

TRPV3 is a polymodal receptor, capable of responding to thermal stimuli, GPCRs, acidic pH, as well as lipid or lipid-like agonists, and may be involved in physiological processes similar to TRPV1.

iii

Dedicated to my parents and sister

iv

ACKNOWLEDGMENTS

I am greatly indebted to my adviser, Mike X. Zhu, for constant guidance, support, encouragement and patients throughout my time in his lab.

I would like to thank my General Exam committee members, Dr. Tsonwin Hai, Dr. John

Oberdick, Dr. Arthur Strauch III and Dr. Mike Zhu.

I would like to thank my Dissertation committee members, Dr. Tsonwin Hai, Dr. John

Oberdick, Dr. Glenn Lin and Dr. Mike Zhu for their efforts in supporting me through my general exam.

I am particularly appreciative of the efforts by Dr. Hong-Zhen Hu and Rui Xiao for the electrophysiological support of my experiments.

I feel privileged to have worked with Chumbo Wang, Hong-Zhen Hu, Rui Xiao and Jisen

Tang and Dina Zhu.

Finally, I could not have completed my research without the support of my friends and family.

v

VITA

July 11, 1971……………………………...…… Born---Altuna, Pennsylvania 1989…………………………………………… Graduate high school Girard High School 1989-1992……………………………………… United States Army 1992-present…………………………………… Ohio Army National Guard 1993-1998……………………………………… B.S. Biochemistry B.S. Microbiology The Ohio State University 2001…………………………………………… Graduate of Officers Candidate School, Army 2004-2005……………………………………… United States Army Deployed to Kosovo 2005…………………………………………… Ohio National Guard Deployed to MS and LA in support of Hurricane Katrina relief effort

PUBLICATIONS

Colton C & Zhu MX. (2006) 2APB activates TRPV1, TRPV2 and TRPV3. Handbook of Experimental Pharmacology, in press

Xiao R, Tang J, Wang C, Colton CK, Tian J, Zhu MX. (2006) Calcium plays a central role in the sensitization of TRPV3 channel to repetitive stimulations. Journal of General Physiology, in revision.

Hu HZ, Xiao R, Wang C, Gao N, Colton CK, Wood JD, Zhu MX. (2006) Potentiation of TRPV3 channel function by unsaturated fatty acids. J Cell Physiol. 208(1): 201-12.

Hu HZ, Gu Q, Wang C, Colton CK, Tang J, Kinoshita-Kawada M, Lee LY, Wood JD, Zhu MX (2004) 2-aminoethoxydiphenyl borate is a common activator of TRPV1, TRPV2, and TRPV3. J Biol Chem. 279(34): 35741-8.

Wang C, Hu HZ, Colton CK, Wood JD, Zhu MX. (2004) An alternative splicing product of the murine trpv1 gene dominant negatively modulates the activity of TRPV1 channels. J Biol Chem. 279(36): 37423-30.

vi

Nguyen KT, Hu X, Colton C, Chakrabarti R, Zhu MX, Pei D. (2003) Characterization of a human peptide deformylase: implications for antibacterial drug design. Biochemistry. 42(33): 9952-8.

FIELDS OF STUDY

Major Fields: Biochemistry Neuroscience Molecular Biology Calcium fluorometric techniques

vii

TABLE OF CONTENTS

Page

Abstract……………………………………………………………………………… ii

Dedication…………………………………………………………………………… iv

Acknowledgments……………………………………………………………...….. v

Vita…………………………………………………………………………………. vi

List of Tables……………………………………………………………………..... xii

List of Figures……………………………………………………………………… xiii

Abbreviations………………………………………………………………..……. xvii

Chapters:

1. Background and Significance…………………………………………………….. 1 1.1. The TRP superfamily………………………………………………….….. 1 1.2. The TRPV family and thermal responsiveness…………………………… 3 1.3. Thermal TRP channels and temperature sensation……………………….. 4 1.4. Pain management is a growing problem in our society……………….…... 5 1.5. The physiology of pain …………………………………………………… 6 1.6. TRPV1 is the classic example of a thermal and pain receptor……………. 9 1.7. Properties of TRPV3 prior to this research…………….……………….... 12

2. 2APB is a chemical activator of TRPV3……………………………………...….. 16 1.8. Introduction…………………………………………………………….…. 16 1.9. Materials and methods……………………………………………………. 19 1.9.1. DNA constructs, cell culture, and transfections……………...….. 19 1.9.2. Whole cell recordings of HEK293 cells……………………….… 21

viii 1.9.3. Intracellular Ca2+ measurements………………………………… 22 1.9.4. Membrane potential measurements…………………….….…..… 23 1.9.5. cRNA synthesis and expression in Xenopus oocytes………….… 23 1.9.6. Two-electrode voltage clamp…………………...…………….…. 24 2.3. Results………………………………………………………………...... … 25 2.3.1. 2APB induces calcium influx through TRPV3 channels in a dose-dependent manner…………………………………………. 25 2.3.2. Activation of TRPV3 by 2APB is inhibited by RR…….….….… 26 2.3.3. TRPV3 is localized to the plasma membrane……………...….… 27 2.3.4. 2APB elicits membrane depolarization in TRPV3 expressing cells……………………………………………………………..... 29 2.3.5. 2APB and heat elicit similar currents in TRPV3 expressing cell... 30 2+ 2.3.6. TRPV3D641N is less sensitive to Ca -dependent inhibition than wild type TRPV3………………………………………....… 31 2.3.7. TRPV3 is sensitized to repetitive stimulation by 2APB.…..….…. 31 2.3.8. 2APB activates currents in Xenopus oocytes expressing TRPV3.. 32 2.3.9. 2APB activates TRPV1 and TRPV2 at higher concentrations.….. 33 2.4 Discussion…………………………………………………………....…….. 34

3. TRPV3 is modulated by signaling events downstream of G-Protein Coupled Receptors……………………………………………………………...…. 54 3.1. Introduction………………………………………………………………... 54 3.2. Materials and methods………………………………………...…..………. 60 3.2.1. DNA constructs, cell culture, and transfections………....……...... 60 3.2.2. For all experimental procedures, see section 2.2……...………….. 60 3.3. Results………………………………………………………………...…… 60 3.3.1. TRPV3 is activated by GPCRs of the Gq/11 pathway……………... 60 3.3.2. RR and the removal of extracellular Ca2+ reduce the 2+ Gq/11PCR induced rise in [Ca ]i through TRPV3………………... 62 3.3.3. Activation of TRPV3 by Gq/11PCRs causes membrane depolarization…………………………………………………….. 62 3.3.4. Gq/11PCRs potentiates the 2APB response of TRPV3…….………. 63 3.3.5. Signaling events downstream of G(q/11)PCRs modulate TRPV3…. 65 3.3.5.1. The effect of PKA and PKC phosphorylation on the G(q/11)PCR induced activation of TRPV3……………. 65 3.3.5.2. The effect of PLC on the G(q/11)PCR-induced activation of TRPV3…………………………………. 66 3.3.5.3. The effect of polyunsaturated fatty acids on the G(q/11)PCR-induced activation of TRPV3……...…….. 67 3.3.5.4. The effect of PIP2 depletion on the G(q/11)PCR- induced activation of TRPV3………………………... 68 3.4. Discussion………………………………………………………….………. 72

4. TRPV3 channel activity is modulated by acid…………………………….……..… 95

ix 4.1. Introduction…………………………………………………………....…… 95 4.2. Materials and methods…………………………………………………...… 98 4.2.1. DNA constructs, cell culture, and transfections………....……...... 98 4.2.2. Acidification of ECS solution for intracellular Ca2+ measurements using the Flex Station…………………………………………….. 98 4.2.3. Intracellular Ca2+ measurements using fura-2……………...…...... 99 4.2.4. For all other experimental procedures, see Section 2.2………...… 99 4.3. Results…………………………………………………………………...….. 99 2+ 4.3.1. Acid causes an increase in [Ca ]i in TRPV3D641N expressing cells at lower temperatures…………………………………………...… 99 2+ 4.3.2. The acid-induced increase in [Ca ]i in both TRPV3 and TRPV3D641N expressing cells is temperature dependent……….... 100 4.3.3. Acid activates TRPV3 on the plasma membrane……………….... 101 4.3.4. Capsaicin inhibits the acid-induced activation of TRPV3D641N…... 103 4.3.5. Acid activates currents in HEK293 cells expressing TRPV3 and TRPV3D641N………………………………………..…………..… 104 4.3.6. Acid potentiates the 2APB response of TRPV3 transfected cells... 105 4.3.7. Optimal pH for TRPV3 activation…….………………………….. 106 4.3.8. Repetitive acid stimulation results in TRPV3 sensitization…….… 108 4.3.9. Signaling events that affect the acid response of TRPV3………… 110 4.3.10. Mutational analysis of Thr378, a PKA consensus site of TRPV3… 113 4.3.11. Acid-induced activation of other thermoTRPV channels………... 114 4.4. Discussion………………………………………………………………….. 115

5. TRPV1 and TRPV3 form functional heterotetramers…………………………..…. 139 5.1. Introduction………………………………………………………………… 139 5.2. Material and Methods……………………………………………………… 142 5.2.1. DNA constructs, cell culture, and transfections……………….….. 142 5.2.2. For all experimental procedures, see Section 2.2…………...…….. 142 5.3. Results……………………………………………………………………… 142 5.3.1. Coexpression of TRPV1 and TRPV3 increases Ca2+-influx and decreases Ca2+-release in response to capsaicin………………….. 142 2+ 5.3.2. The capsaicin-induced rise of [Ca ]i is similar for TRPV1-TRPV3 transfected or TRPV1/TRPV3 cotransfected cells……………….. 144 5.3.3. Pharmacological comparison of cells transfected with TRPV1, TRPV3, or TRPV1/TRPV3……………………...……… 146 5.3.4. Pharmacological comparison of cells transfected with TRPV1, TRPV1-TRPV3 or TRPV1/TRPV3…………………….. 147 5.3.5. Coexpression of TRPV3 with TRPV1 increases the sensitivity of TRPV1 to acid……………………………...…………………….. 149 5.3.6. The effect of 2APB on the acid response of TRPV1/TRPV3 Heterotetramers…………………………………………………… 150 5.3.7. Coexpression of TRPV3D641N with TRPV1 results in capsaicin- sensitive channels with a reduced sensitivity to RR…………….... 152

x 5.4. Discussion………………………………………………………………….. 153

6. Possible mechanisms of action for the 2APB-induced activation of TRPV3……… 164 6.1. Introduction………………………………………………………………… 164 6.1.1. Chemical activators of TRPV1………………………………….…. 165 6.1.2. Chemical activators of TRPV2…………………………………….. 166 6.1.3. Chemical activators of TRPV4…………………………………….. 167 6.1.4. Chemical activators of TRPM8……………………………...…….. 167 6.1.5. Chemical activators of TRPA1…………………………………….. 168 6.1.6. Chemical activators of TRPV3…………………………………….. 169 6.2. Materials and methods……………………………………………………... 169 6.2.1. DNA constructs, cell culture, and transfections…………..……….. 169 6.3. Results……………………………………………………………………… 170 6.3.1. U73122 and U73433 are chemical activators of TRPV3………….. 170 6.3.2. PDBU is a chemical activator of TRPV3……………………...….. 170 6.3.3. Menthol is a chemical activator of TRPV3………………………… 172 6.3.4. DES is a chemical activator of TRPV3……………………………. 173 6.4. Discussion…………………………………………………………...…….. 173 6.4.1. Chemical activators of TRPV3 can be grouped into several categories……………………………………………………..…… 173 6.4.2. Possible vanilloid-like binding site on TRPV3 and other thermoTRPs…………………………………………………….….. 174 6.4.3. 2APB as an activator of TRPV1………………………………...…. 176 6.4.4. 2APB as an activator of TRPV2…………………………….…...… 178 6.4.5. 2APB as an activator of TRPV3…………………………….…..…. 179 6.4.6. The effects of 2APB analogs on TRPV channels……………...…... 179 6.4.7. Structural considerations for 2APB action………………………… 182 6.4.8. Possible site(s) of action for 2APB……………………………..….. 185 6.4.8.1. Does the activation of TRPV3 by 2APB involve a 2APB binding pocket?…………………………….….. 186 6.4.8.2. Does the activation of TRPV3 by 2APB involve the pore?………………………………………………… 188 6.4.8.3. Does the activation of TRPV3 by 2APB involve a signaling pathway?……………………………………. 189 6.4.8.4. Does the activation of TRPV3 by 2APB involve changes in plasma membrane properties?………….…… 190 6.4.9. Concluding remarks…………………………………………….….. 198

Bibliography………………………………………………………………………….. 213

xi

LIST OF TABLES

6.1 Promiscuity of molecules shown to modulate the thermoTRPs ………..…… 212

xii

LIST OF FIGURES

1.1 Architecture of pain pathway………………………………………………… 5

2.1 Activation of TRPV3 by 2APB……………………………………….…….. 43

2.2 TRPV3 is inhibited by RR…………………………………………………... 44

2.3 Activation of TRPV3 by 2APB is dependent on extracellular Ca2+………… 45

2.4 TRPV3 is localized predominantly to the plasma membrane……………….. 46

2.5 Activation of TRPV3 currents by 2APB……………………………….……. 47

2+ 2.6 TRPV3D641N has a reduced sensitivity to Ca ………………………………. 48

2.7 Sensitization of TRPV3 currents by heat……………………………………. 49

2.8 Sensitization of TRPV3 currents by 2APB……………………………….…. 50

2.9 2APB evoked TRPV3 currents in Xenopus oocytes…………………………. 51

2.10 Activation of thermoTRPVs by 2APB……………………………………..... 52

2.11 Capsaicin inhibits the 2APB-induced activation of TRPV3…………….…… 53

3.1 Regulation of TRPV1 by various signaling pathways………………………………. 79

3.2 Activation of TRPV3 by B2R………………………………………………… 80

3.3 Activation of TRPV3 by H1R………………………………………………… 81

2+ 3.4 Effects of RR and removal of extracellular Ca on the B2R-induced activation of TRPV3………………………………………….…………….. 82

3.5 GPCR-induced activation of TRPV3 leads to membrane depolarization……. 83

3.6 Potentiation of the 2APB-induced activation of TRPV3 by B2R……….……. 84 xiii 3.7 Effects of drugs on the BK-induced activation of TRPV3…………….…….. 85

3.8 Overview of arachidonic acid metabolism………………………………...… 86

3.9 Arachidonic acid and ETYA directly activate TRPV3………………………. 87

3.10 PAO prevents the synthesis of PIP2…………………………………………. 88

3.11 PAO activates TRPV3……………………………………………………….. 89

3.12 Removal of extracellular Ca2+ and RR block the PAO-induced activation of TRPV3…………………………………………………………. 90

3.13 PAO-induced membrane depolarization of TRPV3 transfected cells……….. 91

3.14 PAO potentiates the 2APB response of TRPV3……………………………... 92

3.15 PAO potentiates the B2R and H1R-induced activation of TRPV3……....…… 93

3.16 PSD-95 as an example of the ability of PDZ domain containing proteins to form signaling complexes………………………………………... 94

4.1 Activation of TRPV3 by acid……………………………………...……….….. 121

4.2 Activation of wild type TRPV3 by acid at elevated temperatures…………..…. 122

4.3 Acid activates TRPV3 on the plasma membrane…………………...……….…. 123

4.4 Inhibition of acid-induced TRPV3D641N activity by capsaicin………………….. 124

4.5 Acid evoked TRPV3 currents…………………….…...…………………….….. 125

4.6 Acid potentiates the 2APB response of TRPV3…………………………….….. 126

4.7 Simultaneous addition of acid and 2APB to TRPV3 transfected cells………… 127

4.8 The effect of different pHs on the potentiation of the 2APB response of TRPV3…………………………………………………………………….. 128

4.9 pH dependence of the 2APB-induced currents of TRPV3 transfected cells…… 129

2+ 4.10 Sensitization of the acid-induced rise of [Ca ]i in TRPV3D641N transfected cells…………………………………………………………….… 130

4.11 Acid-induced slow activation/sensitization of TRPV3 currents……………….. 131

xiv 4.12 Acid-induced slow activation/sensitization of V3D641N currents………………. 132

4.13 Identification of signaling events that modulate the acid-evoked response of TRPV3D641N…………………………………………………… 133

4.14 Mutational analysis of Thr382, a PKA consensus site of TRPV3………....…… 134

4.15 Acid-induced activation of other TRP family members……………………….. 135

4.16 Acid potentiates the 2APB response of TRPV2……………………………….. 136

4.17 Potentially important amino acid sequences for the regulation of TRPV3……. 137

5.1 Coexpression of TRPV3 with TRPV1 alters capsaicin responsiveness….……. 157

5.2 Coexpression of TRPV3 with TRPV1 mimics the capsaicin responsiveness of cells expressing the TRPV1-TRPV3 concatamer………………………… 158

5.3 Pharmacological comparison of cells transfected with V1, V3, or V1/V3 at a 1:3 ratio……………………………………………………...…… 159

5.4 Pharmacological comparison of cells transfected with TRPV1, or different ratios of TRPV1/TRPV3……………………………..…………… 160

5.5 Model of TRPV1/TRPV3 heterotetramerization…………………………...… 161

5.6 The acid response of (TRPV3 and TRPV1) mimics the acid response of the TRPV1-TRPV3 concatamer………………………………...………… 162

5.7 Coexpression of TRPV3D641N with TRPV1 results in capsaicin-sensitive channels with a reduced sensitivity to RR………………………………...… 163

6.1 Chemical activators of TRPV1………………………………………………… 199

6.2 Chemical activators of various thermoTRP channels………………………….. 200

6.3 Chemical activators of TRPV3………………………………………………… 201

6.4 Activation of TRPV3D641N by U73122 and U73433………………...………… 202

6.5 Activation of TRPV3D641N by phorbol esters………………………………….. 203

6.6 Activation of TRPV3 by menthol……………………………………………… 204

6.7 Activation of TRPV3D641N by DES…………………………………………….. 205

xv 6.8 Various forms of 2APB and several 2APB analogs...... 206

6.9 Possible binding site or sites of action for 2APB and its analogs on TRP channels...... 207

6.10 Ways in which the membrane can affect membrane protein conformation...... 208

6.11 The effect of dissolving lipohilic molecules into phospholipid bilayers……… 210

xvi

ABBREIVIATIONS

A1, TRPA1

2APB, 2-aminoethoxydiphenyl borate

Arg, arginine

Asp, aspartic acid

4-αPDD, 4α-phorbol 12, 13-didecanoate

AA, arachidonic acid

B2R, type 2 bradykinin receptor

BSA, bovine serum albumin

C(1,3,5,6,7), TRPC(1,3,5,6,7)

2+ [Ca ]i, intracellular calcium concentration cDNA, complimentary deoxyribonucleic acid

CFP, cyanin mutant of green-fluorescence protein

CR, capsaicin receptors

DAG, diacylglycerol

DMEM, Dulbecco’s Modified Eagle’s Medium

DRG, dorsal root ganglia

DS, Drosophila signaxplex

xvii ECS, extracellular solution

EET, epoxy-eicosatrienoic acid

ER, endoplasmic reticulum

ETI, eicosatriynoic acid

ETYA, eicosatetraynoic acid

FBS, fetal bovine serum

FRET, fluorescence resonance energy transfer

GFP, green-fluorescence protein

Glu, glutamic acid

GPCR, G-protein coupled receptor

H1R, type 1 histamine receptor

HEK, human embryonic kidney

HPETE, hydroperoxy-eicosatetraenoic acid

IP3, inositol (1, 4, 5) trisphosphate

IP3R, inositol (1, 4, 5) trisphosphate receptor

I-V, current-voltage relationship

LO, lipoxygenase

M(4,5,7,8), TRPM(4,5,7,8)

2+ [Mn ]I, intracellular manganese concentration

NGF, nerve growth factor

OAG, 1-oleoly-2-acetyl-sn-glycerol

P2Y, purinergic receptor

xviii PAO, phenylarsine oxide

PBS, phosphate buffered saline

PCR, polymerase chain reaction

PDZ, (named for PSD-95, Disc large, and ZO-1)

PKA, phosphokinase A

PKC, phosphokinase C

PCa/PNa, permeability ratio for calcium over sodium

PIP2, phosphatidylinositol 4,5-bisphosphate

PLC, phospholipase C

PLD, phospholipase D

PMA, phorbol-12-myristate-13-acetate

PUFA, poly-unsaturated fatty acid

Q10 values, ten degree Celcius temperature coefficient

RNA, ribonucleic acid

RR, ruthenium red

S(5,6), transmembrane domain(5,6)

SD, standard deviation

Ser, serine

SOCC, store-operated calcium channel

SOCE, store-operated calcium entry

STT, spinothalamic tract neurons

Th, temperature threshold

xix ThermoTRPs, thermal-sensitive TRP channels (V1, V2, V3, V4, M8, A1)

ThermoTRPVs, thermal-sensitive TRPV channels (V1, V2, V3, V4)

Thr, threonine

TRP, transient receptor potential

TRPA, transient receptor potential ANKTM1

TRPC, transient receptor potential canonical

TRPM, transient receptor potential melastatin

TRPML, transient receptor potential mucolyptin

TRPN, transient receptor potential NOMP

TRPP, transient receptor potential polycystin

TRPV, transient receptor potential vanilloid

V(1,2,3,4,5,6), TRPV(1,2,3,4,5,6)

641 V3D641N, TRPV3 with Asp mutated to arginine

VLPO, ventrolateral preoptic

YFP, yellow mutant of green-fluorescence protein

xx

CHAPTER 1

BACKGROUND AND SIGNIFICANCE

1.1 The TRP superfamily

The transient receptor potential (TRP) superfamily of cation channels derive their names from a phototransduction mutant in Drosophila. Genetic and electrophysiological techniques eventually identified a gene coding for a non-selective cation channel, the

Drosophila trp, as the gene responsible for a mutation that resulted in the loss of a sustained current as well as Ca2+-influx upon exposure to constant light (Montell, 2005a).

Since the identification of the Drosophila trp, more than 50 trp homologues have been identified from yeast to man, including 28 human homologues (Nilius and Voets, 2005).

Based on sequence homology, the TRP superfamily is divided into six mamilian subfamilies: TRPC (canonical), TRPV (vanilloid), TRPM (melastatin), TRPML

(mucolipin), TRPP (polycystin), and TRPA (ANKTM1). In addition, the TRPN (NOMP) subfamily has members in fish, fly, and worm. All members of the TRP superfamily form channels and/or ionic transporters composed of tetrameric protein assemblies, with individual TRP subunits containing six putative transmembrane segments (see Fig. 4.17

& Fig. 5.7). The pore of a TRP channel is formed when a portion of the P-loop from

1 each of the tetrameric TRP subunits folds back into the membrane. The P-loop is located on the extracellular face of the membrane between the alpha helical 5th and 6th transmembrane domains, which are believed to hold the pore in place. In fact, TRP channels are architecturally related to the cyclic nucleotide channels and voltage-gated K+ channels whose structure has been determined to atomic resolution (Clapham, 1999).

TRP channels are non-selective cation channels with permeability ratios for calcium over

2+ sodium (PCa/PNa) of < 10, with the Ca -selective TRPV5 (V5) and TRPV6 (V6) as well as the monovalent-selective TRPM4 (M4) and TRPM5 (M5) being the exceptions. TRP channels can be thought of as detectors/regulators of environmental changes on both a cellular and physiological level. On a cellular level, TRP channels play a role in receptor and store-operated Ca2+ entry, cell volume regulation, magnesium homeostasis, osmolality regulation, and trace metal detection (Clapham, 2003). On a physiological level, TRP channels are involved in pulmonary reflexes, Ca2+ uptake in kidney and intestine, pain pathways, synaptic transmission, neurite growth and turning, vascular tone, endothelial permeability, temperature sensation, osmolality regulation, fertilization, hearing, olfaction, and the detection of pheromones, redox states, touch and/or pressure, and flavors (Clapham, 2003; Patapoutian et al., 2003; Moran et al., 2004;). This diversity of function supports the idea that these channels play a role in diverse sensory functions.

In addition, the fact that many individual TRPs respond to multiple signals supports the idea that many of the TRP channels are “polymodal” receptors, with TRPV1 (V1) possibly being the best model. Interestingly, most of the TRP channels are expressed ubiquitously, with many mammalian cell types expressing

2 several different TRPs (Clapham, 2003). With the many potential physiological functions that TRP channels play a role in, deciphering the functions of each of the TRP channels should be an important endeavor in our time.

1.2 The TRPV family and thermal responsiveness

The vanilloid family of TRP channels (TRPV) derives its name from the fact that the founding member of the family, TRPV1 (V1), is activated by binding to capsaicin and resiniferatoxin, two compounds that contain a vanilloid moiety but are otherwise unrelated structurally (Caterinal et al., 1999). Interestingly, like with the other members of the TRP superfamily, the TRPV family members are grouped based on sequence homology, rather than on ligand binding. In fact, when both sequence homology and various functional aspects of the TRPV family members are considered, the family can be split into two separate categories, with V1, TRPV2 (V2), TRPV3 (V3), TRPV4 (V4)

(thermoTRPVs) and V5, V6 forming the different categories (Gunthorpe, 2002). From a protein homology perspective, the human thermoTRPVs vary in protein sequence between 41 and 47%. In contrast, human V5 and V6 are 75% homologous with regards to protein sequence, but only 30-33% homologous with the thermoTRPVs. From a functional perspective, V5 and V6 are not thermoreceptors and are the only known

2+ members of the TRP superfamily to show high selectivity for Ca (PCa/PNa > 100) over monovalent cations. Because of their expression in kidney and intestine, as well as their selectivity for Ca2+, V5 and V6 function is presumably related to calcium homeostasis in these tissues. These channels are not relevant to this thesis and no further description is

3 necessary. In contrast, members of the thermoTRPV group are thermosensitive, with

o temperature thresholds (Th) of >43, >52, >31 and >25 C respectively (Clapham, 2003).

Upon reaching their Th, the thermoTRPVs are activated with ten degree Celcius temperature coefficient (Q10 values) of 30 for V3 and between 10 and 20 for V1 and V4.

Q10 values are generally not listed for V2 because temperatures just past the thermal threshold for V2 will result in destruction of the cells. A simplistic definition of a Q10

o value is the fold increase in activity per 10 C increase in temperature. Q10 values for most channels and enzymes are between 1 and 3.

1.3 Thermal TRP channels and temperature sensation

TRPM8 (M8) and TRPA1 (A1) are also thermal receptors, with Ths < 25 and < 17 oC respectively. Therefore, V1, V2, V3, V4, M8 and A1 are classified as thermoTRPs.

The thermoTRPs are characterized by typical TRP channel cation nonselectivity upon activation. This cation nonspecific property of TRP channels has significant consequences for the peripheral nerve fibres in which all thermoTRPs are expressed.

First, the influx of monovalent cations, in particular Na+, results in a depolarization of the membrane in cells in which the thermoTRPs are expressed. This depolarization can lead to action potential firing in the subset of nerve fibres that a particular thermoTRP is expressed. Since the thermoTRPs have different temperature thresholds associated with them, mammals can sense the temperature of a surface and/or the environment based on the pattern of action potential firing that results from activation of thermoTRPs on the peripheral neurons. Second, Ca2+ can act as a messenger, causing both short term and

4 long-term changes in the nerve fibres after an encountering temperatures capable of activating these channels. This can lead to either a sensitized or desensitized response upon repeated stimulation, thereby providing the motivation to avoid a potentially damaging stimulus and reducing the likelihood of injury in the case of noxious temperature encounters.

1.4 Pain management is a growing problem in our society

A great deal of effort in medicine is focused on avoidance or reduction of pain.

Local and/or general anesthetics are administered before even simple surgeries and pain medication is usually prescribed during periods following surgery and/or injury. In addition to the relief of pain as it goes through the natural progression from injury to recovery, 1/3 of the world population suffers from recurrent pain that is due to permanent injury/illness or to a dysfunctional nervous system, that presumably is the result of a permanent modification following a previous injury; a sort of “long term potentiation” resulting in overly sensitive neurons along the pain pathway, to include peripheral neurons (Stucky, 2001, Woolf & Salter, 2000). The result of this pain can be psychologically devastating to its victims and costs the American public more than $100 billion each year in health care, compensation, and litigation. Exasperating this problem is the growing number of elderly people in our society which will undoubtedly lead to an increase in pain associated with degenerative diseases such as arthritis, cancer and osteoperosis, just to name a few. In light of this, the importance of controlling “normal” and chronic pain is not an option of medical science. In addition, since both “normal”

5 and chronic pain are the result of current injury/disease or sensitization that results from previous injury/disease, respectively, the identification of the receptors and/or other factors in the periphery that contribute to pain is the first step for both the reduction of

“normal” pain and prevention of the chronic and/or inflammatory pain state for medical science.

1.5 The physiology of pain

Since the discovery of V1 as the channel responsible for both thermal sensation in the noxious temperature range as well as the painful burning sensation elicited by the chemical capsaicin, thermal receptors have been viewed as potential pain receptors. In fact, the polymodality of V1, allows this channel to function as an instantaneous detector of noxious heat, while also “reminding” us of recent injuries by being sensitized and sending painful stimuli at normal body temperatures, a phenomenon known as thermal hyperalgesia. Using this line of reasoning, along with the observation of both the thermal responsiveness and the polymodality of V1, V2, V4, M8, and A1, these receptors have been identified and accepted as pain receptors. The role of V1 in pain will be talked about in greater detail in Section 1.6. Being a thermally sensitive channel in the warm to noxious temperature range, V3 is a potential modulator of pain pathways. Accordingly, a brief description of the circuitry comprising pain pathways is in order. This discussion will begin with an overview of the entire pain pathway, followed by a focus on the peripheral nervous system, where painful stimuli originate and is most likely the location

6 that receptors/channels such as V3 play a predominant role in not only pain sensation, but in the development of hypersensitive circuitry throughout the nervous system (peripheral and central) that ultimately results in chronic pain.

The “sensation” of pain is undoubtedly complex and ultimately results from the ability of higher brain centers to integrate: 1) Specific afferents that result from the stimulation of receptors in the periphery. 2) Spacio-temporal inputs of direct labeled lines encompassing activity occurring over the entire body. 3) The physical and emotional state of the body (Craig, 2003). The neuronal architecture that comprises ascending pain pathways can be broken down into three groups that include: 1) Peripheral nerve fibres that originate with the dorsal root ganglion. 2) Spinothalamic tract neurons (STT) that receive information from the peripheral fibres and relay that information to higher brain centers in the thalamus and/or cortex. 3) Higher brain centers that receive information from the STT and process it for the perception of pain (Gottschalk & Smith, 2001) (Fig.

1.1). It is important to note that each of these groups have a plasticity associated with them that can result in a “chronic pain state” following prolonged exposure to painful stimuli that originate in the periphery.

Peripheral sensory fibres are comprised of Aα, Aβ, Aδ, and C-fibers. Each group of A-fibres have differing degrees of myelination while C-fibres are either unmyelinated or very thinly myelinated (Craig, 2003). Aα and Aβ are mechanosensitive fibres that contribute to proprioception, as opposed to nociception, and are not the focus of this discussion. Nociceptors consist of both nonmyelinated C-fibres (0.3-1.5 µm) or minimally myelinated Aδ-fibres (1-5 µm). The exact manner in which different subsets of these fibres in the periphery contribute to various sensory states, including temperature

7 and pain sensation, is unknown, with the most popular (although simplified) theory being that various subsets of peripheral sensory fibers make up various labeled lines of sensory input that is deciphered first in the STT and then by higher brain structures. Regardless of how the peripheral and central nervous system integrate signals into the perception of pain, it is the expression of various nociceptive receptors at the terminals of nociceptors that initiate the process.

In general, a painful stimulus starts with a depolarization of the terminal ends of neural fibers, termed nociceptors, where noxious stimuli (pressure, high temperature, harmful chemicals, tissue/cell damage, ect…) are encountered and can potentially result in tissue damage. Channels/receptors located on these nerve terminals are activated by these stimuli, causing a depolarization of the peripheral neuron in which the channel/receptor is expressed. The proteins that recognize the noxious stimuli are either channels that can directly cause a depolarization, or receptors (typically G-protein coupled receptors) whose activation in the presence of a noxious stimuli or inflammatory mediators, activates signal transduction pathways that in turn modulate channels capable of depolarizing the membrane. These peripheral nerves have set membrane potential thresholds, that once surpassed, result in action potential firing. In fact, nociceptor activation requires a high intensity stimulus for activation to occur. Therefore, noxious stimuli have to surpass a preset threshold before a particular subset (labeled line) of peripheral neurons will begin firing the rapid action potentials that “inform” the STT that a noxious stimulus is present. In fact, chronic and/or inflammatory pain is the result of either sensitized nociceptive receptors or sensitized peripheral neurons (less depolarization required for action potential firing) that respond to noxious stimuli that are

8 below the normal threshold (Woolf and Salter, 2000). It should be noted that sensitization of pain pathways can also occur at the STT, which occurs when central neurons respond to a level of action potential firing from the periphery that is lower than that which is normally required for the activation of neurons in the STT (central sensitization), but this form of sensitization is not the focus of this dissertation.

1.6 TRPV1 is the classic example of a thermal and pain receptor

V1, commonly referred to as the capsaicin receptor (CR), is expressed in capsaicin sensitive sensory afferents and is activated by noxious heat, acidic pH, and the alkaloid irritant capsaicin (Caterina and Julius, 1999; Szallasi, 2001). V1 was cloned as a result of a long standing effort to discover the then putative CR which is restricted to a specific subset of primary sensory neurons that give rise to thin unmyelinated C-fibers.

o These fibers terminate at the skin to sense heat and pain. With a temperature Th of 43 C,

V1 is normally inactive at physiological temperatures. Upon contact of the skin to temperatures >43 oC, tetrameric V1 channels located on the adjacent nerve fibers open to allow a rapid influx of Ca2+ and monovalent cations, resulting in an action potential that moves from the terminus of C-fibers to DRG, where the signal is then sent to the spinal cord, and is eventually perceived as noxious heat (Szallasi, 2002). However, by responding to stimuli such as low pH, bradykinin, ATP, prostaglandins, cysteinyl leukotrienes, and glutamate, all of which are released by surrounding tissues upon injury

o and/or inflammation, the Th of V1 drops to below 37 C (Hwang & Oh, 2002). By integrating these stimuli into a V1 response at normal body temperatures, V1 plays an

9 important role in the sensation of pain around damaged or inflamed tissues. In fact, the most debilitating effect of inflammatory diseases is the chronic pain associated with CR afferents.

The fact that V1 is expressed throughout the neuroaxis, including brain and spinal cord, as well in nonneuronal tissues such as bladder, spleen, heart, and stomach, which are not subjected to significant temperature fluctuations, suggests that V1 is involved in other unknown cellular processes that may have relevance to health and disease (Szallasi,

2002). Moreover, capsaicin therapies are used to treat neuropathic pain, uremic pruritus, bladder overactivity, psoriasis, and muscle cramps. V1 has also been shown to be a potential target for the treatment of rheumatoid arthritis, osteo-arthritis, asthma, and migraines (Hwang & Oh, 2002). An effort to understand the factors leading to CR related disorders, as well as the sensation of temperature and pain, has led to the identification of the PKA- and PKC-dependent phosphorylation sites on V1 (Numazaki et al., 2002; Bhave et al., 2002). Three major phosphorylation sites on V1, including

Ser117, Ser503, and Ser801, have been identified and shown to modulate V1 activity.

Ser503 and Ser801 are phosphorylated by PKC and sensitize V1 by decreasing its Th

(Numazaki et al., 2002). Interestingly, the inflammatory mediators bradykinin, ATP, and cysteinyl leukotrienes have been shown to sensitize V1 in a PKC dependent fashion

(Hwang & Oh, 2002). Ser117 is phosphorylated by PKA and was shown to regulate V1 desensitization (Bhave et al., 2002). Interestingly, Ser117 is constitutively phosphorylated in vivo, with dephosphorylation being necessary for V1 to enter a desensitized state. Evidence suggests that the inflammatory mediators glutamate and prostaglandins may be responsible for regulating V1 desensitization at this site in a PKA-

10 dependent manner (Hu et al., 2002). A common denominator of CR related disorders is that hypersensitivity of capsaicin sensitive nerve terminals leads to action potentials that are transmitted to the central nervous system via sensory afferents. Therefore, the most logical approach to combat CR related disorders is to develop drugs capable of reducing

CR hypersensitivity. Capsaicin therapies accomplish this because CRs become less sensitive to noxious stimuli after repeated capsaicin administration, a phenomenon referred to as “capsaicin induced desensitization” (Szallasi, 2002). The pungent burning sensation associated with capsaicin therapies has resulted in a search for drugs that will mimic the effect of capsaicin induced desensitization, while eliminating the unpleasant sensation of capsaicin therapy.

Although drugs capable of mimicking capsaicin induced desensitization are promising, another approach is to develop drugs that prevent CR sensitization. In fact, glutamate and prostaglandins sensitize V1 by blocking capsaicin-induced desensitization, apparently by the PKA-dependent phosphorylation of Ser117. Other inflammatory mediators sensitize V1 by lowering its Th in a PKC-dependent manner, presumably by phosphorylating Ser503 and/or Ser801. It therefore seems logical that drugs capable of antagonizing the receptors that sensitize V1 would be useful in treating CR related disorders. A better understanding of the factors regulating V1 would aid in the development of such a drug and would further enrich our understanding of temperature sensation. Differences observed when comparing the currents of recombinant V1 to those of native CRs in the DRG exasperate this effort. Although single-channel conductance, current-voltage relationship, and the ion selectivity of recombinant V1 are similar to those of the native capsaicin sensitive fibers, cloned V1 appears to have a

11 different sensitivity to vanilloid ligands, such as capsaicin and resiniferatoxin, than the native channels (Szallasi, 2002). The pharmacological differences of cloned V1 to that of native CRs, and the fact that trpv1-knockout mice exhibit a loss of capsaicin responsiveness without a significant loss of thermal nociception, have led to the search for other thermal receptors and for proteins that associate with V1 and affect its sensitivity to vanilloid compounds, of which other thermoTRP channels are good candidates (Caterina et al., 2000).

1.7 Known properties of TRPV3 prior to this research

V3 is a 95 kDa protein with three ankyrin repeats in the N-terminus, six transmembrane domains, several putative phosphorylation sites, and a C-terminus that contains a PDZ (named for PSD-95, Disc large, and ZO-1) recognition site not found in other TRPV family members (see Fig. 3.16; Fig. 4.17; Fig. 5.7) (Xu et al., 2002, Smith et al., 2002; Peier et al., 2002). Like V1, V3 channels are tetrameric, with the highly conserved amino acids between the fifth and sixth transmembrane domain (P-loop) making up the pore region. The gene encoding human V3 has a similar exon topology as

V1 and is located on chromosome 7, approximately 7.5 kb from V1, possibly explaining the observation that V1 and V3 are coexpressed in DRG, brain, and several nonneuronal tissues (Xu et al., 2002, Smith et al., 2002). While low pH or vanilloid ligands such as capsaicin and resiniferatoxin failed to activate V3 in these studies, similarities in the structure, transmembrane sequence, expression pattern, temperature gating, ion selectivity, outward rectification, and antagonism by the cationic dye ruthenium red (RR)

12 suggest that V1 and V3 not only have similar function, but are also ideal candidates for intrafamily heteromerization (Smith et al., 2002). Although, immunocoprecipitation experiments and an increased sensitivity of V1 to capsaicin when coexpressed with V3 support the hypothesis that V1/V3 heterotetramers form at least a subpopulation of the

CR in the DRG, the existence of functional V1/V3 heterotetramers has not been proven.

Similar to currents in native DRG, V3 channels have large Q10 values, were sensitized upon repeated heating, and showed a marked hysteresis on heating and cooling

(Xu et al., 2002). Temperature gated channels, such as V1 and V3, have 10-degree temperature coefficients (Q10 values) of > 10, meaning they exhibit heat evoked currents that increase exponentially above their Th, causing a rapid influx of current over a small critical temperature range. Most non-temperature gated channels have Q10 values of 1.0 to 3.0 and exhibit heat-evoked currents that increase linearly with increases in temperature. The sensitization upon repeated heating and hysteresis exhibited by V3 reflects multiple modes of regulation. Sensitization of V3 upon repeated heating is different than the sensitization of V1 by inflammatory mediators in that it does not result in a decreased Th, but in an increase in the open probability of V3 channels at a particular temperature. Hysteresis upon heating and cooling refers to the fact that V3 responds differently to increases or decreases in temperature. Although V3 currents increase rapidly over a range of temperatures, deactivation occurs upon a very slight decrease in temperature. In fact, VR3 channels close even if the temperature never falls below the Th of V3. Determining how sensitization upon repeated heating and/or hysteresis of V3 are regulated will not only further improve our understanding of temperature sensation, but, since V3 may form heterotetramers with V1, should provide additional targets for drug

13 therapies aimed at CR related disorders as well. In light of the known mechanisms of V1 regulation, the presence of many putative kinase consensus sites as well as a consensus

C-terminal PDZ recognition site on V3, suggest that kinases such as PKA, PKC, as well as phosphatases such as calcineurin, may be involved in these regulatory processes.

14

spinthalamic tract (STT)

VGIC

Dorsal root ganglion Action potential direction

Peripheral neuron

Skin

. . . . .

TRPV1 TRPV3

Fig. 1.1. Architecture of pain pathway. Left shows that the neuronal architecture that comprises ascending pain pathways can be broken down into three groups that include: 1) Peripheral nerve fibres that originate with peripheral fibres extending from the dorsal root ganglion to the skin. 2) Spinothalamic tract neurons (STT) that receive information from the peripheral fibres and relay that information to higher brain centers in the thalamus and/or cortex. 3) Higher brain centers that receive information from the STT and process it for the perception of pain. Boxes with arrows show drugs used to combat pain and the location in which they are effective. The left diagram was recreated based on Gottschalk & Smith (2001). Right shows the peripheral nerve terminal with TRPV1 and TRPV3 expressed on it. Thermosensitive channels such as these are expressed at these terminals and initiate the response to several painful stimuli by depolarizing the membrane. In response to the membrane depolarization, voltage-gated ion channels (VGIC) initiate action potentials (gray block arrow).

15

CHAPTER 2

2APB IS A CHEMICAL ACTIVATOR OF TRPV3

2.1 Introduction

V3 was first characterized by three independent labs (Xu et al., 2002; Smith et al.,

2002; Pier et al., 2002). Although tissue expression patterns varied between these studies, V3 was consistently shown to be a RR-sensitive, thermal-gated channel with temperature thresholds between 34 and 39 oC. These findings, as well as the high degree of sequence homology, shared chromosomal co-localization, and co-expression in the

DRG of a subset of thinly-myleinated C-fibres, led to the hypothesis that V3 and V1 may share functional properties. In fact, an increased expression of V3 following neuronal injury, as well the observation that V3 channel activity increases well into the noxious temperature range, suggests that V3 may play a role in the detection of warm to noxious temperatures as well as ascending pain pathways (Smith et al., 2002). The recent observations that trpv3-knockout mice show a reduced tendency to migrate toward warm surfaces and show deficiencies in noxious temperature sensation support this role

(Moqrich et al., 2005; Zimmermann et al., 2005). Additionally, the fact that V3 is activated at or near body temperatures and is expressed in the hypothalamus, as well as

16 the identification of warm sensitive channels in the ventrolateral preoptic (VLPO) region of the hypothalamus, suggest that V3 may play a role in body temperature regulation and/or sleep-wake cycles (McGinty and Szymusiak, 2000; Griffin, 2004). Finally, the rather ubiquitous expression of V3 in nonneuronal tissue such as testis, skin, tongue, small intestines, placenta, and stomach suggest that V3 may play a role in many unanticipated cellular and/or physiological functions (Xu et al., 2002; Smith et al., 2002;

Peier et al., 2002). Despite these potential functions of V3, the lack of a reliable chemical agonist for V3 has limited the ability to study this important channel.

Regardless of the similarities between V3 and V1, initial attempts to activate V3 by low pH, capsaicin or resiniferatoxin had failed (Xu et al., 2002; Smith et al., 2002;

Peier et al., 2002). In addition, V3 was not responsive to the V4 specific activators 4-

αPDD or low osmolarity (Liedtke et al., 2000; Xu et al., 2002; Peier et al., 2002; Vriens et al., 2004) or the V2 specific activator insulin-like growth factor-I (Kanzaki, 1999; Xu et al., 2002). These observations led to the conclusion that, although V3 shared thermal- gating properties with V1, V2 and V4, it was not gated by other stimuli known to activate the thermoTRPVs. The lack of a reliable chemical agonist of V3 is problematic because the use of thermal ramp protocols can yield inconsistent results that are hard to interpret; as the rate of increasing temperature used during temperature ramp protocols, the expression systems used, as well as many other experimental parameters, are not standard in the scientific community, resulting in observable differences in V3 thermal gating properties such as Ths, current densities at a particular temperature past the Th, and hysteresis (Xu et al., 2002). In fact, the scarceness of V2 and V3 related studies in the

17 literature relative to those of V1 and V4 could be directly attributed to the lack of a chemical agonist for the former relative to the latter pairs of thermally gated channels.

2-Aminoethoxydiphenyl borate (2APB) was first introduced to the biological community in 1997 as an inhibitor of Ca2+-release through inositol 1,4,5-trisphosphate receptors (IP3Rs) (Maruyama et al., 1997). Shortly after, it was demonstrated that 2APB could also block Ca2+-entry through store-operated calcium channels (SOCCs) (Ma et al.,

2000). This was taken as proof of the conformational-coupling hypothesis, which suggests that SOCCs are activated through direct physical interactions with IP3Rs. Thus, the effect of 2APB on SOCCs was thought to be indirect either as a result of IP3R block or inhibition of a mutual regulatory component (see Parekh, 2005 for review of possible mutual regulatory components). This belief was challenged not only by the observation that 2APB blocked store-operated calcium entry (SOCE) in some cells without affecting

Ca2+-release through internal stores, but by the demonstration that 2APB blocked SOCE in a chicken B cell line genetically lacking all three IP3R types (Dobrydneva and

Blackmore, 2001; Prakriya and Lewis, 2001; Diver et al., 2001; Trebak et al., 2002).

Thus, 2APB is recognized as an inconsistent inhibitor of IP3Rs but a rather reliable blocker of SOCCs (Bootman et al., 2002; Lievremont et al., 2005).

Several members of the TRPC family, as well as V6, have been suggested to participate in the SOCEs (Zhu et al., 1996; Yue et al., 2000). As such, some of these channels have been tested for their sensitivity to 2APB in heterologous expression systems. To date, the inhibition by 2APB has been documented for TRPC1 expressed in rat superior cervical neurons (Delmas et al., 2002) and TRPC3, C5, C6, and C7 expressed in HEK293 cells (Trebak et al., 2002, Hu et al., 2004; Xu et al., 2005a; Lievremont et al.,

18 2005). Additionally, the inhibitory effect of 2APB has been shown for ectopically expressed TRPM3, TRPM7, M8, and TRPP2, none of which have been proposed to participate in SOCE (Xu et al., 2005a; Hu et al., 2004; Koulen et al., 2002; Hanano et al.,

2004). In light of these observations, 2APB became recognized as a universal TRP channel blocker (Clapham et al., 2001). However, TRPM2 is unaffected and V5 is only slightly inhibited by 2APB (Nilius et al., 2001; Xu et al., 2005a). More interestingly, the data presented in this chapter demonstrate that 2APB is a chemical activator of V3.

Additionally, we have shown that at higher concentrations V1 and V2, but not V4, V5 and V6, are activated by 2APB.

2.2 Materials and methods

2.2.1 DNA constructs, cell culture, and transfections

All compounds and reagents were purchased from Fisher Scientifics or Sigma unless otherwise stated. 2APB was from Tocris or Cayman Chemical Company.

Lipofectamine 2000 and OptiMEM were from Invitrogen. cDNA for murine V3 was prepared by amplification of reverse-transcribed total RNA from mouse skin using appropriate oligonucleotide primers followed by subcloning into pcDNA3. V3D641N was made by PCR amplification using two sets of primers (0.2 µM): 1) 5’- ggcctgggtaacctgaacatccagcag-3’ and 5’-actagttatctagaggt-3’ (C-terminal product) 2) 5’- atgttcaggttacccaggcctatggtgag-3’ and 5’-tgtcctcatctgggccac-3’ (N-terminal product). The template was pcDNA3-TRPV3 and 0.25 units of Pfu enzyme was used. The reactions

19 were carried out in 50 µl reaction buffer containing 10 mM KCl, 10 mM (NH4)2SO4, 2 mM MgSO4, 20 mM Tris-HCl, pH 8.75, 0.1% Triton X-100, 0.1% bovine serum albumin

(BSA) and 0.2 mM each of dATP, dGTP, dCTP and dTTP. The reactions consisted of one 95 oC incubation for 3 minutes, followed by 35 cycles of 95 oC for 1 minute, 55 oC for 1 minute, and 72 oC for 1 minute, and then terminated by a 10 minute elongation at 72 oC. The PCR product was purified following agarose gel electrophoresis and the C- terminal PCR product was cut with the restriction enzymes NsiI and BstEII and the N- terminal PCR product was cut with BstEII and BamHI. The digested PCR products were subcloned into pcDNA3-TRPV3 that was cut by NsiI and BamHI via three piece ligation.

All sequences were confirmed by DNA sequencing.

HEK293 cells were grown in Dulbecco’s minimal essential medium (DMEM) containing 4.5 mg/ml glucose, and 10% heat-inactivated fetal bovine serum. For intracellular Ca2+- and Mn2+-quenching, as well as membrane potential measurements, the cells were transfected with the desired DNA constructs in the wells of 96-well plates without preseeding using LipofectAMINE 2000 following the protocol provided by the manufacturer. The wells were treated with 20 µg/ml polyornithine for 15 min and rinsed once with Hank’s balanced salt solution without Mg2+ and Ca2+. For each well, 25 ng of plasmid DNA and 0.4 µL of LipofectAMINE 2000 were mixed in 50 µL of OptiMEM and added to the well before the addition of 125,000 cells suspended in 100 µl of the medium without antibiotics. The cells were incubated for 24-28 hours without medium change. The transfection efficiency was about 70% as determined using an enhanced green fluorescence protein expression vector. For whole cell recordings, V3 or the appropriate construct was subcloned into the bicistronic expression vector, pIRES2-

20 EGFP (Clontech). The transfections were performed in 35-mm dishes using

LiptofectAMINE 2000 as described in the manufacturer’s protocol. For expression in

Xenopus oocytes, murine V3 was sublconed into pAGA3 vector.

2.2.2 Whole cell recordings of HEK293 cells

Transfected HEK293 cells were reseeded in 35-mm dishes 1 day after the transfection. Whole cell recordings were performed the following day. Recording pipettes were pulled from micropipette glass (World Precision Instruments Inc, Sarasota

FL) to 2-4MΩ when filled with a pipette solution containing 140 mM CsCl, 0.6 mM

MgCl2, 1 mM EGTA, 10 mM Hepes, pH 7.20, and placed in the bath solution containing

140 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM glucose, and 10 mM

Hepes, pH 7.40. Isolated cells were voltage-clamped in the whole cell mode using an

EPC9 (HEKA Instruments Inc, Southboro, MA) amplifier. Voltage commands were made from the Pulse + Pulse Fit program (version 8.53, HEKA), and the currents were recorded at 5 kHz. Voltage ramps of 100 ms to + 100 mV after a brief (20 ms) step to –

100 mV from holding potential of 0 mV were applied every 0.5 s. The cells were continuously perfused with the bath solution through a gravity-driven multioutlet device with the desired outlet placed about 50 µm away from the cell being recorded. Stock solutions of 2APB were made in DMSO. Ruthenium red (RR) was dissolved in water.

2APB and RR were diluted in the appropriate external solutions to the desired final concentrations and applied to the cell through perfusion. All of the whole cell experiments were performed at room temperature (20-24 oC) unless otherwise noted.

21 Temperature ramps were applied using a CL-100 Bipolar temperature controller connected to a SC-20 dual in-line solution heater/cooler (Warner instruments).

2.2.3 Intracellular Ca2+ measurements

Transiently transfected HEK293 cells seeded in 96-well plates were washed once with an extracellular solution (ECS) containing 140 mM NaCl, 5 mM KCl, 1 mM MgCl2,

1.8 mM CaCl2, 10 mM glucose, and 15 mM Hepes, pH 7.4, and then incubated in 50 µl of ECS supplemented with 2 µM fluo4/AM and 0.05% Pluronic F-127 (both were from

Molecular Probes, Eugene, OR) at 37 oC for 60 min. Probenecid (2mM) was included in all of the solutions to prevent fluo4 leakage from cells. At the end of the incubation, the cells were washed three times with ECS and placed in 80 µL of the same solution.

Intracellular Ca2+ was measured using a fluid handling integrated fluorescence plate reader (Flex Station; Molecular Devices, Sunnyvale, CA). 2APB and other drugs were diluted into ECS at three times the desired final concentrations and delivered to the sample plate by the integrated robotic eight-channel pipettor at the preprogrammed time points. The fluo4 fluorescence was read at an excitation wavelength of 494 nm and emission wavelength of 525 nm from the bottom of the plate at 0.67 Hz. All of the experiments were performed at 32 oC unless indicated otherwise.

22 2.2.4 Membrane potential measurements

Transiently transfected HEK293 cells seeded in 96-well plates were washed three times with ECS before being loaded with 80 µl of a working solution of FLIPR membrane potential dye (Molecular Devices). The working solution was prepared by first diluting the stock dye in 10 ml of ECS. 1 ml aliquots were stored at –80 oC until usage. Aliquots were then thawed and diluted 4-fold in ECS. Membrane potential dye was equilibrated on cells for 30 minutes at 32 oC. As described above for the Ca2+ measurements, fluorescence was read at 535 nm excitation and 565 nm emission while drugs were applied with the integrate robotic eight-channel pipettor.

2.2.5 cRNA synthesis and expression in Xenopus oocytes

V3 in the pAGA3 vector was linearized using HindIII. cRNAs were synthesized using mMessage mMachine reagents and protocols obtained from Ambion (Austin, TX).

The resulting cRNAs were dissolved in diethylpyrocarbonate-treated water. Sexually mature female Xenopus laevis of older than 2.5 years of age were purchased from

Xenopus I, Inc. (Dexter, MI). The frogs were quarantined for at least 2 weeks before being used. For oocyte isolation, small pieces of ovarian lobe were dissected out from anesthetized frogs and shaken gently at 19 oC for 90 min in a solution containing 82.5 mM NaCl, 2 mM KCl, 1mM MgCl2, 5 mM Hepes, pH 7.4, and supplemented with 1 mg/ml collagenase (Worthington Biochem, Lakewood, NJ). Denuded, healthy looking oocytes of more than 1mm in diameter were selected and injected in a volume of 50

23 nl/cell with a total of 5 ng of cRNA. The injected oocytes were incubated at 19 oC for 2 days in sterile Barth’s saline (88 mM NaCl, 1 mM KCl, 0.41 mM CaCl2, 0.33 mM

Ca(NO3)2, 0.82 mM MgSO4, 2.4 mM NaHCO3, 7.5 mM Tris-HCl, pH 7.6, supplemented with 20 units/ml penicillin and 20 µg/ml streptomycin) before experiments were performed. The solution was changed daily.

2.2.6 Two-electrode voltage clamp

The oocytes were placed in a 50 µl chamber that was perfused with a bath solution containing 100 mM NaCl, 2.5 mM KCl, 1 mM MgCl2, 1.5 mM EGTA, 5 mM

Hepes, pH 7.4. The cell was impaled with two intracellular glass electrodes filled with 3

M KCl connected to a TEV-700 two-electrode voltage clamp work station (Warner

Instruments, Hamden, CT). The oocytes were clamped at –40 mV, and the currents were continuously recorded using a chart recorder (Astro Med, Inc., West Warwick, RI) and at the same time digitized at 100 Hz using a PowerLab Data Acquisition System

(ADInstruments, Colorado Springs, Co). 2APB was dissolved in the bath solution at the desired final concentrations and applied to the cells by perfusion. The perfusion temperature was 22 oC unless otherwise stated. Temperature changes were made using a

CL-100 Bipolar temperature controller connected to a SC-20 dual in-line solution heater/cooler (Warner Instruments).

24 2.3 Results

2.3.1 2APB induces calcium influx through TRPV3 channels in a dose-dependent manner

V3 was first characterized as a warm sensitive channel that was not activated by other known activators of TRPV family members (Xu et al., 2002; Smith et al., 2002;

Pier et al., 2002). Therefore, in order to identify a reliable activator of V3, we developed a semi high-throughput screening method by transiently expressing V3, V3D641N, or pcDNA3 in HEK293 cells seeded in 96-well plates and monitored fluo-4 fluorescence using the FLEX-Station. This assay allowed us to screen for the effects of multiple compounds on several cDNAs at one time. Since millions of compounds are available, we chose to limit our search to compounds already known to modulate TRP channels, with an emphasis on TRPV specific compounds. In addition, we mutated Asp641, located at the putative P-loop of V3 to an Asn (V3D641N), in order to facilitate our effort by ensuring that a positive “hit” was specific for V3, as apposed to working through a channel/receptor that is endogenous to the expression system or some other non specific means. This strategy was based on the finding that a D→N mutation at the equivalent position of rat V1 (V1D647N) caused a 10-fold reduction of RR sensitivity (Garcia-

Martinez et al., 2000). Although many compounds that are known to modulate members of the TRP superfamily such as capsaicin (300 µM, V1), capsazipine (50 µM, V1), anandamide (300 µM, V1), arachidonyl dopamine (300 µM, V1), resiniferatoxin (30 µM,

V1), 4α-PDD (30 µM, V4), icilin (30 µM, M8), La3+ (3.0 –1000 µM, C5 and M5), OAG

25 (100 µM, several TRPCs) and thapsigargin (1 µM, several TRPCs) did not cause an increase in fluo4 fluorescence in HEK cells transiently transfected with V3 or V3D641N

(data not shown), 2APB, usually used as a blocker of many TRP family members, activated V3 in a dose-dependent manner at 32 oC (Fig. 2.1). V3 was activated by 2APB between 6.0-500 µM and had an EC50 value of 57.7 µM. The slightly higher intracellular

2+ calcium concentration ([Ca ]i) observed at 500 µM compared to 166 µM is most likely due to an endogenous response of HEK293 cells at 500 µM, as pcDNA3 transfected cells have a slight response at this concentration. Interestingly, at the same temperature,

V3D641N had a left shifted dose-response curve to 2APB with and EC50 value of 23.9 µM.

2.3.2 Activation of TRPV3 by 2APB is inhibited by RR

Like with other thermal sensitive TRPV family members (thermoTRPVs), heat activation of V3 is blocked by 1-10 µM RR (Xu et al., 2002; Smith et al., 2002; Peier et

2+ al., 2002). Therefore, if the 2APB-induced rise of [Ca ]i was due to activation of V3, this phenomenon should be blocked by RR. We therefore applied differing concentrations of RR two minutes prior to the addition of 166 µM 2APB to HEK293 cells transiently expressing V3 or V3D641N with the expectation that V3D641N would

2+ exhibit a reduced sensitivity towards RR. As predicted, the 2APB induced rise in [Ca ]i was blocked by RR in a dose-dependent manner when applied to V3 but not V3D641N transfected cells (Fig. 2.2). IC50 values were not calculated because RR concentrations above 30 µm interfered with fluorescence measurements and were not included, resulting in an incomplete dose-response curve. In addition, although the 2APB induced rise in

26 2+ [Ca ]i did not appear to be completely blocked by 30 µM RR, V3 channels located on the plasma membrane may be completely blocked by 10 µM RR, with the additional

2+ decrease in [Ca ]i resulting from RR crossing the plasma membrane and blocking V3 localized on the endoplasmic reticulum (ER) or some other calcium storage organelle

(Fig. 2.2B). These data not only demonstrate that 2APB is a chemical activator of heterologously expressed V3, but that V3D641N forms a functional, yet RR-insensitive, V3 channel.

2.3.3 TRPV3 is localized to the plasma membrane

Although most TRP channels are expressed predominantly on the plasma membrane, large amounts of V1 have been shown to localize to the endoplasmic reticulum (ER) and/or other intracellular sites when ectopically expressed in some cell types, including HEK293 cells (Hellwig et al., 2005). In addition, 2APB has been hypothesized to inhibit IP3Rs by traversing the plasma membrane and binding to native channels located on the ER (Maruyama et al., 1997). We therefore felt it was important

2+ 2+ to determine the degree in which the 2APB induced rise in [Ca ]i was the result of Ca - influx through V3 localized on the plasma membrane versus Ca2+-release through V3 localized on the ER or some other intracellular Ca2+ store. To test this, HEK293 cells transiently transfected with V3, V3D641N, or pcDNA3 were exposed to 166 µM 2APB in either the presence or absence of extracellular Ca2+ (Fig. 2.3A). Indeed, 2APB was able to induce a fluo4 fluorescence increase in the absence of extracellular Ca2+, possibly by traversing the membrane and activating V3 located on the ER or some other Ca2+-store.

27 It is also possible that Ca2+ leaking from the cells accumulates in the external solution to a concentration high enough to sustain detectable levels of Ca2+-influx in response to activation of V3 on the plasma membrane. Regardless of the source of Ca2+ when using a nominally Ca2+-free extracellular solution, a more robust fluorescence increase was observed in the presence of extracellular Ca2+, indicating that the majority of the 2APB induced fluo4 fluorescence increase was due to the influx of Ca2+ from the extracellular space. These data suggest that 2APB activates V3 predominantly on the plasma membrane.

To further verify that 2APB activates V3 localized on the plasma membrane,

Mn2+-quenching was performed on fura-2 loaded HEK293 cells transfected with V3,

2+ 2+ V3D641N, or pcDNA3. Mn can enter the cell via Ca -permeable channels from the

2+ extracellular space. Although the fura-2 fluorescence is unaffected by changes in [Ca ]i at the isosbestic wavelength of 357 nm, Mn2+ can bind fura-2 and quench its

2+ 2+ fluorescence, providing a direct readout of the intracellular Mn concentration ([Mn ]i).

The addition of 166 µM 2APB resulted in a rapid influx of Mn2+ for both the V3 and

V3D641N transfected cells, but not pcDNA3 transfected cells (Fig. 2.3B). Interestingly,

V3D641N transfected cells displayed a faster fura-2 quenching than the V3 transfected cells, presumably due to the higher sensitivity of V3D641N to 2APB. Another possibility is that V3D641N may be more permeable to divalent cations.

V3 localization studies confirm that V3 is localized predominantly to the plasma membrane of HEK293 cells. Fluorescence signals of V3-GFP fusion proteins were largely localized to the plasma membrane, with a less significant signal found in the cytoplasm (Fig. 2.4, top, left). This localization pattern was not due to the GFP portion

28 of the protein, as expression of GFP protein in HEK293 cells is localized predominantly in the cytoplasm (Fig. 2.4, top, middle). Immunostaining V3 transfected HEK293 cells using a rabbit anti-V3 antibody confirms this localization pattern (Fig. 2.4, top, right).

Interestingly, V3 also localizes to an unidentified intracellular organelle/structure (see arrows in Fig. 2.4), possibly explaining the 2APB induced Ca2+-release in the absence of extracellular Ca2+ that was observed above. Immunostaining of cultured rat DRG neurons with V1 (Fig. 2.4, bottom, left) or V3 (Fig. 2.4, bottom, middle) antibodies shows not only that both of these proteins localize to the plasma membrane in this native tissue, but that they are coexpressed in a subset of DRG neurons (Fig. 2.4, bottom, right).

It should be noted that V3 is also expressed on the plasma membranes of keratinocytes, another tissue in which V3 is believed to have function (Peier et al., 2002; Chung et al.,

2004b; Chung et al., 2005; Xu et al., 2006). These observations demonstrate that, like in the native environment, V3 is localized predominantly on the plasma membrane when expressed in HEK293 cells, verifying this expression system is a good model of V3 in the natural state.

2.3.4 2APB elicits membrane depolarization in TRPV3 expressing cells

When activated by heat, V3 is permeable to both Ca2+ and monovalent cations

(Xu et al., 2002). Therefore, application of 2APB to V3 transfected cells should result in

2+ both Ca -influx and membrane depolarization. To test this, V3, V3D641N, or pcDNA3 transfected HEK293 cells were loaded with both the FLIPR membrane potential-sensitive dye and fluo4 and monitored for membrane depolarization and Ca2+-influx by alternating

29 between appropriate wavelengths (Fig. 2.5A). 166 µM 2APB resulted in both Ca2+-influx and membrane depolarization for V3 and V3D641N transfected cells but not the pcDNA3 controls, demonstrating that 2APB should be capable of eliciting action potentials and

Ca2+-signaling events in the subset of neurons which it is expressed. Although the effect of membrane depolarization on nonexcitable cells such as keratinocytes is not known,

2+ increases in [Ca ]i has been shown to result in release of neurotransmitters that sensitize peripheral sensory nerve fibres (Xu et al., 2006).

2.3.5 2APB and heat elicit similar currents in TRPV3 expressing cells

Heat activation of voltage clamped HEK293 cells transiently expressing V3 display outwardly rectifying I-V curves with a reversal potential around 0 mV (Xu et al.,

2002; Smith et al., 2002; Peier et al., 2002). To test whether activation of V3 by 2APB elicits currents similar to those of heat, we performed whole cell patch clamping experiments on HEK293 cells transiently expressing V3 and V3D641N. Like the

2+ fluorometric Ca data, 2APB dose-dependently activated V3 and V3D641N expressing cells (Fig 2.5B & C), but not cells transfected with pcDNA3 (not shown). Similar to heat activation, the currents displayed outward rectification and a reversal potential of approximately 0 mV. Interestingly, at 100 µM 2APB, and to a greater extent at 300 µM

2APB, V3D641N transfected cells showed a reduced rate of inactivation and displayed I-V curves that were near linear with reversal potentials of approximately 0 mV.

30 2+ 2.3.6 TRPV3D641N is less sensitive to Ca -dependent inhibition than wild type TRPV3

A loss of outward rectification at high 2APB concentrations in V3D641N expressing cells suggests that Asp641 on V3 may play a role in Ca2+-dependent inactivation, which has been proposed to contribute to the outward rectification that is typical of TRP family members. To test this, HEK293 cells expressing V3 or V3D641N were stimulated with 100

µM 2APB in a nominally Ca2+-free solution, with increasing concentrations of Ca2+ added after maximal activity was observed (Fig. 2.6). A comparison of Figure 2.6A & B

2+ shows that Ca inactivates V3 to a much greater extent than V3D641N at +100 mV and –

100 mV, with a more pronounced difference at –100 mV. The I-V curves were also affected, as demonstrated by the fact that V3 displayed a more characteristic outward

2+ rectification in the presence of 0.01mM extracellular Ca than V3D641N did at 2 mM extracellular Ca2+. Inactivation dose-response curves for extracellular Ca2+ are shown in

Figure 2.6C, with V3 having an IC50 value of 1.00 +/- 0.20 mM and 0.23 +/- 0.07 mM at

+100 mV and –100 mV, respectively, and V3D641N having an IC50 value of 11.6 +/- 2.3 mM and 7.5 +/- 1.3 mM at +100 mV and –100 mV, respectively, for Ca2+.

2.3.7 TRPV3 is sensitized to repetitive stimulation by 2APB

Similar to published recordings of sensory neurons, V3 is sensitized upon repeated heating (Cesare et al., 1999; Xu et al., 2002; Fig. 2.7), a phenomenon that is unique among the temperature sensitive TRPs, as V1, V2, V4, TRPM8 and TRPA1 are

31 desensitized upon repeated temperature challenges. We therefore determined if the activation of V3 by 2APB could be sensitized upon repeated 2APB application (Fig. 2.8).

As shown, the V3 currents became progressively larger as the number of applications of

100 µM 2APB increased. This increase in current amplitude eventually leveled off, resulting in currents similar to those seen at higher 2APB concentrations (compare Fig.

2.5B, trace d & e to Fig 2.8, trace c). Additionally, prolonged exposure of V3 to 2APB resulted in a time-dependent increase in current amplitude that also reached a peak value and leveled off (data not shown). These data demonstrate that V3 activation by 2APB or heat share an additional property; repeated applications sensitize V3 to subsequent stimulation. This observation is important, as sensitization of V3 to repeated or prolonged stimuli could play a role in skin irritation and/or thermal hyperalgesia, and

2APB can induce this sensitized state.

2.3.8 2APB activates currents in Xenopus oocytes expressing TRPV3

The studies above were performed using HEK293 cells as the expression system.

In order to confirm that activation of V3 by 2APB is not unique to HEK293 cells, we expressed V3 in defoliculated Xenopus ooctyes. Perfusion of 1 mM 2APB using a nominally Ca2+-free solution evoked large currents in V3 expressing oocytes (Fig. 2.9A), but not in uninjected controls (data not shown). Although 100 µM 2APB elicited strong currents in HEK293 transfected cells, very little current resulted when V3 expressing oocytes were perfused with the same concentration of 2APB. This is not unexpected, as

32 it is well documented that higher agonist concentrations are required for receptor/channel activation in oocytes when compared to mammalian cells such as HEK293 (Hu et al.,

2005).

A synergistic effect is observed when capsaicin and heat are administered to V1 simultaneously (Caterina et al., 1999). Since V3 is also a thermally-activated member of the TRPV family, we hypothesized that heat and 2APB may work synergistically to activate V3. Figure 2.9B shows that application of 100 µM 2APB (i), or a rapid increase in the temperature of the perfusate to 39.5 oC (ii), did not result in a significant current.

In contrast, adding both of these subthreshold stimuli simultaneously (iii) resulted in large currents that were greater than 30-fold of either stimulus alone. This synergistic effect is not the result of V3 sensitization, as application of 100 µM 2APB (Fig. 2.9A, third perfusion) did not induce currents following two sequential 1mM 2APB stimulations (Fig. 2.9B). In addition, a 39.5 oC perfusate alone (Fig. 2.9B, iv) did not activate V3 following synergistic activation by 2APB at 39 oC (Fig. 2.9B, iii). These data demonstrate that activation of V3 by 2APB is not limited to HEK293 cells. In addition,

V3 exhibits synergistic activation toward heat and 2APB in a manner reminiscent of heat and capsaicin for V1.

2.3.9 2APB activates TRPV1 and TRPV2 at higher concentrations.

Other studies in our lab revealed that at higher concentrations, 2APB also activates V1 and V2, but not V4, V5, or V6 (Hu et al., 2004). A representative trace of the response to 333 µM 2APB by HEK293 cells transiently transfected with V1, V2, V3,

33 V4, or pcDNA3 is shown in Figure 2.10A. Dose-response curves at 32 oC revealed that

EC50 values of 2APB for V1 and V2 were 114 +/- 8 and 129 +/- 13 µM respectively.

Under these same conditions, V3 and V3D641N had EC50 values of 57.7 and 23.9 µM respectively (Fig. 2.1B). Similar experiments revealed that 2APB and capsaicin act synergistically to activate V1 (Hu et al., 2004). We therefore tested if 2APB and capsaicin also have a synergistic effect on V3. To our surprise, capsaicin inhibited the response of V3 to 166 µM 2APB in a dose-dependent manner (Fig 2.11). Given that RR inhibits all thermoTRPV family members, the identification of an additional inhibitor of

V3 could prove useful in identifying endogenous thermoTRPV channels responsible for physiological functions in native cells.

2.4 Discussion

In these studies, 2APB has been identified as a reliable chemical activator of V3, with activation properties similar to that of heat. To summarize, 2APB activates both

Ca2+-influx and membrane depolarization in HEK293 cells transiently expressing V3, with activation being blocked by RR and high concentrations of capsaicin. Mn2+- quenching and fluo4 florescence studies in the absence and presence of extracellular Ca2+ demonstrate that 2APB activates V3 predominantly located on the plasma membrane.

Also important is the fact that activation of V3 by 2APB mimics electrophysiologic properties of heat activation, resulting in outwardly rectifying currents with reversal potentials around 0 mV. In addition, V3 currents are not only sensitized by repeated stimulation with heat or 2APB, but these two stimuli work synergistically.

34 These studies showed that activation of V3 by 2APB is not unique to expression in HEK293 cells, as 2APB activated currents in Xenopus oocytes as well. The effective concentrations for the 2APB induced V3 activation varies with method used, with progressively higher concentrations needed for HEK293 fluorometric, HEK293 electrophysiological, and Xenopus oocytes measurements respectively. These differences are comparable to variations in temperature thresholds reported for V3 expressed in CHO cells vs HEK293 cells, and may serve as an indication that V3 is regulated by a multitude of cellular factors (Xu et al., 2002). In fact, temperature, cell density, level of TRP channel expression, lipid composition of the plasma membrane, expression of cofactors, and level of channel phosphorylation are just a few factors that have been shown to effect TRP channel sensitivity to agonists (Caterina et al., 1999;

Numazaki et al., 2002; Bhave et al., 2002; Liu et al., 2003; Chung et al., 2005). Although the fluorometric and electrophysiologic methods were both carried out in HEK293 cells, the fluorometric analysis occurs at very high cell density, whereas the electrophysiologic analysis occurs at very low cell densities. This difference may seem minor, but differences in lipid composition and/or activity of signaling pathways may vary depending on cell density. These findings highlight the importance of using very defined condition when studying V3.

There has been a longstanding debate over what governs the “outward rectification” properties of TRP channels. Two hypothesis hold the most clout: 1) outward rectification is the result of an intrinsic voltage sensor within TRP channels, similar to what exists in voltage-gated potassium, sodium, and calcium channels. 2) outward rectification is the result of a divalent cation-induced blockage within the

35 selectivity filter of TRP channels, with the extent of blockage being dependent on membrane potential (voltage across the membrane). Compared to V3, V3D641N was less sensitive to inhibition by RR and Ca2+, more sensitive to heat and 2APB, and displayed a more linear rectification profile. These observations not only suggest that Asp641 of V3 plays an important role in Ca2+-induced inhibition at the channel pore, but that this inhibition may contribute to the rectification properties of V3. This hypothesis inspired a detailed electrophysiological analysis of V3D641N in other studies in our laboratory (Xiao et al., 2006). The findings of this study were summed up as: 1) Asp641 is located within the selectivity filter of V3. Ca2+ binds Asp641 with high affinity and plugs the pore. 2)

Repeated stimulation “sensitizes” V3 by incrementally decreasing the Ca2+ block at this site. This is most likely caused by a conformational change in the channel that results in a change in the structure/properties of the pore. 3) In the cytoplasm, calmodulin interacts with a calmodulin binding site (aa 108-130) and upon repeated channel stimulation shifts the voltage-dependence of the Ca2+ block, thereby complimenting the sensitizing effects felt by the relief of the Asp641- dependent Ca2+ block. From a practical standpoint, the summation of these results demonstrate that under normal conditions, V3 is in a “semi- quiescent” state. Activation of V3 by either short repeated stimulation or prolonged exposure to a stimulant results in an incremental relief of the voltage-dependent Ca2+ block which is observed as a “sensitization” of V3. In other words, V3D641N is already partially sensitized because it is missing a key element that holds V3 in a semi-quiescent state; the electrostatic affinity for Ca2+ that blocks the pore in a voltage-dependent fashion. This observation explains both the increased sensitivity and linearization of the

I-V curve for V3D641N upon stimulation with 2APB (Fig. 2.5B & C).

36 Although V3D641N was successfully used to understand the phenomenon of sensitization as well as the rectification properties of V3, V3D641N should also prove useful in identifying factors that regulate V3 channel activity, especially when used in conjunction with high throughput screening assays. Based on observations of other

TRPV family members, potential modulators of V3 include endogenous ligands, signaling cascades and/or G-protein coupled receptors, pH and changes in osmolarity.

Some of these factors may only modulate V3 in its sensitized state, and may therefore go undetected when the wild type receptor is used for screening. By using V3D641N, the sensitized state can be mimicked effectively without the burden of developing a consistent protocol of V3 sensitization. In addition, many of the modulators of TRP channels affect endogenous channels/receptors in the host cell, making a rapid determination of specificity difficult, especially since RR has antagonistic effects on many channels, which limits its usefulness as a pharmacological identification tool.

Usually, once a positive “hit” is observed, a detailed electrophysiologic analysis is required to insure specificity toward the candidate channel/receptor; a process that is costly, time consuming, and requires great skill. By comparing the RR sensitivity of potential activators in both V3 and V3D641N expressing cells, the specificity of the modulating factor can be determined rapidly. This could prove to be extremely useful considering the polymodality of many TRP family members, especially the TRPV family.

In addition to activating V3, 2APB activates V1 and V2 but not V4, making

2APB the first chemical compound that shares efficacy toward the thermoTRPV family.

Interestingly, there is a direct correlation between the temperature threshold of the

37 thermoTRPVs and the concentration of 2APB required to activate it; heat and 2APB have the same potency order (V3 > V1 > V2). Additionally, the observation that 2APB inactivates the cold sensitive channel, M8, lends further support to the idea that heat and

2APB may activate thermally gated channels by a similar mechanism. Although the observation that V4 is not activated by 2APB seems to contradict this hypothesis, it should be noted that the temperature range for V4 is very narrow, with progressively increasing current amplitudes between 25 and 42 oC, above which, the current is rapidly inactivated (Guler et al., 2002; Patapoutian et al., 2003). Conversely, at temperatures above their respective thresholds, V1, V2, and V3 current amplitudes increase with increasing temperature without reaching a maximum (Xu et al., 2002). Therefore, the lack of response to 2APB by V4 could be the result of V4 inactivation in the presence of strong stimulation. Considering that V2 and V3 are expected to share equal physiological importance as V1, and that V1 plays a role in chronic pain sensation, asthma, disorders of the bladder, and possibly other undiscovered pathologies, understanding the mechanism of action by 2APB on the thermoTRPV channels/receptors could have a profound impact on therapeutic treatments of disorders related to the dysfunction of heat activated TRPV channels. The possible mechanistic actions of 2APB are discussed in chapter six of this dissertation.

The identification of a reliable chemical agonist for V1, V2 and V3 that mimics heat activation has played a key part in distinguishing the roles that V1 and V3 play in specific tissues. In general, the thermoTRPVs are ubiquitously expressed. Despite this, a large majority of studies focus on the involvement of thermoTRPVs in temperature and/or pain sensation. This focus stems from the observation that each of the

38 thermoTRPVs are expressed in sensory neurons, although it should be noted that the expression patterns of V3 have been a matter of debate. For rats, primates and humans,

V3 transcripts and proteins have been found in DRG neurons (Xu et al., 2002; Smith et al., 2002). However, in mouse, the expression of V3 seems to be restricted to skin keratinocytes and the epithelium of the mouth and nose, where V3 works as a warm sensor at the body surface (Peier et al., 2002; Moqrich et al., 2005; Xu et al., 2006).

Regardless of the controversy, it appears that in higher mammalian species, keratinocytes express both V3 and V4, with sensory neurons expressing each of the thermoTRPVs

(Chung et al., 2004b). This observation brings to question the usefulness of 2APB in identifying specific roles of individual thermoTRPV channels in these tissues. Despite these concerns, 2APB has been successfully used to identify tissue specific functions of

V1 and V3 since the publication by our lab that 2APB is a common activator if V1, V2 and V3.

Chung et al (2004a) has tested the response of 2APB application to cultured mouse keratinocytes. Immunostaining with V3 antibodies revealed that V3 was expressed in most of these cells. However, heat-evoked activation of V3-like sensitizing current in these cells was rare (5/189) (Chung et al., 2004a). However, application of

100 µM 2APB at 40oC resulted in outwardly-rectifying currents that were sensitized upon repetitive heat challenges in the majority of the wild type (22/27) and trpv4-/- (23/30) keratinocytes tested (Chung et al., 2004b). Similar to heterologously expressed V3 in

HEK293 cells, 10 µM RR inhibited inward currents evoked by 2APB at 42oC in keratinocytes derived from trpv4 -/- mice. Together, these data confirm that 2APB can sensitize V3 responses to heat in mouse keratinocytes and that it does not require V4.

39 Furthermore, in an independent study, the 2APB-induced activation of native V3, as well as the potentiation of its heat response in mouse keratinocytes was confirmed; unfortunately, whether these responses are missing in trpv3-/- keratinocytes was not reported (Moqrich et al., 2005). Very recently, 2APB has been used to confirm that a single amino acid substitution in V3, V3G573S and V3G573C in mice and rats, respectively, resulted in hypersensitive V3 channels in keratinocytes. Furthermore, it was demonstrated that these mutations in V3 were responsible for the autosomal dominant phenotype of dermatitis and hair loss inherited in mouse (Ds-Nh) and rat (WBN/Kob-Ht) strains (Asakawa et al., 2006). These studies exemplify the usefulness of 2APB in studying the physiological role of V3 in keratinocytes.

Although the high selectivity of V1 agonists like capsaicin has been useful in identifying the physiological roles of V1, the question has remained if other thermoTRPVs compensate for V1 function. As an example, 2APB (30-300 µM) evoked dose-dependent currents in cultured capsaicin-sensitive rat pulmonary neurons that were sensitive to RR and to a lesser extent capsazipine (Gu et al., 2005). In addition, intravenous bolus injection of 2APB or capsaicin elicited pulmonary chemoreflex responses, characterized by apnea, bradycardia, and hypotention in anesthetized, spontaneously breathing rats. Interestingly, the response to capsaicin and 2APB differ enough to suggest that other thermoTRPVs may be involved in these processes.

Although these data cannot distinguish the relative contributions of V1 and V3 and perhaps V2 in these responses, these studies showed that other thermoTRPVs may play a role in many physiological processes believed to be V1 specific based on observations that the response is effected by capsaicin. Similar studies using 2APB and several TRPV

40 knockout mice could determine the importance of individual TRPV members in the pulmonary chemoreflex response. In fact, by comparing the heat responses of skin- saphenous nerve preparation and cultured DRG neurons from wild type and trpv1-/- mice, Zimmermann et al (2005) showed that the 2APB-induced sensitization to thermal stimulation in mouse C-fibres was a TRPV1-facilitated process. Of course, whether or not other thermoTRPVs play a similar role in humans is still in question considering the observed expression patterns differences between mice and humans that has already been discussed.

Recently, several compounds including carvacrol, eugenol, thymol and ethyl- vanillin have been shown to be V3 specific activators, with the exception of eugenol which appears to have an ability to activate V1 at high concentrations (Xu et al., 2006).

Unlike sensory fibres or cultured DRG from mouse, the sensory fibres of rat, primates and humans appear to express V3. Although knockout studies in rats are not as practical as with mice, the use of these recently identified V3 specific pharmacological activators should prove useful in determining if physiological processes believed to V1 specific are compensated by other thermoTRPVs in rat, primate, or human studies. For example,

2APB-evoked currents have been demonstrated in neurons from rat DRG and nodose/jugular ganglia (Hu et al., 2004; Gu et al., 2005). Although these studies concluded that V1 was responsible for the 2APB-invoked currents, the conclusion was based on the fact that capsazipine completely blocked the capsaicin-induced responses.

By repeating these studies with carvacrol, it can be concluded whether or not V3 plays a role in these processes. The advantage of using 2APB is that the combined effect that

V1, V2 and V3 play in physiological processes can be determined and compared to the

41 effects of specific activators of the other thermoTRPVs. One thing is clear, the expression patterns of thermoTRPVs in peripheral neurons, as well as the role that individual thermoTRPVs play in these neurons, is complicated and will require the use of many pharmacological tools, including 2APB, to decipher the role of individual thermoTRPV family members in native tissues.

42

A pcDNA3 TRPV3 TRPV3 D641N 16 16 16 [2APB] (µM) ) -4 0 12 12 12 2.06 6.17 (x 10 (x

0 8 8 8 18.5 55.6 F-F 4 4 4 167 0 0 0 500 01 201 201 2 Time (min) Time (min) Time (min)

1.0 B 0.8

0.6 TRPV3D641N TRPV3 23.9 µM 57.7 µM 0.4

0.2 pcDNA3

Normalized response 0 6543 -Log [2APB] (M)

Fig. 2.1. Activation of TRPV3 by 2APB. A, dose-dependent activation of pcDNA3, TRPV3D641N and TRPV3 transfected cells by 2APB. Fluo4-loaded cells were stimulated with different concentrations 2APB at 20 sec. B, Least square fits of data points for V3 (open circles), V3D641N (filled circles), and pcDNA3 (open triangles) EC50 values were determined using the Hill equation. Data are averages of triplicate samples 60 seconds after 2APB application.

43

A TRPV3 TRPV3 2APB (166 µM) D641N 2APB (166 µM) No RR 20 20

)

-4 No RR 15 15 10 µM RR

(x10 10 10 0 RR RR

F-F 10 µM RR 5 5 0 0 0123456 0246135 Time (m) Time (m)

B 20 TRPV3 D641N

) -4 15

(x10 0 10 TRPV3

F-F 5

0 7654 -Log [RR] (M)

Fig. 2.2. TRPV3 is inhibited by RR. A, Ruthenium red (RR) block of the 2APB response in TRPV3 and TRPV3D641N cells. Different concentrations of RR were applied at 20 seconds followed by the addition of 166 µM 2APB at 140 seconds. Representative traces are the average of triplicate trials with (red) or without (black) 10 µM RR. B, Least square fits of data points for V3 (filled circles) or V3D641N (open circles). Each point is the average of 3 data points 2 minutes after addition of 2APB.

44

A

B 0.5 mM Mn2+

166 µM 2APB

pcDNA3

4

RFU 2 x 2 x 10 1 min TRPV3

TRPV3 D641N

Fig. 2.3. Activation of TRPV3 by 2APB is dependent on extracellular Ca2+. A, extracellular Ca2+-dependence of the 2APB-induced fluorescence increases in TRPV3 (left) TRPV3D641N (middle), and pcDNA (right) transfected cells. Experiments were performed in the presence (dark blue) or absence (red) of 1.8 mM Ca2+ in the extracellular solution. Data points for each trace are the average for two experiments. B, 2+ The Mn -quenching of fura2 signal by 2APB in cells transfected with TRPV3D641N and TRPV3 but not cells transfected with pcDNA3 confirms that 2APB activates V3. Fura2- loaded cells were monitored using excitation and emission wavelengths of 357 and 510 nm, respectively. RFU, relative fluorescence unit. A & B, traces are the average of triplicate experiments.

45

C

GFP

Fig. 2.4. TRPV3 is localized predominantly to the plasma membrane. Localization of TRPV3 to the plasma membrane. TRPV3-GFP (top, left), GFP-only (top, middle) and TRPV3 transfected HEK293 cells stain with TRPV3 antibodies (top, right). Coimmunostaining of rat DRG using a guinea pig anti-V1 (bottom, left) and a rabbit anti- V3 antibody (bottom, middle). V1 and V3 superimposed images (bottom, right) Scale bar = 10 µm. The top panel was provided by Chumbo Wang and the bottom panel by Hong- Zhen Hu.

46

AB 6 nA e [2APB] 10 30 100 300 1000 2APB (166 µM) (µM) d

) pcDNA3 8 4 -5 2 MP Ca2+ 4 0 0 +100 mV (x10

0 -100 mV F-F nm, 520 Fluo4, 2 8 2APB (166 µM) c MP -100 -50 ba 6 c 2+ 50 100 mV 4 Ca 16 a 1 nA 12 b c 15 sec d 2 8 e -2 TRPV3 4 d 0 0 e 0 (x10 C [2APB] 310 30 100 300 nA 8 2APB (166 µM) (µM) 8 e

MP -4 d 6 ) +100 mV 4 c 4 Ca2+ 16 -100 mV 12 -100 -50 ba 2 8 TRPV3 c 50 100 mV D641N 4 a b 4 nA Membrane potential dye, 565 nm, F-F nm, 565 dye, potential Membrane 0 0 c -4 02468 10 30 sec d Time (m) e -8 d e

Fig. 2.5. Activation of TRPV3 currents by 2APB. A, Simultaneous membrane 2+ potential and [Ca ]i measurements for 2APB-induced responses. Cells were loaded with fluo4 and then the membrane potential dye. Alternating fluorescence signals for the excitation/emission pairs of 494/520 and 530/565 nm were acquired once every 6 sec. The Ca2+ traces from 494/520 (red) are superimposed with the membrane potential traces from 530/560 (black). Separate experiments have shown that there is no interference between the two fluorescence channels (not shown). Traces represent triplicate experiments. B & C, 2APB-evoked currents in TRPV3 (B) and TRPV3D641N (C) cells. After making the whole cell, voltage ramps of 100 ms from –100 to +100 mV were applied from the holding potential of 0 mV at 2Hz. Solutions with different 2APB concentrations were applied through perfusion. Currents are shown for –100 and +100 mV (left). I-V curves are shown for the time points indicated by the letters (right). (B &C) were performed by Hong-Zhen Hu.

47

2+ 2+ [Ca ]o (mM) ~0.01 [Ca ]o A 0.03 nA ~0.01 0.1 15 0.03 0.5 0.1 2 10 0.5 10 2 30 mM 5 10 -100 -50 30 TRPV3 baseline 50 100 mV C -5 100 5 nA 5 nA -10 80 V3D641N 11.6 ± 2.3 mM 15 sec15 sec 60 2+ -15 [Ca ]o (mM) TRPV3 40 +100 mV ~0.01 1.00 ± 0.20 mM at +100 mat V nA +100 mat V -100 mV 0.5 20 15 2 2+ 10 R elative elative res res pons pons e e ~0.01 [Ca ]o 10 0 B 0.03 30 0.01 0.1 1 10 100 0.1 5 [Ca2+] (mM) 0.5 baseline o 2 -100 -50 10 100 30 mM 50 100 mV -5 80 +100 mV V3D641N -100 mV-10 60 7.5 ± 1.3 mM TRPV3D641N -15 40 TRPV3 at –100 mat V 20 0.23 ± 0.07 mM

response Relative 0 5 nA 0.01 0.1 1 10 100 2+ [Ca ]o (mM) 15 sec

2+ Fig. 2.6. TRPV3D641N has a reduced sensitivity to Ca . A&B, HEK293 cells expressing TRPV3 (A) or TRPV3D641N (B) were stimulated with 100 µM 2APB in a nominally Ca2+-free external solution. Increasing concentrations of Ca2+ were added after the activity reached the maximum. As shown to the right, extracellular Ca2+ also affected the shape of the I-V curves. C, dose-response curves to extracellular Ca2+ for currents at +100 mV (upper) and –100 mV (lower) for TRPV3 (open circles, n = 5) and

TRPV3D641N (filled circles, n = 4). Error bars are SD and EC50 values were determined from Hill equation. These experiments were performed by Hong-Zhen Hu.

48

°C 37 27 h

1.2 nA g

f 0.8 e 0.3 nA d +100 mV 1 min 0.4 c b a -100 mV -100 -50 50 100 mV a b c d e f g h -0.4

Fig. 2.7. Sensitization of TRPV3 currents by heat. heat-evoked currents from HEK293 cells transfected with V3. After making the whole cell, voltage ramps of 100 ms from –100 to +100 mV were applied from the holding potential of 0 mV at 2Hz. The thermal ramp protocol (left, above) was applied through perfusion. Currents are shown for –100 and +100 mV (left). I-V curves are shown for the time points indicated by the letters (right). This experiment was performed by Rui Xiao.

49

Fig. 2.8. Sensitization of TRPV3 currents by 2APB. 2APB-evoked currents from HEK293 cells transfected with V3. After making the whole cell patch, voltage ramps of 100 ms from –100 to +100 mV were applied from the holding potential of 0 mV at 2Hz. 100 µM 2APB was applied as indicated (left, above). Currents are shown for –100 and +100 mV (left). I-V curves are shown for the time points indicated by the letters (right). This experiment was performed by Rui Xiao.

50

A 2APB: 1 mM 1 mM 0.1 mM

1 µA

1 min

B 0.1 mM 2APB 39.5oC

i ii v

1 µA 1 min iv iii

Fig. 2.9. 2APB evoked TRPV3 currents in Xenopus oocytes. A, activation of TRPV3 expressed in a Xenopus oocyte by 2APB. The TRPV3-expressing oocyte was clamped at -40 mV and continuously perfused with a nominally Ca2+-free external solution at 22oC. 1 mM 2APB was applied twice as indicated. There was not significant inactivation but the return to the baseline was complete after drug washout. 0.1 mM 2APB-only evoked a very small current increase. B, augmentation of 2APB effect by heat. Application of 0.1 mM 2APB at 22oC (i) or immediately switching to an external solution heated at 39.5oC (ii) caused small current increases in the TRPV3-expressing oocyte. The same 2APB solution heated at 39.5oC caused more than 30-fold increase of the current (iii). Sequential heating followed by the heated 2APB solution also evoked a large current increase (iv). For all treatments, the washout was complete and the heat-evoked current was relatively stable (v). These experiments were performed by Hong-Zhen Hu.

51

A B µ 120 23 300 M 2APB 100

) 18

-4 V1 80 V2 13

(X 10 (X 10 60

o V3 F - 40 F 8 V4

Normalized response pcDNA 20 3 0 0 50 100 150 200 0.01 1 100 10000 Time (sec) [2APB] (µM)

Fig. 2.10. Activation of thermoTRPVs by 2APB. A, Shows representative traces of cells transfected with one of the thermoTRPVs or pcDNA and stimulated with 333 µM 2APB at 20 seconds. Traces are the average of triplicate experiments. B, Shows the least square fits of data points for TRPV1 (open circles), TRPV2 (filled circles), TRPV3 (closed triangles) and TRPV4 (open triangles). Data are averages ± SD of triplicate samples.

52

A B 2APB (166 µM) TRPV3 TRPV3D641N 20 20 2APB (166 µM) No Cap )

-4 ) 15 -4 15 No Cap

100 µM Cap 100 µM Cap (x1 0 TRPV3 0 10 (x10 10 0 TRPV3D641N F-F

F-F 5 Cap 300 µM Cap Cap 300 µM Cap 5

0 0 0123456 0123456 65 43 Time (m) Time (m) -log [Capsaicin] (M)

Fig. 2.11. Capsaicin inhibits the 2APB-induced activation of TRPV3. A, Fluo4- loaded cells that were transfected with TRPV3 (left) or TRPV3D641N (right) were sequentially exposed to capsaicin (20 seconds) followed by 166 µM 2APB (160 seconds). Representative traces are the average of duplicate samples at the concentrations shown. B, Inhibition dose-response curve of capsaicin for the 2APB-induced activation of TRPV3 (purple closed circles) or TRPV3D641N (open red circles). Data are the average of duplicate experiments for the time points 1 minute after 2APB application.

53

CHAPTER 3

TRPV3 IS MODULATED BY SIGNALLING EVENTS DOWNSTREAM OF G-PROTEIN COUPLED RECEPTOR ACTIVATION

3.1 Introduction

The polymodal nature of TRPV channels has been nicely illustrated for V1. The primary, or the best-known, function of V1 is temperature sensing of noxious heat. V1 is expressed in a subset of primary sensory neurons that have thin unmyelinated axons known as C-fibres, which terminate at the skin to sense heat and pain (Szallasi, 2001). A similar sensation is caused by capsaicin, the pungent ingredient of hot chili peppers, leading to the hypothesis that V1 is a transducer of noxious temperature sensation

(Caterina et al., 1997). Indeed, most of the properties of native CRs are recapitulated by

o heterologous expression of V1. With a Th of 43 C, V1 is inactive at normal body temperatures. Upon contact of the skin to temperatures >43 oC, the receptors located on the adjacent nerve fibres open to allow a rapid influx of Ca2+ and Na+, causing an action potential that moves from the nerve terminal to DRG, where the signal is then sent to the dorsal horn of the spinal cord, and is eventually perceived as noxious heat in the brain

(Woolf, 2000; Craig, 2003). However, at room and/or body temperature, V1 is activated by acid, 2APB, vanilloid compounds, as well as a growing number of natural products

54 (Caterina et al., 1997; Tominaga et al., 1998; Jordt et al., 2000; Tominaga and Tominaga et al., 2005; Calixto et al., 2005). Moreover, several endogenous ligands that are often associated with inflammatory states, including anandamide (AEA), N- arachidonoyldopamine (NADA), N-oleoyldopamine, and several lipoxygenase products of

AA, have been found to stimulate V1 at room temperature, indicating that V1 is involved in cellular processes outside of temperature sensation (Smart et al., 2000; Hwang el al.,

2000; Huang et al., 2002; Chu et al., 2003). In fact, studies using trpv1 knockout mice have demonstrated that V1 may play a secondary role to V3 with regard to noxious temperature sensation, but that V1 is an essential component of inflammatory thermal hyperalgesia (Caterina et al., Davis et al., 2000; Moqrich et al., 2005). Interestingly, capsaicin therapies have been successfully used to combat the chronic pain associated with arthritis or other long-term hyperalgesic conditions associated with inflammatory and/or neurogenic pain, an observation that has put V1 at the center of many drug development programs aimed at finding therapeutic pain remedies (Planells-Cases et al.,

2005). A summary of V1 regulatory components is shown in Figure 3.1.

Under normal conditions, nociceptive nerve terminals have a relatively high threshold for action potential firing. On the contrary, inflammatory and/or neurogenic pain is the result of oversensitive peripheral nerve terminals and/or receptors such as V1 that are located at these terminals. Nociceptive peripheral nerve terminals are sensitized by several mechanisms to include: 1) Autosensitization caused by the release of neuropeptides such as substance P and calcitonin-gene-related peptide immediately following nociceptive nerve firing (Woolf and Salter, 2000). In addition to sensitizing the terminals themselves, these neuropeptides recruit immunological cells and sensitize

55 adjacent epithelial cells (Koizumi et al., 1994). 2) Release of inflammatory mediators such as nerve growth factor (NGF), histamine, cytokines and prostaglandins by immunological cells (Woolf and Salter, 2000). 3) Release of NGF, ATP, and other proalgesic mediators from adjacent skin cells (Ruissen et al., 1995; Xu et al., 2006;

Asakawa et al., 2006). 4) release of acid, ATP and other cellular components upon tissue damage or inflammation to surrounding areas (Woolf and Salter, 2000; Chuang et al.,

2001). Regardless of the source, the inflammatory mediators released activate or potentiate the activity of V1 either indirectly by triggering cell signaling events through their respective receptors, or directly for protons and/or lipoxygenase or cyclooxygenase products of AA (Tominaga et al., 2001; Chuang et al., 2001; Hu et al., 2002; Shin et al.,

2002). Several PKA- and PKC-dependent phosphorylation sites (S116, S502, and S800 of rat V1) have been shown to be involved in the GPCR-induced modulation of V1 activity

(Tominiga et al., 2001; Bhave et al., 2002; Numazaki et al., 2002; Bhave et al., 2003).

Phosphorylation of S502 and S800 by PKC sensitizes V1 by decreasing its Th (Numazaki et al., 2002; Bhave et al., 2003). This could explain the sensitizing effect of BK and ATP, which has been shown to be PKC dependent (Chuang et al., 2001; Numazaki et al., 2002;

Carr et al., 2003). The effect of phosphorylation of S116 by PKA is different from the other two sites. S116 is constitutively phosphorylated in vivo, with dephosphorylation being necessary for V1 to enter a desensitized state. Thus, prostaglandins may minimize

V1 desensitization by maintaining the phosphorylation of S116 through PKA (Hu et al.,

2002; Bhave et al., 2002). In addition to phosphorylation, the breakdown of PIP2 is critical for the enhancement of V1 activity elicited by the stimulation of phospholipase C

(PLC) through either GPCRs or receptor tyrosine kinases (Chuang et al., 2001). It has

56 been shown that PIP2 binds directly to a C-terminal site (aa 777-792) of rat V1 and inhibits its activation (Prescott et al., 2003). Thus, distinct molecular mechanisms are involved in the regulation of V1 channel by different signals. By transducing these stimuli into depolarization of peripheral nerve terminals at normal body temperatures, V1 plays a critical role in integrating nociceptive signaling around damaged and inflamed tissues. It should also be noted that the ability of V1 to respond to so many stimuli underlies its’ role in asthma and nonproductive cough, as well as bladder dysfunction and migraine headaches (Birder et al., 2002; Szallasi, 2002; Hwang & Oh, 2002).

The identification of V3 as a thermosensitive channel that gates in response to increasing temperatures in the warm to noxious range as well as the high degree of sequence homology, shared chromosomal co-localization, identification of several putative PKA and PKC sites, and co-expression in the DRG of a subset of thinly- myleinated C-fibres, naturally led to the speculation that V3 would play a role in physiological processes similar to V1. In fact, several properties of V3 suggest it may be better suited to act as an integrator of multiple signals in an inflammatory environment.

First, V3 is the only thermosensitive TRPV family member to contain a C-terminal PDZ recognition sequence (Xu et.al., 2002). PDZ domain containing proteins assemble macromolecular complexes, thereby increasing signaling speed and specificity to the receptors and/or channels bound to them (Kim et al., 1995; Songyang et al., 1997;

Fanning & Anderson, 1999; McGee and Bredt, 1999, Nix et al., 2000; Huber et al.,

2001). In fact, the PDZ recognition domain on V3 is consensus for binding to PSD-95, a neuronal specific scaffolding protein shown to bind several channels/receptors (including

NMDA receptors) as well as the A kinase anchoring protein, AKAP79/150, which itself

57 provides a scaffold in which PKA, PKC, and the calcium-dependent phosphatase calcinurin can bind (Colledge et al., 2000; Fig. 3.18). Second, V3 is the only thermosensitive TRPV channel shown to “sensitize” upon repeated stimulation (Xu et al.,

2002; Section 2.3.7). This intrinsic property of V3 has been shown to occur for all known activators of V3, including heat, 2APB, diphenylboronic anhydride, and the recently identified skin sensitizers camphor, carvacrol, thymol, ethyl vanillin, and eugenol (Chung et al., 2005; Xu et al., 2006). Third, in addition to being expressed in peripheral C-fibres, V3 is expressed in keratinocytes as well as the epithelial cells of the nose and tongue (Peier et al., 2002; Moqrich et al., 2005; Xu et al., 2006). These cells are not only the first to encounter exogenous irritants and/or harmful stimuli, but they encounter the same inflammatory mediators that peripheral neurons encounter during inflammation. Despite these properties/characteristics, the lack of data demonstrating that V3 responds to low pH, capsaicin, 4α-PDD, or reduced osmolarity, has lead researchers to presume that the main function of V3 is to act as a warm sensor in the skin and/or DRG. In addition, studies using trpv1 or trpv3 knockout mice revealed that, although V3 played a larger role than V1 in noxious heat sensation, it was V1, not V3, that was essential for the development of thermal hyperalgesia in response to inflammatory conditions, a conclusion that may be misleading when considering that V3 expression is lacking in the sensory neurons of mice but not in the sensory neurons of rats, primates and humans (Xu et al., 2002; Smith et al., 2002; Peier et al., 2002; Moqrich et al., 2005; Zimmermann et al., 2005).

Several recent findings may change the way in which V3 is viewed. First, several plant derived compounds such as carvacrol, eugenol, and thymol have recently been

58 shown to activate V3. These compounds are not only known to act as skin sensitizers and allergic irritants, but are derived from spices such as oregano, savory, clove, and thyme, each of which invoke a warm sensation to the tongue (Xu et al., 2006).

Interestingly, the epithelium of murine tongue and nose were shown to express V3 in this study, consistent with previous reports in primates. Second, spontaneous mutant strains of mouse (DS-Nh) and rat (WBN/Kob-Ht) develop dermatitis and hairlessness in an autosomal dominant fashion. Breeding, linkage, and sequence analysis revealed that a single amino acid substitution in V3, V3G573S and V3G573C in mice and rats respectively, was responsible for these phenotypes. Consistent with this mutation being inherited in a dominant fashion, these mutations were shown to result in hypersensitive V3 channels.

In addition, the mutant strains showed increased numbers of mast cells in skin lesions, increased concentration of histamine and NGF in the skin, and an abnormal spacing between hair follicles when compared to age matched nonmutant strains. The conclusion of this study was that overreactivity of V3 results in release of higher than normal amounts of NGF from keratinocytes as well as the recruitment of mast cells to skin lesions, and ultimately development of dermatitis and hair loss. Finally, studies in this chapter will show that V3 activity is enhanced by the G-protein coupled receptors

(GPCR) histamine type-1 (H1R) and bradykinin type-2 (B2R), and that several downstream signaling events associated with activation of these receptors modulate V3.

These observations suggest that the role of V3 in various inflammatory conditions may be more significant than recently suggested.

59 3.2 Material and Methods

3.2.1 DNA constructs, cell culture, and transfections

All compounds and reagents were purchased from Fisher or Sigma unless otherwise stated. 2APB was from Tocris or Cayman Chemical Company. Bradykinin and histamine were from Cayman Chemical Company. All receptor cDNAs were cloned into pcDNA3.

3.2.2 For all experimental procedures, see section 2.2.

3.3 Results

3.3.1 TRPV3 is activated by GPCRs of the Gq/11 pathway

Inflammatory mediators, such as bradykinin (BK), histamine (His), and ATP, modulate V1 function through activation of their respective receptors (Gq/11 pathway)

(Premkumar & Ahern, 2000; Chuang et al., 2001; Shin et al., 2002). To determine if these mediators affect V3 function, we cotransfected the human type 2 bradykinin receptor (B2R) or the guinea pig type 1 histamine receptor (H1R) with V3, V3D641N or

2+ pcDNA3 in HEK293 cells and monitored changes in [Ca ]i by loading the cells with fluo4 (Fig. 3.2 & 3.3). Figure 3.2 shows the responses of transfected cells to three different concentrations of BK (0.03-3.0 µM) at 32 and 37 oC. In control cells, BK dose-

60 2+ dependently induced a rapid rise in [Ca ]i, which then declined to near the basal level.

Typically for HEK293 and many other cell types, the rapid initial rise is due to Ca2+- release from inositol 1,4,5-trisphosphate (IP3)-sensitive stores whereas the small, sustained phase at the later part of the stimulation is due to store-operated Ca2+ entry

(Zhu et al., 1998; Groschner et al., 1998; Parekh & Putney, 2005). In V3D641N and V3-

2+ cotransfected cells, the initial BK-induced rise in [Ca ]i always appeared smaller, most likely because of a lower expression of B2R due to the competition for transcription/translation machinery under conditions of coexpression. More

2+ significantly, the sustained phase of [Ca ]i elevation was increased in the V3D641N and

V3-cotransfected cells as compared to the pcDNA3 cotransfected cells. This effect was more pronounced at 37 oC (Fig. 3.2, lower) than at 32 oC (Fig. 3.2, upper) and was much stronger in V3D641N cells than in V3 cells. Similar results were obtained when the corresponding experiments were performed with H1R instead of B2R as the receptor and

His (0.30-30.0 µM) as the agonist instead of BK (Fig. 3.3). In addition, stimulation of endogenous purinergic P2Y and muscarinic acetylcholine receptors with ATP and carbachol, respectively, enhanced V3 activity (not shown). Based on these results, it was determined that 1-3 µM BK and 30 µM His should be the optimal concentration for receptor induced activation of V3/V3D641N. In addition, it was determined that the GPCR

o induced modulation of V3D641N and V3 should be studied at 32 and 37 C respectively.

61 2+ 3.3.2 RR and the removal of extracellular Ca reduce the Gq/11PCR induced rise in 2+ [Ca ]i through TRPV3

Consistent with activation of V3, removal of Ca2+ from the extracellular solution

2+ dramatically reduced the receptor induced rise in [Ca ]i for V3 and V3D641N following addition of 1 µM BK (Fig. 3.4) or 30 µM HIS (not shown) to cells cotransfected with the corresponding receptor. In addition, the simultaneous addition of 10 µM RR with 1 µM

BK in the presence of extracellular Ca2+ resulted in a significant reduction in the rise of

2+ o [Ca ]i for V3 at 37 C (Fig. 3.4B, top), but only a slight reduction in V3D641N (Fig. 3.4A, top) cells at 32 oC when compared to the equivalent experiments without RR. Cells only transfected with receptors were not affected by RR but showed a reduction in the rise of

2+ 2+ [Ca ]i when Ca was removed from the extracellular solution, a result that is expected

2+ for these controls (Fig. 3.4, bottom) The H1R induced rise in [Ca ]i for cells cotransfected with V3 and V3D641N yielded similar results as that seen in the B2R experiments (not shown). Since V3D641N has a reduced affinity toward RR compared to wild type V3, this observation confirms specificity of Gq/11-linked receptor pathways toward V3.

3.3.3 Activation of TRPV3 by Gq/11PCRs causes membrane depolarization

V3 is permeable to both Ca2+ and Na+ and should therefore depolarize the membrane upon activation. To further confirm that V3 is activated through Gq/11 pathways we cotransfected B2R or H1R with V3, V3D641N or pcDNA3 into HEK293 cells

62 and monitored changes in membrane potential using the FLIPR membrane potential

(FMP) dye (Fig. 3.5). In cells cotransfected with pcDNA3 and either B2R or H1R, application of 1 µM BK (Fig. 3.5A) or 30 µM His (Fig. 3.5B) to cells expressing the corresponding receptor did not cause a significant change in membrane potential at all

o temperatures tested (26, 32, 37 C). However, in cells cotransfected with V3D641N and

B2R or H1R, application of BK or His to the corresponding cells induced depolarization at all temperatures. For V3 cells, BK or His induced depolarization was not seen at 26 oC, but began to appear at 32 oC and was more pronounced at 37 oC. Interestingly, the membrane potential assay also revealed a temperature dependent rise of basal fluorescence signal in V3-exressing cells, especially between 26 and 32 oC, indicative of heat-induced depolarization, which is consistent with V3 being activated by warm temperatures. Overall, these results confirm that V3 activity is enhanced by the activation of the Gq/11 signaling pathway.

3.3.4 Gq/11PCRs potentiates the 2APB response of TRPV3

As a polymodal receptor/channel, V1 is able to transduce multiple signals into

2+ membrane depolarization and a rise in [Ca ]i, ultimately resulting in action potential firing and activation of Ca2+-specific signaling events in C-fibres. In addition, subthreshold combinations of V1 specific signals result in channel activity that either would not occur in the presence of one stimulus alone, or enhances V1 activity to a much greater extent than the additive effects of each individual stimulus. This property of V1 makes it an ideal receptor/channel for the transduction of painful stimuli during

63 inflammation, when additional motivation to avoid agitation of affected areas is necessary in order to prevent further injury/infection and promote healing. To determine if activation of G(q/11)PCRs can enhance the activity of other stimuli known to activate

V3, thus demonstrating that V3 may be regulated in a similar manner as V1 under inflammatory conditions, we cotransfected HEK293 cells with B2R and either V3,

V3D641N or pcDNA3 and applied differing concentrations of 2APB three minutes prior to the addition of 1 µM BK (Fig. 3.6). The experiments were performed at 26 oC to minimize the direct activation of V3 by the B2R receptor. As is clearly demonstrated,

B2R stimulation potentiates the response of V3 to 2APB at concentrations between 2.0 and 55 µM (Fig 3.6A). Interestingly, B2R activation does not seem to increase the sensitivity of V3 to 2APB to a significant extent, but instead causes a larger increase in

2+ [Ca ]i at a particular 2APB concentration. In fact, at maximal 2APB (166 µM) induced

2+ activation of V3, stimulation of B2R does not cause a further rise in [Ca ]i. Similar results were observed when V3D641N was used instead of V3, although only traces for 0.0,

2.0, 6.0 µM 2APB are shown (Fig. 3.6B, left). In addition, 2APB did not affect the response of control cells cotransfected with B2R and pcDNA3 at any of the concentrations tested (Fig. 3.6B, right). Similar results were obtained when performing the corresponding experiments with H1R receptor and 30 µM HIS instead of B2R and 1

µM BK (not shown). These data demonstrate that the 2APB induced activation of V3 is potentiated by G(q/11)PRC pathways, suggesting that V3 should display enhanced activity under conditions of inflammation; an observation that has been nicely demonstrated for

V1 and is critical for the development of thermal hyperalgesia and asthma, as well as other disorders related to V1 hypersensitivity.

64 3.3.5 Signaling events downstream of G(q/11)PCRs modulate TRPV3

As shown in Figure 3.1, V1 is a polymodal receptor, as it is regulated by a multitude of cellular factors. This characteristic of V1 is essential for a receptor that plays a role in so many inflammatory pathologies. The high degree of sequence homology, shared chromosomal co-localization, co-expression in the DRG of a subset of thinly-myelinated C-fibres, as well as similarities with regard to thermal gating, led us to investigate the possibility that V3 is also regulated by a multitude of cellular factors.

3.3.5.1 The effect of PKA and PKC phosphorylation on the G(q/11)PCR induced activation of TRPV3

G(q/11)PCRs such as B2R and H1R modulate channel activity indirectly through several signaling pathways including PKA, PKC and PLC activation. In addition, events downstream of PLC activation have been shown to modulate V1 (Hwang & Oh, 2002).

We therefore preincubated blockers of various signaling pathways 30 minutes prior to the addition of 1 µM BK to HEK293 cells cotransfected with B2R and either V3, V3D641N or pcDNA3 in order to determine which signaling events downstream of G(q/11)PCR were responsible for the enhancement of V3 activity (Fig. 3.7). Experiments were originally

o o performed using V3D641N at 32 C and then confirmed using V3 at 37 C. Because results were similar for V3 and V3D641N, only relevant traces are shown (Fig. 3.7A, C), and only the summary for V3D641N is shown (Fig. 3.7B). Unless otherwise stated, the receptor activity was not affected by drugs in cells cotransfected with B2R and pcDNA3. The

65 PKC blockers GF109203X (GF, 0.1 µM, data not shown) and CalphostinC (1.5 µM) did not effect B2R induced activation of V3 or V3D641N. The PKA blocker H89 (10 µM) consistently caused a slight increase in the receptor-induced activation of V3 and

V3D641N. In addition, 30 µM GF showed a similar degree of potentiation as 10 µM H89.

Generally, 0.1 µM GF will fully block PKC, with higher concentrations having efficacy for other kinases, including PKA. Neither the irreversible adenylyl cyclase inhibitor,

MDL-123330A (50 µM) nor the adenylyl cyclase activator, Forskolin (3 µM) had a significant effect on the B2R induced activation of V3 or V3D641N, suggesting that the effect of the PKA inhibitor is not affected by changes in cAMP concentration.

3.3.5.2 The effect of PLC on the G(q/11)PCR-induced activation of TRPV3

The phospholipase D (PLD) inhibitor D609 (3 µM) only caused a slight inhibition of the B2R induced activation of V3 and V3D641N. In contrast, the phospholipase C inhibitor U73122 (15 µM) blocked the B2R induced activation of V3 and V3D641N strongly and consistently (Fig. 3.7 A & B). In addition, the BK induced rapid initial rise

2+ in [Ca ]i was almost completely blocked by U73122 in cells cotransfected with B2R and pcDNA3 (not shown), indicating that PLC activation is necessary for receptor induced activation of V3. It should be noted that the inactive analog of U73122, U73433, potentiated the receptor-induced activation of V3 and V3D641N, a phenomenon that will be discussed in Section 6.3.1. PLC generates diacylglycerol (DAG) and IP3 by cleaving phosphatidylinositol 4,5-bisphosphate (PIP2), a negatively charged phospholipid found on the inner leaflet of plasma membranes. DAG lipase can then cleave DAG, generating

66 free AA, a phenomenon that can result in either the accumulation of AA in the plasma membrane and/or the production of several AA metabolites. Of relevance, several TRP channels (Drosophila TRP and TRPL, C6, V1, V4, A1) are directly activated by AA and/or products of AA metabolism (Chyb et al., 1999; Zygmunt et al., 1999; Basora et al., 2003; Watanabe et al., 2003; Bandell et al., 2004; Park and Vasko, 2005; Hardie,

2006). ETYA (20 µM), a non metabolizable triple-bonded analogue of AA that prevents the production of all derivatives of AA via inhibition of cycloxygenase, lipoxygenase, and epoxygenase activity, consistently reduced the B2R induced activation of V3 and

V3D641N, suggesting that AA and/or one of the products of AA metabolism modulate V3 activity (Fig. 3.7, see Fig. 3.8 for overview of AA metabolism). In addition, the cycloxygenase inhibitor piroxicam (60 µM) consistently showed a slight potentiation of the response, indicating that either AA itself, or a metabolic product(s) of AA derived from lipoxygenase or cytochrome P450 epoxygenase activity are responsible for the modulation of V3. Alternatively, piroxicam itself could act as a weak chemical activator of V3. In summary, PLC, PKA, and AA and/or one or more of the oxidized AA metabolites downstream of either lipoxygenase or cytochrome P450 epoxygenase activity appear to be responsible for the modulation of V3 downstream of G(q/11)PCR activation.

3.3.5.3 The effect of polyunsaturated fatty acids on the G(q/11)PCR-induced activation of TRPV3

Since AA modulates several TRP channels, we applied differing concentrations of

o AA to V3D641N transfected cells and monitored fluo4 fluorescence at 32 C to determine if

67 2+ AA modulates V3 (Fig. 3.9). As shown, 33.3 - 300 µM AA caused an increase in [Ca ]i that was not seen in control cells (Fig. 3.9A). To ensure that activation by AA was not unique to V3D641N, we applied 300 µM AA to cells expressing either V3 or V3D641N and

o compared the level of activation at 32 C (Fig. 3.9B). As expected, V3D641N cells were activated to a greater extent than V3 cells, but V3 was activated at this concentration of

AA. Interestingly, 20 or 100 µM ETYA applied two minutes prior to 100 µM AA, not only failed to block the activation of V3 by AA, but 100 µM ETYA appeared to potentiate the response (not shown). Since ETYA is both nonmetabolizable and able to inhibit each of the enzymes that metabolize AA, we applied differing concentrations of

ETYA (33.3-300 µM) to cells expressing V3D641N to determine if ETYA, like AA, could activate V3 directly. Figure 3.9C shows that 300 µM ETYA also directly activated

V3D641N. These observations, along with a decrease in the B2R-induced activation of V3 by 20 µM ETYA, suggest that V3 is directly activated by AA and ETYA, but does not rule out the possibility that one or more derivatives of AA metabolism modulate V3 with higher efficacy, especially in the sensitized state. In fact, if high potency derivatives of

AA for V3 exist, it would explain the inhibition of the B2R-induced activation of V3 by

ETYA.

3.3.5.4 The effect of PIP2 depletion on the G(q/11)PCR-induced activation of TRPV3

Production of DAG, followed by the generation of free AA, is not the only manner in which cleavage of PIP2 by PLC has been shown to modulate TRP channels, as decreasing the concentration of PIP2 in the membrane can have either stimulatory

68 (TRP/TRPL, V1) or inhibitory (M4, M5, M7, M8, V5) effects on TRP channels (Estacion et al., 2001; Runnels et al., 2002; Prescott & Julius, 2003; Liu & Liman, 2003; Rohacs et al., 2005; Zhang et al., 2005; Liu & Qin, 2005). Phenylarsine oxide (PAO) inhibits phosphatidylinositol-4-kinase (PI4 kinase), resulting in a net depletion of PIP2 from the plasma membrane (Prescott and Julius, 2003). The rate at which PIP2 is depleted from the membrane depends on various conditions in the cell that regulate the natural turnover of PIP2 (see Fig. 3.10 for schematic of the PIP2 cycle). We therefore applied differing concentrations of PAO to HEK293 cells expressing V3D641N and monitored fluo4 florescence in order to determine if depletion of PIP2 modulates V3. Concentrations of

PAO between 1.1 µM and 90 µM dose-dependently activated V3D641N (Fig. 3.11A, left).

30 µM PAO was determined to be the optimal concentration since at 90 µM a slight endogenous response was observed (Fig. 3.11A, right). We next compared the rise in

2+ o [Ca ]i of V3 or V3D641N transfected cells using 30 µM PAO at 32 C (Fig. 3.11B). V3

2+ transfected cells exhibited a delayed activation with a slow rise in [Ca ]i occurring after

2+ 5 minutes. For V3D641N transfected cells, the [Ca ]i began rising immediately upon addition of PAO, with the activity of V3D641N still increasing after 8 minutes of stimulation. The slow activation rate is most likely due to a time-dependent depletion of

PIP2 in the membrane, suggesting that PIP2 may hold V3 in an inactive state in a manner

o o similar to V1. At 37 C, the rate of V3 activation was similar to that of V3D641N at 32 C, indicating that V3D641N is more sensitive than wild type V3 (Fig. 3.11B). Removal of

2+ 2+ Ca from the extracellular solution completely prevented the PAO-induced rise in [Ca ]i for both V3 (not shown) and V3D641N (Fig. 3.12). In addition, 10 µM RR almost

2+ completely prevented the PAO induced rise of [Ca ]i for V3 (not shown) transfected

69 cells, but only caused a slight reduction in V3D641N transfected cells, confirming specificity for V3 (Fig. 3.12A). A summary of the inhibition of the PAO response by 10

µM RR for V3 and V3D641N is shown in Figure 3.12B. Also consistent with V3 activation, 30 µM PAO caused membrane depolarization for both V3 and V3D641N transfected cells at 37 and 32 oC respectively, with a stronger response being observed in

V3D641N cells (Fig. 3.13). No change in membrane potential was observed in pcDNA3 transfected cells. These data suggest that PIP2 holds V3 in an inactive state, with depletion of PIP2 from the plasma membrane resulting in enhancement of V3 activity.

The observation that 2APB and heat work synergistically to activate V3 (Fig. 2.9) reveals a common characteristic between V3 and V1, as acid, capsaicin, heat, and 2APB have a synergistic effect on V1 activation when more than one stimuli is added simultaneously (Kress and Zeilhofer, 1999; Hu et al., 2004). Therefore, we reasoned that

PAO and 2APB may work synergistically to activate V3. To test this, 30 µM PAO was added to V3 expressing HEK293 cells four minutes prior to the addition of differing

2+ concentrations of 2APB and the [Ca ]i was monitored via fluo4 fluorescence (Fig.

3.14A). These experiments were performed at 32 oC to avoid the direct activation of V3 by PAO. As shown, PAO caused a left-shift in the dose response curve of V3 to 2APB at this temperature, as an EC50 value of approximately 6 µM was obtained for 2APB after exposure to 30 µM PAO for 4 minutes compared to 57.7 µM without PAO treatment (Fig

3.14B; compare to Fig. 2.1). Concentrations as low as 490 nM 2APB caused a rise in

2+ [Ca ]i. These data show that PAO and 2APB have a synergistic effect toward V3

70 activation, exemplifying a common characteristic between V1 and V3; both have an intrinsic ability to coordinate multiple subthreshold signals into channel activation that would not occur with a single stimulus alone.

Activation of G(q/11)PCRs such as B2R and H1R activate PLC, resulting in a rapid depletion of PIP2 from the plasma membrane. Since PAO slowly depletes PIP2 from the plasma membrane by inhibiting PI4 kinase, activation of B2R or H1R in the presence of

PAO should result in a faster depletion of PIP2 from the membrane, a phenomenon that should potentiate the receptor induced activation of V3 (Fig. 3.10). To test this, 30 µM

PAO was applied four minutes prior to the addition of 3 µM BK to HEK293 cells

o coexpressing B2R and V3 at 32 C (Fig. 3.15A) and the cells were monitored for changes

2+ in fluo4 fluorescence. Interestingly, 30 µM PAO caused an instantaneous rise in [Ca ]i that is not seen in cells transfected with V3 alone at this temperature (compare to Fig.

3.11B). Overexpression of B2R should result in a higher turnover rate of PIP2 when compared to cells not expressing B2R due to the fact that G(q/11)PCRs have basal activity associated with them, consistent with V3 being release from an inactive state by PIP2 hydrolysis. Furthermore, addition of 3 µM BK resulted in a much higher receptor induced activation compared to cells that were not treated with PAO, suggesting that

PAO potentiates the receptor-induced response of V3. In corresponding studies using

H1R and 300 µM HIS instead of B2R and 3 µM BK, similar observations were made (Fig

3.15B). To what extent the G(q/11)PCR-induced enhancement of V3 activity is the result of phosphorylation, release of AA, or PIP2 hydrolysis is not known, but PAO obviously works synergistically with events downstream of G(q/11)PCR activation to enhance V3 activity.

71 3.4 Discussion

To summarize, V3 is directly activated by G(q/11)PCRs such as H1R and B2R at 37 oC, and this activation results in both Ca2+-influx and membrane depolarization.

Although G(q/11)PCRs do not activate V3 to a significant extent below body temperature, they do potentiate the 2APB response at these temperatures. Pharmacological analysis reveals that events downstream of PLC activity are responsible for V3 activation by

G(q/11)PCR, with release of AA, PKA activity, and depletion of PIP2 from the plasma membrane all contributing to modulation of V3. Although the relative contribution that each of these signaling events contributes to V3 activation will depend on conditions within and around the cells in which V3 is expressed, the polymodal nature of V3 should integrate even small contributions from several of these signals into a strong activation of

V3. As has been seen with 2APB, V3D641N is more sensitive to G(q/11)PCR activation, as using V3D641N instead of V3 yields similar results at lower temperatures and/or lower agonist concentration, lending further support to the hypothesis that Asp641 is a critical site of Ca2+-induced inhibition (Xiao et al., 2006, Section 2.4).

Data in this chapter demonstrate that AA or ETYA, a nonmetabolizable triple bonded analogue of AA, are capable of activating V3 directly. In fact, other studies in our lab have recently shown that V3 is not only directly activated by several unsaturated fatty acids, but that these fatty acids potentiate the 2APB response, with the most potent unsaturated fatty acids containing double bonds that start at the fifth position from the carboxylated head group (AA is one such fatty acid) (Hu et al., 2006). In addition these studies demonstrated that the triple bonded fatty acids, 5,8,11-eicosatriynoic acid (ETI)

72 and ETYA, also directly activated V3 currents and potentiated the 2APB-induced currents in V3 transfected cells. These observations are relevant to human health when considering recent findings that have implicated the involvement of V3 in inflammatory processes such as dermatitis and irritation of skin, mouth and nose (Xu et al., 2006;

Asakawa et al., 2006). Under inflamed conditions, AA is either released from skin cells or sensory fibres following GPCR activity, or it is released from infiltrating lymphocytes.

In fact, AA concentrations of up to 100 µM have been reported in the epidermis of involved psoriasis, with 13 µM AA being common in uninvolved psoriasis

(Hammarstrom et al., 1975; Brash, 2001). With high local levels of AA existing under inflammatory states, the enhancement of V3 activity by AA may play a role in the skin related disorders discussed above as well as in thermal hyperalgesia, chronic pain and other unidentified physiological disorders that V3 may play a role in.

Interestingly, at concentrations capable of inhibiting lipoxygenase, epoxygenase, and cyclooxygenase activity, ETYA causes a reduction in the G(q/11)PCR induced activation of V3 (Fig. 3.7). This observation suggests that one or more oxidized AA metabolites may be at least partially responsible for V3 activation in the context of

G(q/11)PCR stimulation. In fact, it is not expected that AA would reach the concentrations (> 30 µM) necessary to directly activate V3 by receptor activation alone, although the combined affects of low concentrations of AA as well as PIP2 hydrolysis and PKA and/or PKC phosphorylation may ultimately activate V3 under the conditions of G(q/11)PCR activity. Interestingly, phosphorylation of V3 by PKC has been shown to enhance the AA induced activation of V3, demonstrating that phosphorylation events may serve to increase the sensitivity of V3 in a similar manner as V1 (Hu et al., 2006).

73 Regardless, our lab has demonstrated that cinnamyl-3,4-dihydroxy-cyanocinnamate

(CDC) and nordihydroguaiaretic acid (NDGA) also directly activate V3 and potentiate the activity of the 2APB response (Hu et al., 2006). Interestingly, piroxicam structurally resembles these molecules, possibly accounting for the slight potentiation of the response to GPCRs by piroxicam that was observed in Figure 3.7 (chapter six provides a detailed analysis of structural considerations of molecules shown to activate V3). Although these compounds are not fatty acids, CDC and NDGA are reversible inhibitors of lipoxygenase.

Reversible inhibitors of enzymes often work because they structurally resemble either the reactant or product of the enzyme in which they inhibit (Fersht, 1985). Taking this into consideration, it is possible that one or more of the lipoxygenase products of AA are partly responsible for the G(q/11)PCR induced activation of V3. Although low concentrations of lipoxygenase products should be expected under the conditions of the experiments performed in this chapter, the combined effects of lipoxygenase products with the other signaling events downstream of G(q/11)PCR activity could activate V3 considering the polymodal nature of V3. This type of mechanism is not unfounded, as lipoxygenase products have been shown to activate V1 and play a role in inflammatory thermal hyperalgesia (Huang et al., 2002). It is worth emphasizing that the activation of

V3 by ETYA, as well as the fact that AA activates V3 in the presence of inhibitors of AA metabolism (Hu et al., 2006; Sections 3.3.5.2 & 3.3.5.3), demonstrates that AA itself is capable of activating V3, but at the same time does not rule out the possibility that, under the appropriate conditions, derivatives of AA may contribute to the activation of V3.

The PIP2 concentration in the plasma membrane has been shown to influence the activity of many receptors and channels. By adjusting the rate of PIP2 synthesis and/or

74 hydrolysis in response to signaling events, cells can mediate the responsiveness of key channels and receptors in accordance with metabolic or physiological need. The PI4 kinase inhibitor PAO, was shown to slowly enhance the activity of V3, presumably via the slow depletion of PIP2 from the plasma membrane. Further support of this was shown not only by the potentiation of the G(q/11)PCR induced activation of V3 by PAO, but by the enhanced rate of V3 activation by PAO when V3 was coexpressed with B2R, which should result in a faster depletion of PIP2 in the plasma membrane due to the constitutive activity of B2R (Fig. 3.15). Linking PIP2 hydrolysis to enhanced V3 activity is a significant finding when considering that PIP2 hydrolysis has been shown to release

V1 from an inactive state following G(q/11)PCR activity, contributing to enhancement of

V1 activity during conditions of inflammation (Prescott and Julius., 2003). PIP2 has been proposed to inhibit V1 by binding to a putative PIP2 binding domain (aa 777-820 of murine V1) in the C-terminus of V1. Interestingly, sequence analysis revealed that V3 is lacking a PIP2-binding domain, suggesting that inhibition of V3 by PIP2 is mediated by a mechanism different than that used by V1, although it should be noted that PIP2 binding domains are loosely defined, with the only true structural determinants being stretches of positively charged amino acids intermixed with hydrophobic amino acids. It is quite possible, that V3 has a structural determinant for PIP2 binding that is less conventional than the PIP2 binding domain of V1. Alternatively, many channels and receptors preferentially localize to lipid rafts and PIP2 has been implicated in their formation.

Lipid rafts are small lipid domains known to act as scaffolds for various signaling transduction components (Helms & Zurzolo, 2004). These rafts differ from the bulk membrane in both lipid composition and biophysical properties (van Meer, 2002;

75 Gallegos et al., 2006). Interestingly, lipid rafts have been shown to be concentrated in

PIP2 and cholesterol, two lipid molecules that have been shown to have inhibitory effects on V1 at higher concentrations (Hope and Pike, 1996; Liu et al., 2003; Hur et al., 2004).

It is possible that V3 preferentially localizes to lipid rafts, and that this localization is dictated by structural and/or biophysical means unrelated to PIP2 binding domains.

Association of V3 with lipid rafts in such a scenario may increase the concentration of signaling proteins such as PLC in the vicinity of V3, contributing to enhancement of V3 activity following G(q/11)PCR activation.

As mentioned, V3 is the only thermoTRPV family member that contains a PDZ- recognition sequence, suggesting that V3 is capable of forming signaling complexes in the cells in which it is expressed. In fact, the proper functioning of Drosophila TRP

(TRP) in photoreceptor cells is dependent upon the formation of a signaling complex termed “Drosophila signaxplex” (DS) (Montell, 2005b). A PDZ recognition sequence located on the C-terminus of TRP serves as a link between TRP and INAD. INAD is a

PDZ domain containing scaffolding protein that also binds the photon transducing receptor rhodopsin, PLC, PKC, calmodulin, and the actin binding protein NINAC myosin

III, thus forming the DS (Montell, 2005b). In addition, by forming heterotetramers with

TRPL, the PDZ recognition sequence of TRP effectively brings TRPL proteins into the

DS. Interestingly, TRP/TRPL is regulated by calmodulin, PKC, PLC, PIP2 depletion as well as AA and other PUFAs, and signaling speed and specificity is imparted to

TRP/TRPL by localizing these otherwise promiscuous signaling molecules to the complex. The observation that V3 is regulated by the same signaling molecules as

TRP/TRPL, coupled to the fact that V3 is the only thermoTRPV family member to

76 contain a PDZ recognition sequence, merits further investigation in order to determine whether or not a similar signaling complex exists for V3 in native tissues. Since PDZ containing proteins are best known for their ability to form signaling scaffolds at cellular synapses, the membrane-membrane appositions between keratinocytes and sensory nerve terminals would be a likely candidate for the formation of signaling complexes in keratinocytes and/or sensory neurons (Peier et al., 2002). Also, since it has been suggested that V1 and V3 form heterotramers, the possibility of V3 bringing V1 into these signaling complexes in a similar manner as TRP and TRPL is worth exploring. It is worth noting that V1 has been shown to be part of a PKA/AKAP/V1 signaling complex that is essential for V1 regulation by prostaglandins during inflammatory states (Rathee et al., 2002). Figure 3.16 demonstrates the manner in which PDZ domain containing proteins such as PSD-95 form signaling complexes.

The observation that G(q/11)PCR activation enhances V3 activity implies that V3 plays a role in inflammatory states. Whether V3 plays a more significant role in function of keratinocytes or sensory neurons is a matter of debate, but the fact that BK, HIS, substance P as well as other neurotransmitters activate their respective receptors on keratinocytes, sensory neurons and lymphocytes, highlights the possibility that these tissues display a high degree of cross-talk between each other. For instance, it has long been known that activation of sensory neurons results in the release of neurotransmitters such as substance P, resulting in lymphocyte recruitment, autocrinal sensitization of

2+ peripheral nerves, as well as PIP2 depletion and increases in the [Ca ]i of keratinocytes

2+ (Koizumi et al., 1994; Woolf and Salter, 2000). In addition, increases in the [Ca ]i of keratinocytes results in the release of neurotransmitters such as NGF and IL-1α, which

77 sensitize sensory nerve terminals and recruit lymphocytes (Xu et al., 2006; Asakawa et al., 2006). Lymphocytes in turn release AA, histamine, NGF as well as numerous other sensitizing molecules. This type of scenario creates a “perfect storm” in which one event causes signal amplification via a continuous feed back loop between keratinocytes, lymphocytes and sensory nerve terminals. Although the way in which the V3 specific signaling events identified in this chapter affect V3 function in vivo is not yet known, it can no longer be assumed that V3 functions as just a warm sensor.

78

Heat Vanilloids Neurotransmitters H+ Neurotransmitters Inflammatory factors Inflammatory factors 5 mV Na+, Ca2+ 20 ms Peptides Peptides NGF ATP

VGIC GPCR GPCRGPCR TRPV1 GPCR AC P P P P PLC PLA2 PLD ATP P Gq/11 Gq/i Gq/i Gsi

PIP2 cAMP Anandamide LM prostaglandins

IP3 AA DAG PKC PKA Ca2+ ER

Fig. 3.1. Regulation of TRPV1 by various signaling pathways. Several receptors modulate V1 activity through activation of signaling pathways downstream of receptor activation. Both GPCRs and tyrosine kinase receptors (NGF-TrkA) have been shown to modulate V1 activity. Examples of receptors and the pathways they utilize to modulate V1 in nociceptive C-fibers are shown as follows: 1) Activation of B2R by bradykinin sensitizes V1 via PKC phosphorylation. In addition, B2R activates V1 by producing leukotrienes via activation of PLA2 and subsequent lipoxygenase activation (Haung et al., 2002). The leukotrienes bind to the capsaicin binding site. 2) Cysteinyl leukotriene receptors sensitize V1 via PKC-dependent phosphorylation (Haung et al., 2002). 3) Activation of P2 purinergic receptors by extracellular ATP sensitizes V1 via PKC- dependent phosphorylation (Tominaga et al., 2001). 4) Prostaglandins activate prostaglandin E2 receptors leading to increases in cAMP and PKA activity. PKA-mediated phosphorylation prevents V1 desensitization (Hu et al., 2002). 5) Glutamate activates metabotropic glutamate receptors and PLC, which produces diacylglycerol (DAG). DAG is then converted to arachodinic acid (AA) via DAG lipase. AA is then converted to prostaglandin E2 (PGE2) by cycloxygenase. PGE2 activates adenylyl cyclase, ultimately leading to PKA activation, which prevents V1 desensitization (Hu et al., 2002). Note: AC = adynylyl cyclase; PLC = phospholipase C; PLA2 = phospholipase A2; PLD = phospholipase D; LO = lipoxygenase; PKC = protein kinase C; PKA = protein kinase A; cAMP = cyclic AMP; ER = endoplasmic reticulum PKC is activated by PLC dependent production of diacylglycerol (DAG) and elevation of cytosolic Ca2+.

79

2+ Fig. 3.2. Activation of TRPV3 by B2R. BK-evoked changes in [Ca ]i in cells co- expressing B2R and pcDNA3, V3, or V3D641N. Fluo4-loaded cells were stimulated with different concentrations of BK as indicated at 32oC (upper) or 37oC (lower). Traces represent the average of duplicate experiments. RFU = relative fluorescence units

80

2+ Fig. 3.3. Activation of TRPV3 by H1R. HIS-evoked changes in [Ca ]i in cells co- expressing H1R and pcDNA3, V3, or V3D641N. Fluo4-loaded cells were stimulated with different concentrations of HIS as indicated at 32oC (upper) or 37oC (lower). Traces represent the average of duplicate experiments. RFU = relative fluorescence units

81

2+ Fig. 3.4. Effects of RR and removal of extracellular Ca on the B2R-induced 2+ activation of TRPV3. A, BK-evoked changes in [Ca ]i in cells co-expressing B2R o 2+ and V3D641N (top) or pcDNA3 (bottom) at 32 C. B, BK-evoked changes in [Ca ]i in o cells co-expressing B2R and V3 (top) or pcDNA3 (bottom) at 37 C. A & B, blue traces are in the presence of 1.8 mM extracellular Ca2+, red traces have 10 µM RR in the presence of 1.8 mM extracellular Ca2+ and green traces are in a nominally 1.8 mM extracellular Ca2+-free solution. Each trace represents the average of duplicate experiments. All cells were loaded with fluo4.

82

A

) 1 µM BK 1 µM BK -5 10 1 µM BK B R/V 3 26oC 32oC 37oC 2 B2R/V 3DN 6 B R/pc DNA 2 RFU (X10 2 012 012 0123 Time (m) Time (m) Time (m)

B 30 µM His 30 µM His 30 µM His ) 10 -5 o 37 C H1R/V 3 32oC H R/V 3DN 26oC 1 6 H R/pc DNA 1

RFU (X10 (X10 2 0 12 01 2 0123 Time (m) Time (m) Time (m)

Fig. 3.5. GPCR-induced activation of TRPV3 leads to membrane depolarization. A, BK-induced depolarization in B2R and V3 co-transfected cells. B, HIS-induced depolarization in H1R and V3 co-transfected cells. A & B, cells were loaded with FMP dye and monitored in FlexStation at 26oC (left), 32oC (middle) or 37oC (right). The increase in fluorescence indicates depolarization. Each trace represents the average of duplicate experiments. RFU = relative fluorescence units.

83

1 µM BK

14 2APB

12 500 µM )

-4 167 µM 10 55.5 µM 8.0 18.5 µM 10 (X 6.17 µM

o 6.0 2.05 µM 4.0 0.69 µM

F - F 0.0 µM 2.0 0 012345 Time (m) 1 µM BK 1 µM BK 2APB 2APB 8 ) ) -4 -4 6 [2APB] (µM) [2APB] (µM) 6 6

(X 10 (X 2 2

(x 1 0 4 o 0 0 0

F-F 2 F - F

0 0123456 0123456 TimeTime (m) (m) Time (m) (m)

Fig. 3.6. Potentiation of the 2APB-induced activation of TRPV3 by B2R. A, dose dependent activation of cells cotransfected with B2R and V3. Fluo4-loaded cells were stimulated with different concentrations of 2APB at 20 sec followed by 1ìM BK at 140 sec. B, dose dependent activation of B2R and V3D641N cotransfected cells. Fluo4-loaded cells were stimulated with different concentrations of 2APB at 20 sec followed by 1ìM BK at 200 sec. A & B, traces represent the averages of duplicate experiments.

84

o 32 C 1 µM BK 3 1 µM BK 3

) 15 µM U73343

-4 2 2

(x 10 (x Buffer only A 0 1 1 F-F 1.5 µM CALC 20 µM ETYA 10 µM H89 Buffer only 15 µM U73122 0 0 0 1234560123456 Time (m) Time (m)

BK-evoked response of TRPV3 2 D641N 1.5 B 1 0.5 0 e 2 3 A 9 X C 9 A in n 2 4 8 3 0 0 l m o 1 3 Y in 6 3 o a 3 3 T H 0 t k ic N 7 7 E 2 s D ,3 s x Normalized response M 9 o 2 r o U U µ 0 h M 1 o ir M 1 p µ - F M M µ 0 F l L p µ µ 0 1 a 3 D M 2 G C M 5 5 M µ µ 1 1 M M 3 0 µ µ M 6 0 µ 3 .5 0 1 5 3737 oCoC 3 µM BK 3 µM BK 12 20 µM H89 10 pcDNA3 )

-4 8 Buffer only

6 (x 10 (x 20 µM H89 C 0 4 20 µM ETYA 20 µM ETYA F-F Buffer only 2 TRPV3 0 0123401234 Time (m) Time (m)

Fig. 3.7. Effects of drugs on the BK-induced activation of TRPV3. A, shows example traces for H89 (PKA blocker), Calphostin C (CALC, PKC blocker), U73122 (PLC blocker), U73433 (inactive form of U73122) and ETYA (non-metabolizable AA analog). B, shows the summary for all drugs tested. Data are averages ± SE for 4-5 samples or averages ± range for duplicated samples for fluorescence increases normalized to the control cells at 100 sec after the addition of BK. C, shows example traces for H89 and ETYA on BK-induced activation of V3 (left) and control (right) cells coexpressing B2R. A-C, Drugs were added >30 min before BK.

85

G-Protein Coupled Receptors

Membrane phospholipids piroxicam ETYA phospholipases Arachidonic acid

ET YA CyclooxygenaseLipoxygenase (LO) Cytochrome P450 Epoxygenase

5-LO 12-LO 15-LO Prostaglandins Epoxides 12-hydroxyeicosatetraenoic acid Prostacyclins

Thromboxanes EETs

HPETEs

Fig. 3.8. Overview of arachidonic acid metabolism. Cycloxygenase converts arachidonic acid (AA) into short-lived prostagladin intermediates that are then converted into several prostacyclins of thromboxanes. Several lipoxygenase (LO) enzymes convert AA into a variety of hydroperoxy-eicosatetraenoic acid (HPETE) derivatives. Being unstable, the resulting HPETE derivatives are further modified resulting in a multitude of metabolites that either act directly on channels and/or receptors such as V1, or act indirectly by activating G-protein coupled receptors such as cysteinyl leukotriene receptors. Cytochrome P450 epoxygenase converts AA into several epoxy-eicosatrienoic acids (EETs) derivatives. Specific lipids of the cyclooxygenase and lipoxygenase pathway directly activate V1, while lipids of the cytochrome P450 epoxygenase pathway activate V4. As shown, piroxicam blocks cycloxygenase activity while ETYA blocks the all three pathways.

86

A B AA 300 µM AA 16

) 300 µM 6 ) -4 100 µM TRPV3 ) -4 D641N 12 33.3 µM -4

(X 10 (X 4

o 8 (X 10(X

(x10 TRPV3 o 0 F - F 2 4 F - F F-F pcDNA3 0 0 012345012 Time (min) Time (min) Time (m) C ETYA 5

300 µM ) 4 -4 100 µM 33.3 µM 3 (X 10(X

o 2

F - F 1

0 012345 Time (min)

Fig. 3.9. Arachidonic acid and ETYA directly activate TRPV3. A, Dose- dependent activation of V3D641N transfected cells by AA. B, 300 µM AA activates 2+ both V3D641N and V3 transfected cells, with a greater rise of [Ca ]i occurring in V3D641N transfected cells. C, Dose dependent activation of V3D641N transfected cells by ETYA, a nonmetabolizable triple bonded analog of AA. A-C, Transfected cells were loaded with fluo4. Traces represent the averages of duplicate experiments.

87

membrane cytosol

PI(3,4,5)P3 PI(4,5)P2

DAG Ins(1,4,5)P3

Ins(1,4)P2 PI(3,4)P2 PA PI4P

PAO Ins(4)P CDP-DG LiCl

PI3P myo-inositol PI

4 3 OH P HO 2 P H 5 OH P DAG 6 1

Ins(1,4,5)P3

Fig. 3.10. PAO prevents the synthesis of PIP2. PIP2 [PI(4,5)P2] is a negatively charged phospholipid located on the inner leaflet of the plasma membrane. Cleavage of PIP2 by PLC results in diacylglycerol (DAG) and IP3 [Ins(1,4,5)P3]. Basal level activity of PLC results in a constant turn over of PIP2 in the cell. The PIP2 is replaced by a series of phosphorylations on phosphatidylinositol (PI), with PAO inhibiting the first kinase in the series, phosphatidylinositol 4-kinase. The balance between PIP2 turn over and PIP2 synthesis is tightly regulated to determine the resting [PIP2] in the cell. By inhibiting phosphatidylinositol 4-kinase, the basal level of turnover will result in a slow depletion of PIP2 from the membrane. In addition, PLC-linked GPCR activation will result in a rapid depletion of PIP2 from the membrane when PAO is present. DAG anchors the phosphatidylinositol derivatives to the membrane. The small gray numbers surrounding IP3 indicate the positions on the inositol sugar in which phosphates are placed on both the inositol derivatives (in the cytosol) and the phosphatidylinositol derivatives (membrane).

88

A

TRPV3D641N pcDNA3

B 30 µM PAO

o 8 V3 (37 C) V3 (32oC)

) o -4 6 V3D641N (32 C)

(X10

0 4

F-F 2

0 02468 Time (m)

Fig. 3.11. PAO activates TRPV3. A, PAO dose-dependently activates V3D641N (left) but not pcDNA3 (right) transfected cells. B, V3D641N is more sensitive to PAO than V3 at 32oC. At elevated temperatures (37 oC), V3 is activated much stronger when compared to 32 oC. A & B, Transfected cells were loaded with fluo4. Traces represent the averages of duplicate experiments.

89

A 30 µM PAO

10 µM RR 8

TRPV3D641N

) 6 -4

(X10 4

0 1.8 mM Ca2+/10 µM RR

F-F 1.8 mM Ca2+/ (-) RR 2 0 Ca2+/10 µM RR 0 Ca2+/ (-) RR 0 0246810 Time (m)

B TRPV3 TRPV3D641N 120 100

80

60

40

20 RR by µM 10 block % 0 (-) (+) (-) (+) RR RR RR RR

Fig. 3.12. Removal of extracellular Ca2+ and RR block the PAO-induced 2+ activation of TRPV3. A, the PAO induced rise of [Ca ]i in cells expressing 2+ TRPV3D641N is blocked by the removal of extracellular Ca , but is only partially blocked by RR. Only the traces for TRPV3D641N are shown, but TRPV3 was blocked by both RR and the removal of extracellular Ca2+. Transfected cells were loaded with fluo4. Traces represent the averages of duplicate experiments. B, Summarization of the RR block for both TRPV3 and TRPV3D641N for data points 8 minutes after the addition of PAO. Data points are the average of two experiments and error bars represent the range.

90

30 µM PAO

4

3 ) TRPV3 -5 TRPV3D641N

2 pcDNA (X10 0

F-F 1

0 0246810 Time (m)

Fig. 3.13. PAO-induced membrane depolarization of TRPV3 transfected cells. PAO caused large changes in the membrane potential of TRPV3 and TRPV3D641N, but not pcDNA3 transfected cells at 32oC. Cells were loaded with FLIPR membrane potential dye and changes in membrane potential were monitored with excitation/emission wavelengths of 530/565 nm. Traces represent the average of three experiments.

91

A 2APB 30 µM PAO 14

12

) [2APB] -4 10 40.0 µM 8 13.3 µM 4.44 µM 6 1.48 µM 4 0.49 µM F – 10 (X Fo 0.00 µM 2 0

0 123456 Time (min) B

EC50 ~ 6 µM

[2APB] µM

Fig. 3.14. PAO potentiates the 2APB response of TRPV3. A, TRPV3 and pcDNA3 cells were treated with 30 µM PAO 4 minutes prior to adding different concentrations of 2APB. pcDNA3 transfected cells did not respond to any of these concentrations so the pcDNA3 traces were subtracted from the TRPV3 traces. Traces are the average of duplicate experiments. B, dose-response curves for the PAO induced potentiation of 2APB. Points were obtained from (A) 1 minute after the addition of 2APB to cells pre-exposed to 30 µM PAO for 4 minutes.

92

Fig. 3.15. PAO potentiates the B2R and H1R-induced activation of TRPV3. A, PAO potentiates the B2R-induced activation of V3 in cells cotransfected with B2R and V3 (left). PAO does not affect the B2R response of cells cotransfected with B2R and pcDNA3 (right). B, PAO potentiates the H1R-induced activation of V3 in cells cotransfected with H1R and V3 (left). PAO does not affect the H1R response of cells cotransfected with H1R and pcDNA3 (right). A & B, Fluo4-loaded cells were stimulated with 30 µM PAO at 20 sec followed by 3µM BK or 300 µM His at 140 sec. Traces represent the averages of triplicate experiments.

93

Channels/receptors

PDZ1 PDZ2 PDZ3 SH3 GUK AKAP

PKC PKA

CLN channels/receptors and their C-term Signaling proteins and their C-term PDZ recognition sites aminos PDZ recognition sites aminos V3 FPETSV Cyclooxygenase 1 and 2 RPSTEL V1 AASGEK Phospholipase C-β1 EFDTPL Shaker K+ Channel KLLTDV Phospholipase C-β2 AQESRL Prostaglandin E2 Receptor SGLSHF Phospholipase C-β3 EENTQL Bradykinin Receptor 2 WAGSRQ Phospholipase C-β4 RPATVV P2 Purinergic (P2Y-1) Receptor NGDTSL Metabotropic Glutamate Receptor 5 QSSSSL Cysteinyl Leukotriene Receptor 1 EELCKV Cysteinyl Leukotriene Receptor 2 RKETRV

Fig. 3.16. PSD-95 as an example of the ability of PDZ domain containing proteins to form signaling complexes. PSD-95 is a modular protein that forms signaling complexes by clusters multiple channels, receptors, and signaling proteins to sites of cell-cell contact. PDZ domains 1-3 (PDZ1, PDZ2, PDZ3) bind to channels and receptors, as well as various isoforms of phospholipase C as long as the proteins have a PDZ recognition sequence that is specific for a particular PDZ domain. PDZ domains are classified as either type I or II depending on whether they recognize carboxyl-terminal peptides with the consensus sequence Thr/Ser-X-Phe/Val/Ala-COOH or Phe/Tyr-X-Phe/Val/Ala-COOH respectively. In addition to PDZ domains, PSD-95 has a src homology 3 (SH3) domain, and a protein-binding domain with homology to guanylyl kinase (GUK). The SH3 and GUK domains each bind the A kinase anchoring protein, AKAP79/150 (Colledge, et al., 2000). AKAP79/150 provides a scaffold in which PKA, PKC, and the calcium-dependent phosphatase calcineurin (CLN) can bind. By placing several protein binding domains in one protein, PSD-95 simultaneously interacts with several different proteins, thereby coupling the activity of channels and receptors to components of downstream signaling events. This type of complex promotes signaling speed and specificity by eliminating cross-talk to other signaling cascades and eliminating diffusion times for the interaction of activated signaling components. The box shows examples of channels/receptors and signaling proteins that have PDZ recognition sequences in their C-terminus and have been shown to bind PDZ domains. TRPV3 has a consensus PDZ-recognition sequence for PSD-95, TRPV1 does not.

94

CHAPTER 4

TRPV3 CHANNEL ACTIVITY IS MODULATED BY ACID

4.1 Introduction

Application of exogenous acid to the skin or a wound causes pain, resulting in the classification of acid as a nociceptive stimulus. In addition, the extracellular fluid of inflamed tissues becomes acidic, with a pH as low as 5.5 being common (Reeh and Steen,

1996; Lee et al., 2005). In fact, acid induced pain effects millions of people, as tissue acidosis occurs not only during wound healing, but in many pathophysiological states, including inflammation, arthritis, ischemia, and cancer (Steen et al., 1996; Issberner et al., 1996; Jones et al., 2004; Gopinath et al., 2005; Coutaux et al., 2005). Unfortunately, the pain associated with these pathologies can have severe physically and psychologically debilitating effects and cost the American public billions of dollars each year on compensation, litigation, and medical expenses (Stucky et al., 2001). Regardless of the origin, acid causes pain by modulating various receptors (nociceptive receptors) that are located at the terminals of the peripheral neurons that initiate the ascending pain pathway

(nociceptors). V1 and ASICs 1-4 are the best characterized acid-sensitive channels located at nociceptive terminals (Waldmann et al., 1997; Caterina et al., 1997; Tominaga

95 and Tominaga, 2005). ASIC channels respond to acid by causing a rapidly desensitizing influx of Na+, and are believed to be major components of the initial “sharp” sensation of acid induced pain (Zhang & Canessa, 2002; Hesselager et al., 2004). On the other hand, activation of V1 by acid results in an influx of both Na+ and Ca2+, with a more prolonged current that contributes to long-term pain, especially in the presence of other inflammatory mediators (Kress and Zeihofer, 1999). Regardless of which channels/receptors are modulated on nociceptors, activation of acid-sensitive channels/receptors are a major contributor to the inflammatory pain experienced by millions of people daily. Unfortunately, the properties of acid-induced currents on native nociceptors is not completely accounted for by the combined properties observed by the heterologous expression of V1 or ASIC channels, leading to the hypothesis that other unidentified channels/receptors contribute to acid-induced nociception (Jones et al.,

2004).

The discovery that protons can not only potentiate the capsaicin-induced activation of V1, but also directly activate V1 at pHs below 6.0, led to efforts to identify the mechanism in which protons modulate V1 channel activity (Tominaga and

Tominaiga, 2005). Mutagenesis studies using murine V1 revealed that protonation of a glutamic acid residue, Glu600, located in the extracellular loop (P-loop) between S5 and

S6 is responsible for potentiation of the capsaicin-induced activation of rat V1 (Jordt et al., 2000). In addition, this study demonstrated that direct activation by acid is a result of protonation of another glutamic acid residue, Glu648, which is also located on the P-loop

(Fig. 4.17). The identification of these sites represents an important step in the development of therapeutic strategies to manage inflammatory pain, a focus of many

96 drug discovery programs. In fact, therapeutically targeting V1 through capsaicin therapies has been a successful chronic pain management strategy, as evidenced by the development of several over-the-counter topical pain remedies with capsaicin as the active ingredient. Unfortunately, a comprehensive strategy for pain management will only come via the therapeutic targeting of V1 and ASIC channels, as well as other nociceptive channels/receptors involved in the induction of chronic/inflammatory pain, a process that will require the identification of the channel/receptors responsible.

V3 is the newest member of the heat-activated TRPV channels, with reported Ths between 34 and 39 oC (Xu et al., 2002; Smith et al., 2002; Peier et al., 2002). Although heterologously expressed V3 is activated at innocuous temperatures, heat-induced currents through V3 increase in amplitude well into the noxious temperature range, and may therefore contribute to the detection of painful stimuli in a manner similar to V1. In addition, thermal activation of V3 in keratinocytes requires temperatures that are in the noxious temperature range (Chung et al., 2004a). In fact, studies using trpv1 and trpv3 knockout mice suggest that V3 may play a larger role in noxious temperature sensation than V1 (Zimmermann et al., 2005; Moqrich et al., 2005). Upregulation of V3 in injured nerves, as well as in patients experiencing severe breast cancer pain, suggest V3 plays a role in pain sensation (Smith et al., 2002; Gopinath et al., 2005). These findings, as well as the high degree of sequence homology, shared chromosomal co-localization and co- expression in the DRG of a subset of thinly-myelinated C-fibres, suggest that V3 and V1 may share functional properties such as acid gating. In spite of this, the original characterization of V3 concluded that it was irresponsive to acid. However, our data

97 demonstrates that V3 is modulated by acid when expressed heterologously in HEK293 cells, making V3 a potential target for therapies of inflammatory pain management and/or other pathologies associated with inflammation.

4.2 Material and Methods

4.2.1 DNA constructs, cell culture, and transfections

All compounds and reagents were purchased from Fisher or Sigma unless otherwise stated. 2APB was from Tocris or Cayman Chemical Company. All cDNAs were cloned into pcDNA3.

4.2.2 Acidification of ECS solution for intracellular Ca2+ measurements using the Flex Station

To acidify the ECS buffer in 96-well plates, a modified ECS solution was made:

140 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2, 10 mM glucose, and 45 mM

Hepes. By not adjusting the pH of this solution, addition of 1/3 the volume to the bathing solution resulted in a pH of 5.6. For the rest of the procedure see section 2.2.3.

98 4.2.3 Intracellular Ca2+ measurements using fura-2

For fura-2 measurements, cells were plated at low densities on glass coverslips the night before. Cells were loaded with fura2 and fluorescence intensities of a group of cells were read by a photomultiplier attached to a Nikon Eclipse TE200 inverted microscope with alternating excitation wavelengths of 340 and 380 nm and the emission wavelength of 510 nm. Drugs were diluted to the final concentration with ECS and applied through perfusion.

4.2.4 For all other experimental procedures, see Section 2.2.

4.3 Results

2+ 4.3.1 Acid causes an increase in [Ca ]i in TRPV3D641N expressing cells at lower temperatures

Although V1 and V3 are both heat-gated channels with similar expression patterns, the original characterization of V3 concluded that, unlike V1, V3 is not activated by capsaicin or low pH. Additionally, chapter 2 demonstrates that mutating

641 Asp to Asn in V3 (V3D641N) results in a similar loss of RR-sensitivity as has been reported for an equivalent mutation in V1 (V1D647N) (Garcia-Martinez et al., 2000). We therefore hoped to demonstrate that coexpression of V1 with V3D641N in HEK293 cells would result in a capsaicin or pH responsive channel/receptor with reduced RR-

99 sensitivity, thus unequivocally showing that V1 and V3 form functional heterotetramers

(see Chapter 5). To our surprise though, lowering the pH of the bath solution from 7.4 to

o 2+ 5.6 at 26 C in HEK293 cells expressing V3D641N resulted in a rise of [Ca ]i to a similar extent as V1 and V1D647N expressing cells (Fig. 4.1). Conversely, the acid response of cells expressing wild type V3 was not increased relative to the endogenous response observed in cells transfected with pcDNA3. The source of the endogenous acid response is not known.

2+ 4.3.2 The acid-induced increase in [Ca ]i in both TRPV3 and TRPV3D641N expressing cells is temperature dependent

A careful analysis of the temperatures used in our studies (26 oC) and others

(room temperature), as well as the observation that higher temperatures work synergistically with 2APB to activate V3 (Fig. 2.9B), suggest that the lack of a response to acid by V3 may be a result of the low temperatures in which these studies were performed (Xu et al., 2002; Smith et al., 2002; Peier et al., 2002). Additionally, our studies of the effects of 2APB on V3 revealed that V3D641N was more sensitive to 2APB than wild type V3 (Fig. 2.1). In light of these observations, we reasoned that the response

o of V3D641N, but not wild type V3, to acid at 26 C could be due to an increase in sensitivity of V3D641N, and that wild type V3 may respond to acid at higher temperatures.

Figure 4.2A shows the acid response (pH 5.6) of cells expressing V3, V3D641N, or pcDNA3 at 26, 32, 37, 42 and 45 oC. As should be expected for a thermally gated channel, the basal level activity increased at higher temperatures for V3 and V3D641N

100 expressing cells, but not cells transfected with pcDNA3 (summarized in Fig. 4.2B). The basal response at higher temperatures was much higher for V3D641N than for V3, further demonstrating that V3D641N is more sensitive than the wild type channel (Section 2.4).

2+ The rise in [Ca ]i following application of acid to V3 transfected cells is not seen at 26 oC, but increases with increasing temperature starting at 32 oC. In contrast, the acid-

2+ induced rise in [Ca ]i for V3D641N transfected cells increases with increasing temperature between 26 and 37 oC, but declines at temperatures higher that 37 oC. It should be noted

o that the decline in V3D641N activity above 37 C is likely due to the high basal level activity of V3D641N at these temperatures, as the steepness of the increase in fluo4 fluorescence does not decrease at temperatures above 37 oC. Interestingly, the rate of

2+ acid induced rise in [Ca ]i was much slower for V3 transfected cells compared to

V3D641N transfected cells at all temperatures tested. Based on these observations, we designated 32 oC as the optimal temperature for fluorometric analysis of the acid-induced modulation of V3 and V3D641N.

4.3.3 Acid activates TRPV3 on the plasma membrane.

Localization studies (Fig. 2.4), as well as experiments using 2APB in the presence and absence of extracellular Ca2+, suggest that V3 is localized predominantly to the

2+ plasma membrane (Section 2.3.3). Therefore, if the acid-induced rise in [Ca ]i is the result of V3 activation, the response should be drastically reduced upon removal of Ca2+ from the extracellular solution. Figure 4.3A shows the acid response to both V3 and

2+ o V3D641N in absence and presence extracellular Ca at 32 C. In the absence of

101 2+ extracellular Ca , the acid response of both V3 and V3D641N is equivalent to the response

2+ of the pcDNA3 controls. The transient rise of [Ca ]i that remains in the nominally calcium-free solution appears to involve Ca2+-release from an intracellular storage organelle such as the endoplasmic recticulum (ER). Although studies with 2APB in a nominally Ca2+-free solution revealed that 166 µM 2APB activates V3 located on both the plasma membrane and intracellular Ca2+ storage organelles (see section 2.3.3), the results observed in Fig 4.3A are expected because, while 2APB is lipophilic and can traverse the plasma membrane, protons carry a positive charge and should not cross the plasma membrane to a significant extent under the conditions of these experiments.

Mn2+-quenching experiments confirm that V3 is localized to the plasma

o membrane, as fura-2 fluorescence quenching at 37 C is observed in V3D641N and V3 but not pcDNA3 transfected cells (Fig. 4.3B; see section 2.3.3 for an explanation of Mn2+- quenching). The level of Mn2+-quenching observed in V3 expressing cells is reduced compared to V3D641N. Although this general trend is expected, Figure 4.2A suggests that the acid induced V3 activation is significantly slower than V3D641N at this temperature.

This apparent anomaly can be explained by the difference between the two methods.

2+ 2+ Fluo4 measurements of [Ca ]i are affected not only by the influx of Ca into the intracellular space (increased fluorescence), but by extrusion of Ca2+ from the intracellular space back out of the cell as well as into intracellular storage organelles

(each resulting in decreased fluorescence). Therefore, a slow activation rate for V3

2+ 2+ would result in a very small net increase in [Ca ]i as long as the rate of Ca efflux is comparable. Conversely, Mn2+ rapidly accumulates in cells because it is not extruded by

102 the transporters/channels that remove Ca2+ from the intracellular space, which explains the difference in the apparent relative activation of V3 compared to V3D641N when using the two different methods.

4.3.4 Capsaicin inhibits the acid-induced activation of TRPV3D641N

To confirm that the response to acid in the above experiments was due to the activation of V3 and V3D641N, we tested the ability of the V3 antagonist, RR (not shown) and capsaicin (Fig. 4.4), to block the acid response. Application of differing concentrations of capsaicin two minutes prior to lowering the pH to 5.6 inhibited the acid-induced response to V3D641N in a dose-dependent manner with an IC50 value of approximately 25 µM (Fig. 4.4B), confirming specificity of the acid response to V3 (see

Section 2.3.9 and Fig. 2.11). A precise IC50 value was not calculated because capsaicin is not soluble at concentrations greater than 300 µM, preventing a thorough construction of the inhibition curve. Unlike with capsaicin, addition of 10 µM RR two minutes prior to lowering the pH to 5.6 did not block the acid response of V3 or V1 at 37 oC (data not shown). Presumably, a strong electrostatic interaction between positively charged ammonium groups of RR and negatively charged acidic residues on the pore of TRPV channels is critical for blockade (Garcia-Martinez et al., 2000). In fact, the mutation underlying the loss of RR-sensitivity in V3D641N mimics protonation at this critical site, resulting in a loss of a critical negative charge required for interaction with RR. Since protonation of acidic residues on the P-loop of V1 has been shown to underlie the acid induced modulation of V1, it is reasonable to assume a similar process underlies the acid-

103 induced activation of V3. If in fact, Asp641 and/or other acidic residues crucial to RR binding are protonated at pH 5.6, this would explain the inability of RR to inhibit the acid response of V1 and V3.

4.3.5 Acid activates currents in HEK293 cells expressing TRPV3 and TRPV3D641N

RR-sensitivity is a fingerprint of TRPV channels. Since the acid-induced activation of V3 hinders the use of RR in identifying specificity of acid towards V3, we monitored voltage-clamped HEK293 cells expressing V3 and V3D641N in the whole-cell- mode in order to determine if acidification of the extracellular solution would yield currents with similar characteristics as those induced by 2APB (Fig. 2.5B & C) or heat.

As with 2APB, the V3 response to pH 5.5 resulted in outwardly-rectifying currents with a reversal potential near 0 mV (Fig. 4.5B). The acid response to V3D641N yielded a similar shape of the I-V curve at pH 6.5, but became increasingly linear at pH 5.5 (Fig 4.5A). A similar phenomenon occurs when V3D641N is activated by 2APB, with higher concentrations leading to a progressively more linear I-V curve (Fig. 2.5C). These data are also in agreement with those seen with the fluorometric analysis, as the V3 response to pH 5.5 was not as pronounced as that of V3D641N. Although these data confirm that the acid response is specific for V3, it is important to note that not all cells transfected with

V3 or V3D641N (as determined by green fluorescence as well as responsiveness to 2APB) respond to acid, a phenomenon that is similar to that of the heat response of V3 in keratinocytes (Chung et al., 2004a; Section 2.4).

104 4.3.6 Acid potentiates the 2APB response of TRPV3 transfected cells

The observation that 2APB and heat work synergistically to activate V3 (Fig

2.9B) reveals a common characteristic between V3 and V1, as acid, capsaicin, heat, and

2APB have a synergistic effect on V1 activation when more than one stimuli is added simultaneously. Therefore, we reasoned that acid and 2APB may also work synergistically to activate V3. To test this, different concentrations of 2APB were applied to fluo4 loaded HEK293 cells transfected with V3, V3D641N, or pcDNA3, two minutes prior to lowering the pH from 7.4 to 5.6 (Fig. 4.6A). As clearly demonstrated, acid potentiates the effects of 2APB at all concentrations. Dose response curves for

2APB reveal a left-shift at pH 5.6, with V3 having an EC50 value of 57.7 µM at pH 7.4 versus 3.83 µM at pH 5.6 and V3D641N having an EC50 value of 23.9 µM at pH 7.4 versus

1.13 µM at pH 5.6 (Fig. 4.6B). As expected, the simultaneous addition of acid and 2APB results in a similar potentiation of the 2APB response for V3 transfected cells when compared to the step-wise addition of 2APB followed by acidification, with an EC50 value of approximately 3 µM for the (acid + 2APB)-induced activation of V3 at pH 5.6

(Fig. 4.7). These data demonstrate that V3, like V1, has an inherent propensity for synergistic behavior between stimuli. This is important because, being expressed in both keratinocytes and peripheral sensory neurons, V3 may often be subjected to multiple subthreshold stimuli, that when encountered together, may result in V3 activation. Such a situation exists in the case of inflammation, where several potential activators of V3 surround keratinocytes and/or peripheral sensory neurons, two tissues in which V3 is expressed and hypothesized to have function.

105 4.3.7 Optimal pH for TRPV3 activation

So far, our experiments using “acid” as an activator have been performed at a pH of 5.6. This pH was chosen as an experimental convenience, as mixing 80 µl of the bathing solution with 40 µl of our stock acidic solution results in a final pH of 5.6. It is not unreasonable to imagine that pH 5.6 is not the optimal pH for activating V3.

Unfortunately, the combined effects of a low level of acid-induced activation using fluorometric analysis, as well as the presence of an endogenous Ca2+-release in HEK293 cells, complicates efforts to determine an optimal pH for direct activation of V3 by acid.

Additionally, the acknowledgement that part of the acid activation mechanism could

641 entail the protonation of Asp cautioned us not to use V3D641N as a model for direct activation by acid without verification that V3 has a similar pH profile. We therefore measured the ability of different pHs to potentiate the response of 6 µM 2APB in V3 and

V3D641N transfected cells and used this as an estimate of the optimal pH for V3 activation

(Fig. 4.8). Figure 4.8A shows traces of the potentiation of different pHs by 6 µM 2APB in cells transfected with V3 (V3D641N traces not shown). Figure 4.8B is a summary of the flou4 fluorescence observed from V3D641N transfected cells for different pHs 45 seconds after the addition of 6 µM 2APB. As seen, the acid induced potentiation of the 2APB response for both V3 and V3D641N reaches a maximum at pH 5.6, with an increase in potentiation occurring as the pH is incrementally lowered from 7.4 to 5.6 and decreasing upon a further reduction below pH 5.6. Although these data demonstrate that pH 5.6 is the optimal pH tested, a more thorough analysis using small incremental changes of pH in the vicinity of pH 5.6 would be necessary for a true quantitative determination of the

106 optimal pH for V3 activation and/or potentiation, but these results confirm that the pH 5.6 is a good estimate of the optimal pH for enhancement of V3 activity. It should be noted

2+ that the acid only-induced rise of [Ca ]i in control cells transfected with pcDNA3 were similar to those of V3 transfected cells at all pHs tested (not shown).

We next tested the ability of different pHs to potentiate the 2APB-induced currents in HEK293 cells transfected with V3 (Fig. 4.9). Cells were voltage-clamped in the whole-cell mode and recorded using voltage ramps. For pH 5.5, 6.5, and 7.0, cells were stimulated with 100 µM 2APB without prior stimulation, whereas for pH 8.0, cells were stimulated with 100 µM 2APB several times prior in order to cause a maximal response (see Sections 2.3.7 & 2.4 for an explanation of V3 sensitization by 2APB). I-V curves were generated when the following stimuli reached a plateau; 2APB at pH 7.4 (a),

2APB at a different pH (b), and 2APB at pH 7.4 following coapplication of 2APB and the different pH (c). At pH 8.0 (Fig. 4.9A), 2APB induced currents decreased relative to those at pH 7.4. Lowering the pH back to 7.4 in the continued presence of 2APB caused only a slight decrease in the current level compared to that observed before raising the pH. After perfusing the cell with bathing solution for several minutes, the simultaneous addition of 2APB and pH 8.0 resulted in similar currents as those seen previously under these same conditions, indicating that any effect of raising the pH is washed out quickly.

Figure 4.9(B, C, D) shows that a slight drop in pH from 7.4 to 7.0 results in a minor potentiation of the response to 100 µM 2APB, whereas a larger drop in pH to 6.5 or 5.5 each have a significant potentiating effect. Interestingly, for each pH below 7.4 that was tested, raising the pH of the 100 µM 2APB perfusate back to 7.4 resulted in a current amplitude that was higher than the previous 2APB perfusion at pH 7.4 (compare

107 trace a and c of Figure 4.9 B, C, D), a phenomenon that is consistent with V3 sensitization. Interestingly, the acid-induced potentiation of the 2APB response is observed in all V3 transfected cells, indicating that the lack of direct acid activation observed in some V3 transfected cells is the result of acid needing a costimulator in those cells (Section 4.3.5). Interestingly, other studies have shown that heat induced activation of V3 in keratinocytes, as well as oocytes, requires 2APB as a costimulant for consistant activation of V3 (Chung et al., 2004a; Asakawa et al., 2006; Fig. 2.9B, Section 2.4).

4.3.8 Repetitive acid stimulation results in TRPV3 sensitization

.

Sensitization of V3 occurs when either a repetitive challenge of heat (Peier et al.,

2002; Fig. 2.7) or chemical activator (Fig. 2.8) are administered in a short time interval following the preceding stimulation, or when the activating stimulus is administered over an extended period of time. In fact, sensitization of V3 by repetitive activation has been shown for 2APB, diphenylboronic anhydride, and the recently identified skin sensitizers camphor, carvacrol, thymol, ethyl vanillin, and eugenol (Xu et al., 2005; Xu et al., 2006).

To see if acid causes V3 sensitization in a manner similar to heat or chemical activators,

2+ we monitored the acid induced change of [Ca ]i in HEK293 cells transiently transfected with V3D641N or pcDNA3. Cells were loaded with fura-2 and monitored at room temperature (Fig 4.10). Consistent with previous results, acid (pH 5.0) induced a slow

2+ 2+ rise in [Ca ]i that was still increasing after 25 minutes. Removal of Ca from the

2+ extracellular buffer resulted in a reduction of the [Ca ]i, that was further reduced upon removal of the acid stimulus. Reapplication of acid while maintaining the absence of

108 2+ 2+ extracellular Ca resulted in a small, sudden rise of [Ca ]i that quickly reached a maximum value and resembled that seen under the same conditions in pcDNA3 transfected cells (Fig. 4.10A & B). Adding Ca2+ back to the extracellular buffer resulted

2+ in a more rapid rise in [Ca ]i than was observed upon the original application of acid in the presence of extracellular Ca2+. In addition, the second application of acid in the

2+ 2+ presence of extracellular Ca resulted in [Ca ]i that was much greater than that which was observed at the end of the first stimulation of acid in the presence of extracellular

Ca2+. It was not determined if the removal of extracellular Ca2+ while maintaining the presence of acid stimulation, or just the fact that the acid stimulus was removed for a time interval before subsequent stimulation, was responsible for the large increased activity of

V3, but the results clearly show that acid activation causes sensitization by both the continued presence of stimulation and by the removal of stimulus followed by subsequent stimulation. It should also be noted that not all V3 transfected cells responded to acid using this method (section 4.3.5). Like with the electrophysiology experiments, but unlike the fluorometric studies using fluo4 in the FLEX station, cells are seeded at a low cell density for these experiments, suggesting cell density may influence the “sensitivity” of V3. In fact, nearby cells excrete factors that effect several signaling pathways that may sensitize V3 (Section 3.4).

We next measured voltage-clamped HEK cells expressing V3 or V3D641N to determine if successive challenges of acid would sensitize V3 currents. As shown, application of pH 6.0 does not result in V3 activation (Fig. 4.11A). Lowering the pH to

5.5 results in a small, outwardly-rectifying current, with slow activation kinetics (Fig.

4.11A & B, Fig. 4.12C). Consistent with sensitization, each subsequent application of

109 pH 5.5 resulted in increasingly larger currents with faster activation kinetics. The sensitization displayed a lasting effect, as further sensitization occurs following a 4.5 minute washout (see Fig 4.11A & B, 5th pH 5.5 application). Application of 2APB at pH

5.5 and 7.4 confirms the specificity toward V3. Additionally, V3D641N was activated and displayed sensitization of subsequent stimulations at pH 6.0 and 5.5, with much faster activation kinetics than wild type V3 at pH 5.5 (Fig 4.12). In addition, V3D641N activation kinetics reached a near maximum value by the second stimulation with pH 5.5, consistent with the finding that Asp641 is a critical site for inactivation by Ca2+ (Xiao et al., 2006; section 2.4). These results demonstrate that, like with other known activators of V3

(heat, 2APB, diphenylboronic anhydride, camphor, carvacrol, thymol, ethyl vanillin, and eugenol), sequential stimulation with acid results in sensitization of V3 currents. This property of acid-induced V3 activation should have a profound impact on cells expressing V3 under inflammatory conditions, where a significant reduction in the pH of the extracellular fluid is the norm.

4.3.9 Signaling events that affect the acid-induced activation of TRPV3

The experiments above have focused on characterizing the response of V3 to acid, but do not demonstrate the mechanism in which V3 is activated and/or potentiated by acid. As discussed throughout chapter 3, activation of GPCRs enhances V3 activity through events downstream of phospholipase C (PLC) activation. To summarize, activation of PLC results in the cleavage of PIP2 and generation of IP3 and diacylglycerol (DAG). It is well established that PIP2 holds V1 in and inactive state,

110 with depletion of PIP2 from the plasma membrane resulting in a more sensitive channel

(Prescott and Julius, 2003). Additionally, it was shown in Section 3.3.5.4 that V3 activity is also enhanced by depletion of PIP2 from the plasma membrane. Furthermore, DAG can either activate PKC or be cleaved by DAG lipase, releasing arachidonic acid (AA) and its metabolites, each of which have been shown to enhance V1 (AA metabolites) and

V3 (AA) activity (Fig 3.1 and section 3.4). IP3, results in Ca2+ release from the endoplasmic reticulum by binding to IP3 receptors, which has been shown to activate several TRP channels in a process known as store operation (Parekh, 2005). Since application of acid to HEK293 cells results in Ca2+ release from an unknown intracellular storage organelle (Fig 4.3A), we reasoned that acid could activate an endogenous GPCR in HEK293 cells, and that the acid induced activation of this receptor could modulate V3 indirectly through any one of the signaling events downstream of PLC activation.

Therefore, blockers/activators of various signaling pathways were applied to V3D641N or pcDNA3 transfected HEK293 cells 30 minutes prior to acid addition (pH 5.6) and

2+ changes in [Ca ]i were monitored via fluo4 fluorescence (Fig. 4.13). None of the drugs used in these experiment had an affect on the endogenous acid response (data not shown).

In addition, only the traces in which the acid response was significantly affected by drugs are shown in Figure 4.13A, with a summary of the effect of each drug tested displayed in

Figure 4.13B. Neither the PLD inhibitor D609 (3 µM) nor the PLC inhibitor U73122 (15

µM) blocked the acid-induced activation of V3D641N. Interestingly, U73122 slightly potentiated the acid response of V3D641N. Subsequent testing using an inactive analog of

U73122, U73343, also potentiated the acid response of V3D641N (not shown), demonstrating that the potentiation is unrelated to PLC activity (the enhancement of V3

111 activity by U73122 and U73134 is discussed in Chapter 6). In addition, the PKC blockers GF109203X (GF) (0.1 µM) and CalphostinC (1.5 µM) did not effect the acid induced activation of V3D641N or the endogenous response, demonstrating that the acid induced modulation of V3 does not involve PKC activation. Neither ETYA (20 µM), a non metabolizable triple-bonded analog of AA that inhibits AA metabolism, nor the cycloxygenase inhibitor Piroxicam (60 µM), were able to block acid induced V3D641N activation, demonstrating that metabolites of arachidonic acid are not responsible for the acid response.

PKA has been shown to enhance V1 activity by phosphorylation of S502 on the murine channel (Tominaga & Tominaga, 2005). The PKA blocker, H89 (10 µM) significantly blocked the acid response of V3D641N (shown), suggesting that phosphorylation of V3 by PKA may also modulate V3 activity. In addition, higher concentrations of GF (1-30 µM) dose-dependently inactivated the acid response of

V3D641N, with 30 µM having a similar effect as 10 µM H89. Generally, 0.1 µM GF will fully block PKC, with higher concentrations having inhibitory effects on other kinases, including PKA. Neither the irreversible adenylyl cyclase inhibitor, MDL-12,330A (50

µM) nor the adenylyl cyclase activator, Forskolin (3 µM) had a significant effect on the acid induced activation of V3D641N. These result suggest that phosphorylation of V3 by

PKA may be essential for acid-induced activation. Whether preincubation of PKA inhibitors results in dephosphorylation of a constitutively phosphorylated site or prevention of phosphorylation by acid induced activation of PKA is not clear by these data, but the effect of PKA on the acid induced activation of V3 does not appear to be enhanced by increasing the production of cAMP.

112 4.3.10 Mutational analysis of Thr378, a PKA consensus site of TRPV3

V3 has one consensus PKA site, Thr378 (human) that is located on the N-terminal cytoplasmic domain (Smith et al., 2002). We therefore mutated the equivalent site

378 (Thr ) in the murine V3D641N background to see if the phosphorylation state at this site would affect the ability of H89 to inhibit the acid induced activation of V3D641N. As

378 shown, HEK293 cells expressing V3D641N or V3D641N with Thr mutated to alanine

2+ (V3DNTA), responded with a similar increase in [Ca ]i to both 166 µM 2APB or pH 5.6

(Fig. 4.14A & B; Fig. 4.16, black box, lower middle). In addition, 10 µM H89 caused a significant reduction in the acid response for cells expressing V3DNTA or V3D641N, suggesting that the phosphorylation of Thr378 by PKA is not responsible for the acid-

378 induced activation of V3 (Fig. 4.14C). We also mutated Thr to Asp (V3DNTD), a mutation that should mimic constitutive phosphorylation at this site, and performed

2+ similar experiments. Both the pH 5.6 and 166 µM 2APB induced rise in [Ca ]i was reduced relative to V3D641N, indicating that phosphorylation of this site results in V3D641N channels that are reduced in activity regardless of the mode of stimulation (Fig 4.14A and

B). The acid induced activity of V3DNTD was also reduced by H89, confirming that the phosphorylation state of V3 at this site is not responsible for the PKA induced modulation of the acid response (Fig 4.15D).

113 4.3.11 Acid-induced activation of other thermoTRPV channels

It is well established that acid not only potentiates the response of other activators of V1, but can also directly activate the channel below pH 6.0. In addition, although it has been reported to be acid-insensitive, the acid-induced activation of V4 has been demonstrated (Suzuki et al., 2003). Likewise, we have shown that V3 is modulated by acid in contrast to previous reports. We therefore applied pH 5.6 to each member of the

TRPV family, as well as M8 and TPC1, two members of the TRP superfamily that are not expected to show a significant response to acid, in order to see if other members of the TRPV family are acid-sensitive (Fig 4.15). Only traces for the thermoTRPVs are shown (Fig 4.15A), with a summary for all TRPs tested shown in Figure 4.15B. As previously reported, V1 and V4 are activated by acid (pH 5.6) at 32 oC. Interestingly, V2 is also strongly activated by acid at this temperature, a result that has not been previously reported. Like V3, but unlike V4 (data not shown), acid strongly potentiates the response of V2 to 2APB (Fig. 4.16, see Fig. 4.6 for V3). V5 and TPC1 showed a slight increase in the endogenous acid response, whereas M8 showed a slight decrease in the basal level of

2+ [Ca ]i. V6 is a calcium transporter with a high level of constitutive activity. The

2+ relatively large decrease in [Ca ]i in response to acid indicates that V6 is inactivated by acid. These data demonstrate that each of the thermoTRPVs show enhanced activity in response to acid. Interestingly, V3 appears to be the most resistant to acid activation alone, explaining previous reports that it is insensitive to acid. Interestingly, this quality of V3, coupled to the expression of V3 in keratinocytes and peripheral C-fibres, suggests

114 V3 is an ideal candidate for transduction of stronger signals during periods of inflammation compared to normal conditions (discussed in more below in Section 4.4).

4.4 DISCUSSION

In these studies, acid has been shown to enhance V3 activity, with a maximum effect around pH 5.6. To summarize, acid causes calcium influx in HEK293 cells transiently expressing V3, but does so only at temperatures at or above body temperature, indicating that protons are relatively weak agonists of V3. Nevertheless, acid strongly and consistently potentiates the 2APB-induced activation of V3, left-shifting the EC50 value of 2APB toward V3 approximately 15 fold. Using electrophysiological techniques, acid-induced currents in some, but not all HEK293 cells expressing V3. The acid induced currents of V3 transfected cells resembled those of heat induced currents, displaying outward rectification with a reversal potential around 0 mV. At pH 5.5, the acid-induced currents were slow to develop but sensitized in response to repeated stimulation. Interestingly, acid potentiated 2APB-induced currents in all V3 expressing cells tested, indicating that acid requires a coactivator for consistent activation of V3.

This observation is similar to that observed with heat activation of V3 in keratinocytes, as

2APB is required as a costimulant for heat activation of V3 in these cells (Chung et al.,

2004a; Asakawa et al., 2006; Section 2.4).

The mechanism by which protons enhance V3 activity was not determined in these studies. The most plausible hypothesis is that protonation of acidic residues between transmembrane domain five and six are responsible via a mechanism similar to

115 V1. A careful analysis of this region of murine V3 shows that E647 is at the equivalent position as E648 of murine V1 (Fig 4.17, red box, top). Protonation of E648 results in direct activation of V1 by acid. Interestingly, V1 and V3 are the only members of the

TRPV family to have an acidic residue at this position. In addition, N608 is at the equivalent site of V3 as E600 in on the murine V1. Protonation of this site is responsible for acid induced potentiation of V1 activity. An arginine located at this site should mimic protonation of an acidic residue, although the significance of having an arginine instead of glutamic acid at this position on V3 can only be determined via a mutational analysis.

Most likely, the modulation of V3 by protons will depend on the protonation state of several acidic residues on V3 (Fig 4.17, blue box, upper, upper). This can be a daunting task to determine by mutational analysis because many acidic residues can be protonated, some that inhibit activity, some that enhance activity, and some that only effect activity under certain conditions. In addition, a comparison of the acidic residues between transmembrane domains 5 and 6 of the TRPV family shows a great level of diversity with regards to specific location of acidic residues, making a sequence dependent selection of critical sites difficult (Fig 4.17, blue box, upper & red box, upper). This type of complexity mirrors that observed when trying to identify the importance of individual serines or threonines on channels/receptors that are regulated by phosphorylation.

Although the mechanism in which protons enhance V3 activity was not determined in these studies, the fact that the PKA inhibitor H89 strongly decreased the acid-induced activation of V3D641N (Fig. 4.14) suggests that phosphorylation by PKA may play a role in the acid-induced enhancement of V3. It should be noted that H89 could cause a constitutively phosphorylated PKA site to become unphosphorylated, with H89

116 only serving to prevent a PKA-dependent phosphorylation event that takes place in the absence of acid stimulation, resulting in a less sensitive V3 channel. In fact, Ser116 on murine V1 is constitutively phosphorylated by PKA in vivo, with dephosphorylation being necessary for V1 to enter a desensitized state. Prostaglandins minimize V1 desensitization by maintaining the phosphorylation of Ser116 through PKA, a phenomenon that allows V1 activity to persist in the presence of inflammatory mediators.

Although V3 has an arginine (Arg121) at the equivalent position, other serines or threonines located on V3 could impart a similar type of regulation on V3 (Fig. 4.17 green box, lower, left). In fact, PKA has been shown to phosphorylate Ser502 on the human V1 receptor, with phosphorylation being dependent on a PKA/AKAP/V1 signaling complex and resulting in a channel that is not desensitized upon activation (Rathee et al., 2002).

Interestingly, V3 has a serine (Ser511) located at the equivalent position as Ser502 of rat V1

(Fig. 4.17, brown box, lower, right). It is possible a similar PKA-dependent mechanism operates in V3, especially since V3 is the only thermoTRPV channel that has a consensus

PDZ domain binding sequence at its C-terminus (Fig. 3.16; Section 3.1 & 3.4). By clustering V3 and PKA into the same signaling complex, a small constitutive activity of

PKA can result in constitutive phosphorylation of PKA sites on V3. Such a mechanism would suggest that extracellular acidification does not induce activation of V3 by stimulating PKA activity, but that a decrease in PKA activity by H89 results in less sensitive V3 channels that mimic a desensitized state. The possibility that acid induces a

PKA-dependent phosphorylation of V3 should still be considered. Mutagenesis of Ser511

117 on murine V3 followed by similar experiments performed in Fig. 4.14 may be useful in determining the role that PKA phosphorylation plays in the acid-induced activation of

V3.

In heterologous expression systems, acid activation of V1 is followed by rapid desensitization (Jordt et al., 2000). Desensitization of V1 is prevented in nociceptive C- fibres by the phosporylation of V1 following activation of various receptors on the nerve terminals (Fig. 3.1). As an example, NGF dramatically increases in concentration under inflammatory conditions. Upon binding to TrKA receptors located at peripheral nerve terminals, NGF causes PKC activation and ultimately phosphorylation of V1. This scenario prevents the desensitization that is observed with heterologously expressed V1.

2+ Interestingly, increases in [Ca ]i of keratinocytes results in release of several inflammatory mediators including NGF (Shu and Mendell, 2001; Xu et al., 2006,

Asakawa et al., 2006). Since V3 activity is enhanced by acidification of the extracellular solution, a phenomenon that is associated with inflammatory states, this should increase

2+ the [Ca ]i in keratinocytes and ultimately the release of NGF as well as other inflammatory mediators. The increased concentration of NGF should in turn enhance V1 activity via the PKC-dependent phosphoryltation of V1 (Hwang and Oh, 2002).

Interestingly, in rats, primates, and humans, V3 is also expressed on nociceptive C-fibres, suggesting that, like with V1, activation of V3 in keratinocytes may enhance the activity of V3 on adjacent nociceptors, resulting in both autocrinal (keratinocytes to keratinocytes) as well as paracrinal (keratinocytes to peripheral nociceptors)

118 amplification of V3 activity. A more thorough hypothesis for the amplification of V3 activity via cross-talk between immonocytes, keratinocytes, and nociceptors was provided in Section 3.4.

As previously mentioned, acid is recognized as being a central player in the transduction of pain during inflammation, where the pH of the extracellular solution can drop to as low as 5.5. Additionally, each of the thermoTRPVs are believed to play a role in the transduction of painful signals from the periphery to the central nervous system

(Patapoutian et al., 2003; Lee et al., 2005). The observation that each thermoTRPV family members except V3 was directly activated by acid (pH 5.6) at 32 oC (Fig 4.15) may lead to the hypothesis that V1, V2 and V4 play a larger role in the acid-induced transduction of painful stimuli under conditions of inflammation than V3. This line of thinking can easily be overcome when considering that, in addition to low pH, the extracellular milieu under inflammatory conditions is associated with several properties that have been identified as enhancing V3 activity including infiltration of immunocytes, higher than normal concentrations of agonists specific to various G(q/11)PCRs, elevated levels of AA and or oxidized products of AA metabolism, lower than normal pH, and increased temperature due to vasodilation (especially near the skin). For example, in

Section 2.4 it was discussed that V3 is normally in a semi-quiescent state, and that sensitization of V3 results from the relief of a Ca2+-induced inhibition (Asp641- dependent). Interestingly, V3D641N mimics V3 in the sensitized state, and this chapter has demonstrated that V3D641N is strongly activated by acid. In fact, compared to V3, V3D641N was directly activated by acid at lower temperatures (Fig. 4.2), exhibited faster activation kinetics (Fig. 4.11 & 4.12), was nearly fully sensitized upon the first stimulation (Fig.

119 4.12), and displayed a more linear rectification profile (Fig. 4.5). These observations are consistent with those observed when comparing the 2APB (Section 2.3.6), G(q/11)PCR

(Section 3.3.1) or PAO (Section 3.3.5.4) induced activation properties of V3 to V3D641N, nicely demonstrating not only the specificity of acid for V3 activation, but that in the sensitized state, acid should act as a strong agonist of wild type V3. Furthermore, it was demonstrated in Chapter 3 that V3 activity is enhanced by signaling events downstream of G(q/11)PCR activation including PKA phosphorylation/dephosphorylation, PIP2 hydrolysis and AA and/or the oxidized products of AA metabolism. Putting this all together, the extracellular solution surrounding keratinocytes and nociceptors should provide all the factors necessary to keep V3 in a sensitized state under inflammatory conditions, allowing acid to act as a strong agonist for V3 activation.

This chapter demonstrates that another inflammatory mediator, protons, enhances

V3 activity, further supporting the hypothesis that V3 has function during inflammatory conditions. In particular, protons play a central role in the transduction of pain during inflammatory conditions, suggesting that V3 could have an important role in the transduction of inflammatory pain, which is especially relevant to the rapidly growing population of elderly people in the United States, where conditions such as arthritis, cancer and infections of the skin are common. Since the data presented in this chapter, as well as Chapter 2 & 3 are consistent with V3 having an important role in the transduction of painful stimuli under inflammatory conditions, the possibility of V3 as potential target for the development of therapeutic drugs for pain management should be considered.

120

o 26 C pH5.6 pH5.6 pH5.6 pH5.6 pH5.6 4 ) -4 3 pCDNA 3 (x10 2 0 TRPV1 TRPV1D647N TRPV3 TRPV3D641N

F-F 1 0 012345 012345 012345 012345 012345 Time (m) Time (m) Time (m) Time (m) Time (m)

2+ Fig. 4.1. Activation of TRPV3 by acid. Acid-induced changes in [Ca ]i in HEK293 cells expressing murine TRPV1, TRPV1D647N, TRPV3, TRPV3D641N, and pcDNA3 (as negative control). Cells were seeded in a 96-well plates and transfected with the corresponding cDNA. After 24 hrs, cells were loaded with fluo4-AM, washed and 2+ [Ca ]i monitored in a microplate reader with excitation and emission wavelengths at 494, and 520 nm, respectively. An acidic solution was added as indicated to make the final pH to 5.6. Individual traces represent the average of duplicate experiments.

121

A pH 5.6pH 5.6 pH 5.6 pH 5.6 pH 5.6

) 2 D641N -5 TRPV3

TRPV3 1 pcDNA3 RFU (x 10 (x RFU 26 oC32 oC 37 oC 42 oC 45 oC 0 024602 46024602 460246 Time (min)

B 16

) -4 12

8 RFU (x10 RFU pcDNA3 4 TRPV3 Before acid stimulation TRPV3 4 min after acid sti. D641N 0 25 30 35 40 45 Temp (oC)

Fig. 4.2. Activation of wild type TRPV3 by acid at elevated temperatures. A, Temperature dependence of acid-stimulated activities of TRPV3 and TRPV3D641N. Flou4-loaded cells were monitored at different temperatures as indicated. Traces in the upper graphs show the average of duplicate experiments at 26, 32, 37, 42, and 45oC. B, Summary of (A), showing the fluorescence before (open symbols and dashed lines) and 4 min after (closed symbols and solid lines) the acid stimulation at different temperatures in control (circles), TRPV3 (triangles), and TRPV3D641N (squares) cells. Shown are averages ± ranges of duplicated samples.

122

A o pH5.6 [Ca2+] = 1.8 mM pH5.6 Ca2+-free 32 C e 6 6

) -4 4 pcDNA3 4 TRPV3 (x 10 (x 0 2 2 TRPV3D641N

F-F

0 0

-2 -2 02468 02468 Time (min) Time (min)

B Mn2+ pH 5.6 pcDNA3

4 V3 2 x 10 RFU 1 min

V3D641N

Fig. 4.3. Acid activates TRPV3 on the plasma membrane. A, Extracellular Ca2+ dependence of the acid-induced fluorescence increases in TRPV3D641N cells. Experiments were performed in the presence (left) or absence (right) of 1.8 mM Ca2+ in 2+ the extracellular solution. B, Mn -quenching of fura2 signal in acid-stimulated V3D641N and V3, but not control cells at 37 oC. Fura2-loaded cells were monitored using excitation and emission wavelengths of 357 and 510 nm, respectively. RFU, relative fluorescence unit. (A) and (B) are the average of triplicate experiments.

123

AB 4 pH 5.6 1.0

3 0.8 )

-4 TRPV3D641N (No Cap) 2 0.6 (x 10 (x

0 Cap 0.4 1 TRPV3D641N F-F 100 µM 0.2 Capsaicin 0 0 pcDNA (no Cap)

Normalized response Normalized 1101001000 02468 [Capsaicin] (µM) Time (m)

Fig 4.4. Inhibition of acid-induced TRPV3D641N activity by capsaicin (Cap). A, Shown are sample traces for the inhibition of the acid response by 100 µM capsaicin using cells transfected with TRPV3D641N. Traces for TRPV3D641N and pcDNA3 transfected cells with no capsaicin added are shown as controls. B, Best-fit inhibition curve for the capsaicin induced inhibition of the acid response of TRPV3D641N transfected cells. Data points represent the average ± SD of triplicate samples.

124

A pH8.5 pH6.5 pH5.5 nA d d 8

TRPV3D641N 4 nA 4 c c 30 sec a b a mV -100 -50b 50 100

-4 +100 mV -100 mV -8 B 0.2 nA pH5.5 g TRPV3 0.2 nA

e 10 sec emV -100 -50 50 100 g +100 mV f -100 mV f -0.2

Fig. 4.5. Acid evoked TRPV3 currents. A&B, acid-evoked currents in V3D641N (A) and V3 (B) cells. After making the whole cell, voltage ramps of 100 ms from –100 to +100 mV were applied from the holding potential of 0 mV at 2Hz. Solutions with different pH were applied through perfusion. The base solution is pH 7.4. Left shows the currents at –100 (open) and +100 mV (filled). Right shows I-V curves for the time points indicated by the letters. Note, I-V traces b & f are for the endogenous ASIC currents. This experiment was performed by Hong-Zhen Hu.

125

D641N A TRPV3 TRPV3 pcDNA3 pH 5.6 pH 5.6 pH 5.6 30 2APB 2APB 2APB

2APB (µM) ) 25 2APB (µM)

-4 166 2APB ( µM) 20 166 55 55 18 166 18

(x 1 0 55 15 6 0 18 2 10 6 F-F 6 2 0 5 0 2 0 0 02460246024 6 Time (min)

D641N B TRPV3 TRPV3 100 100

80 80

60 60 pH5.6 pH7.4 pH5.6 pH7.4 1.13 µM 23.9 µM 3.83 µM 57.7 µM 40 40

20 20 Norm alized alized respo respo nse nse Norm alize d resp onse 0 0 876543 876543 -Log [2APB] (M) -Log [2APB] (M)

2+ Fig. 4.6. Acid potentiates the 2APB response of TRPV3. A, 2APB-induced [Ca ]i rise in cells transfected with pcDNA3, V3 and V3D641N and its potentiation by subsequent addition of acid in the presence of the same concentrations of 2APB. B, dose response curves for 2APB obtained at pH 7.4 and pH 5.6. Data points represent the average ± SD of triplicate samples and EC50 values were obtained using the Hill equation.

126

TRPV3 A 2APB at pH 5.5 20

16 ) 166 µM

-4 55.3 µM 12 18.4 µM 6.14 µM (X 10 (X

o 8 2.05 µM 683 nM

F - F 4 228 nM 0 µM 0 012345 Time (min)

B

EC ~ 3 µM 50

Fig 4.7. Simultaneous addition of acid and 2APB to TRPV3 transfected cells. A, 2+ (Acid + 2APB)-induced [Ca ]i rise in cells transfected with V3. B, dose response curves for 2APB obtained at pH 7.4 and pH 5.6. Data points represent the average of triplicate samples.

127

A 6 µM 2APB acid 16

5.54

) 12

-4 5.63 5.82 5.96

(X 10(X 8

o 6.08 7.4 F - F 4

0 0 1234 5 Time (min)

B 18 16 14 ) V3D641N -4 12

10 (X 10 8 o 6

F - F 4 2 0 4 3 2 6 8 0 .5 .6 .8 .9 .0 .4 5 5 5 5 6 7 pH

Fig. 4.8. The effect of different pHs on the potentiation of the 2APB response of TRPV3. A, Acid at various pHs was added at 20 seconds followed by 6 µM 2APB in cells transfected with V3. Similar results were obtained for V3D641N transfected cells (not shown). B, Summary of the results obtained for V3D641N transfected cells 45 seconds after addition of 2APB. . Data points represent the average ± range of duplicate samples. For A & B, the pH profile for the potentiation by 2APB was similar for V3 and V3D641N.

128

Fig. 4.9. pH dependence of the 2APB-induced currents of TRPV3 transfected cells. Solutions were changed from 7.4 to pH 8.0 (A), 7.0 (B), 6.5 (C), and 5.5 (D) as indicated with the concentration of 2APB (100 µM) held constant. For B-D, the cells were stimulated by 2APB for the first time. For A, the cell was previously stimulated with 2APB several time in order to obtain a maximal activation level. For A-D, left is the current at +100 mV (closed circles) and –100 mV (open circles) and right is the I-V curves with letters a-d corresponding to the time points labeled on the left. In D, trace d represents the endogenous ASIC current. This experiment was performed by Hong-Zhen Hu.

129

2+ 2+ A Ca Ca pH 5.0 pH 7.4 pH 5.0 3.0

2.5

2.0

1.5

(340/380) Ratio 1.0

0.5 0 102030405060

Time (min)

B Ca2+ 1.3 pH 5.0

1.1 Ca2+ free 0.9

0.7 Ratio (340/380) Ratio 0.5 0 1020304050 Time (min)

2+ Fig. 4.10. Sensitization of the acid-induced rise of [Ca ]i in TRPV3D641N transfected cells. A, Fura-2 loaded HEK293 cells that were transfected with TRPV3D641N display a 2+ time-dependent increase in [Ca ]i upon stimulation with acid. Removal of extracellular 2+ 2+ Ca results in a decrease in the [Ca ]i . A second application of acid in the presence of 2+ 2+ 1.8 mM Ca results in more pronounced rise in [Ca ]i than occurred upon the first stimulation.. B, pcDNA3 transfected cells demonstrate a much smaller response to acid that corresponds to the acid induced rise in A with the nominally Ca2+-free solution.. Fura2-loaded cells were monitored using excitation and emission wavelengths of 357 and 510 nm, respectively.

130

Fig. 4.11. Acid-induced slow activation/sensitization of TRPV3 currents. A, TRPV3- transfected HEK293 cells were voltage-clamped in the whole-cell mode and recorded using voltage ramps. Acidic solutions and 2APB (100 µM) were perfused onto the cells as indicated. Left, membrane current at –100 mV (open circles) and +100 mV (filled circles) for a representative cell. The dashed line indicates zero current. Right, I-V curves obtained by the voltage ramps at the indicated time points in the left. Note, I-V traces a & f contain the contribution of an endogenous ASIC currents known to be present in HEK293 cells. B, comparison of activation kinetics for the acid-evoked response at +100 mV. Solid lines represent fit using the formula: . -t/τ y It = Imax (1-e ) where Imax is the maximal current, τ is the activation time constant, and y is a coefficient for cooperativity. Performed by Hong-Zhen Hu.

131

Fig 4.12. Acid-induced slow activation/sensitization of V3D641N currents. A, TRPV3D641N-transfected HEK293 cells were voltage-clamped in whole-cell mode and recorded using voltage ramps. B, comparison of activation kinetics of TRPV3D641N current at +100 mV activated at different pH. Shown are first acid stimulation for pH 6.5, 6.0 and 5.5 of different cells as well as the 2nd stimulation for pH 5.5. I/V curves at the end of the stimulation are shown at the top. C, activation time constants of currents at +100 mV for TRPV3 (see Fig. 4.14) and TRPV3D641N by low pH solutions. * p < 0.01 different from TRPV3. n = 3-5. Performed by Hong-Zhen Hu.

132

A pH 5.6 5 pH 5.6 4 Buffer only 4

)

-4 3 3 1.5 µM CALC 2 (x1 0 0 2 15 µM U73122

1 20 µM ETYA F-F 10 µM H89 1 Buffer only 0 0 0123456 0123456 Time (m) Time (m) B

Acid-evoked response of TRPV3D641N 1.5 1

0.5 0 e 9 2 9 n 8 M M M M C A A li m n µ µ µ µ n 2 0 0 Y a o H ti 1 6 3 o c N 0 0 1 1 3 D 3 k i M 3 1 . s 7 , ET rs x µ 0 o U M 2 o ro

Normalized response h p µ -1 M F i 0 l M µ p 1 µ 3 0 M GF109203X Ca DL 2 M 5 M µ µ M 1 3 0 µ M 6 5 µ . 0 1 5

Fig 4.13. Identification of signaling events that modulate the acid-evoked response of TRPV3D641N. Drugs were added 30 min before the application of acid. A, shows example traces for H89 (PKA blocker), Calphostin C (CALC, PKC blocker), U73122 (PLC blocker) and ETYA (non-metabolizable AA analog). B, shows the summary for all drugs tested. Data are averages ± SE for 4-5 samples or averages ± range for duplicated samples for fluorescence increases normalized to the control cells at 100 sec after the addition of protons.

133

A B pH 5.6 166 µM 2APB 10 22

V3D641N 8 18 ) V3DNTA ) -4 -4 V3DNTD 6 pc DNA3 14 (X 10 (X 10 V3D641N o 4 o 10 V3 DNTA V3 DNTD F - F F - F pc DNA 3 2 6

0 2 0 0 12345 012345 Time (min) Time (min)

CD

378 Fig. 4.14. Mutational analysis of Thr , a PKA consensus site of TRPV3 . A, Acid (pH 5.6) activation of V3D641N, V3DNTA, V3DNTD and pcDNA3 transfected cells. 378 V3DNTA and V3DNTD were created by mutating Thr to alanine and aspartate, respectively, in the V3D641N background. Acid was added at 20 seconds. Traces are the average of duplicate experiments. B, Same as in A except 166 µM 2APB was used to activate the cells instead of acid. C, Light blue bars are the summary for data points for the V3DNTA mutant 3 minutes after stimulation with 166 µM 2APB, pH 5.6 or (pH 5.6 + 10 µM H89). The gray bar is a control to show that V3D641N is inhibited by H89 to a similar extent as V3DNTA. D, Same as C only V3DNTD was used instead of V3DNTA. For C & D, data are averages ± ranges for duplicated samples for fluorescence increases normalized to the level of activation of V3D641N by either 166 µM 2APB (left most bar) or pH 5.6 (all but the left bar).

134

A pH 5.6 5

) 4 V1 -4 V2 3 V3 2 V3DN (X 10 (X

o V4 1 pcDNA3

F - F 0

-1 01234 Time (min)

B

Fig 4.15. Acid-induced activation of other TRP family members. A, Acid (pH = 5.6) induced activation of cells transfected with individual members of the thermoTRPVs or pcDNA3 (control). Traces are the average of triplicate experiments. B, shows the summary for all TRP channels tested. Data are averages ± SD for 3 samples from the experiments in (A) 120 sec after the addition of acid. The pcDNA3 data were subtracted out. V1DN and V3DN represent V1D647N and V3D641N respectively.

135

pH 5.6 2APB 16

14 1000 µM ) 12

-4 333 µM

10 111µM

37.0µM 8 (X 10 12.3 µM o 4.11 µM 6 0 µM 4 F - F

2

0

0123456

Time (min)

2+ Fig. 4.16. Acid potentiates the 2APB response of TRPV2. 2APB-induced [Ca ]i rise in cells transfected with V2 and its potentiation by subsequent addition of acid (pH = 5.6) in the presence of the same concentrations of 2APB. Each trace represents the average of two trials with the results from pcDNA3 transfected cells subtracted out.

136

Fig. 4.17. Potentially important amino acid sequences for the regulation of TRPV3. The blue box (upper, upper) has asterisks located above the acidic amino acid residues that are located between the 5th and 6th transmembrane domain of V3 and V1. These amino acids represent the most likely candidates for regulation by protons. The red box (upper) has the amino acids of V1 that are known to be important for the activation of V1 by protons bolded and in large font. In addition, the equivalent amino acid positions of V3, V2, V4, V5 & V6 are also bolded and in large font. This box illustrates that neither of the acidic residues at the equivalent positions of V1 are necessary for the acid-induced activation for the entire TRPV family, as V3 and V4 are both modulated by acid and only have an acidic amino acid at one of the equivalent positions as V1. In addition, V2 does not have an acidic amino acid at either of the positions even though it is modulated by protons. Therefore, it is likely that the mechanism for each thermoTRPV involves acidic residues at different positions. For the blue & red box, blue sequences are part of the 5th or 6th transmembrane domain and red sequences are part of the P-loop. The middle is an illustration of one subunit of a V3 channel. Ovals with an “A” represent ankyrin repeats and the oval with the “T” represents the TRP-box. The green box (lower, left), black box (middle, left), and brown box (lower, right) show critical serines or threonines (bold and large font) that are phosphorylated by PKA on V1. In addition, the equivalent positions for V3 are bolded and in large font. Red amino acids outside the boxes are critical V1 amino acids and blue amino acids outside the boxes are V3 amino acids that are in the equivalent position as the V1 amino acids. The significance of each of these positions is described in the text.

137

** * * * * * * V1: 596 VTLIEDGKNNSLPVES---PPHKCRGSACRPGNSYNSLYSTCLELFKFTIGMGDLEFTENYDFKA 658 *** * * * * V3: 604 VTLLNPCTNMKVCDED----QSNCTVPTYPACRDSETFSAFLLDLFKLTIGMGDLEMLSSAKYPV 667 Glu600 Asn608 Glu648 Glu655

V1: 596 VTLIEDGKNNSLPVES---PPHKCRGSACRPGNSYNSLYSTCLELFKFTIGMGDLEFTENYDFKA 657

V3: 604 VTLLNPCTNMKVCDED----QSNCTVPTYPACRDSETFSAFLLDLFKLTIGMGDLEMLSSAKYPV 664

V2: 550 VSLSREARSPKAPENSNTTVTEKPTLGQEEEPVPYGGILDASLELFKFTIGMGELAFQEQLRFRG 620 V4: 631 ASLIEKCSKDKKDCSS------YGSFSDAVLELFKLTIGLGDLNIQQNSTYPI 697 V5: 508 YIIFQTEDPDNLGEFS------DYPTAMFSTFELFLTIIDGPANYRVDLPF 558 V6: 514 YIIFQTEDPDELGHFY------DYPMALFSTFELFLTIIDGPANYDVDLPF 564

132645

T T T T T T A A N A C

Ser116 Thr370 Ser502 V1: 112 YDRRSIFDAVAQS 124 V1: 366 SRKFTEWAY 374 V1: 498 QRRPS-LKS 505

V3: 117 RLKKRIFAAVSEG 129 V3: 374 SRKFTDWAY 382 V3: 507 LLRPSDLQS 515

Arg121 Thr378 Ser511 Figure 4.17

138

CHAPTER 5

TRPV1 AND TRPV3 FORM FUNCTIONAL HETEROTETRAMERS

5.1 Introduction

The question as to weather or not V3 and V1 form functional heterotetramers has been debated since V3 was first characterized (Smith et al., 2002). Several observations have contributed to the notion that V1/V3 heterotetramers may make up at least some of the native capsaicin receptors (CRs) of peripheral C-fibres. First, V3 and V1 are both located on chromosome 7 (mouse), only 7.5 kb apart and in the same transcriptional orientation as V1, partially explaining the coexpression of V1 and V3 in a subpopulation of the DRG of peripheral C-fibres (Smith et al., 2002; Xu et al., 2002; Fig. 2.4). Second,

V3 and V1 have a very high degree of sequence homology in the transmembrane regions, an observation that should be required for heterotetramerization. Third, several TRP channels have been shown to form heterotetramers, with ectopically expressed heterotetramers more closely recapitulating the properties of currents exhibited in their native tissues (Montell, 2005). Groups of TRPs that have been shown to form heterotetramers include mammalian TRPC1, TRPC4, and TRPC5 or TRPC3, TRPC6 and

TRPC7, as well as TRPV5 and TRPV6 (Hellwig et al., 2005). In fact, the founding

139 members of the TRP family, Drosophila TRP and TRPL form heterotetramers that make up the functional channel in photoreceptor cells (Montell, 2005). TRPL has also been shown to form heterotetramers with TRPγ. Finally, DRG neurons appear to be more sensitive to agonists than recombinant V1, and coexpression of V3 with V1 in HEK293 cells results not only in channels with increased capsaicin responsiveness, but in the ability to pull-down V3 proteins using V1 antibodies and vice-versa (Szallasi, 2002,

Smith et al., 2002).

Although increased capsaicin responsiveness and the ability to pull-down V1 proteins with V3 antibodies (and vice-versa) support the formation of V1/V3 heterotetramers, other data support the notion that V1 and V3 form separate homotetrameric channels. Most of the biophysical properties of native capsaicin receptors, such as single-channel conductance, current-voltage relationship, and ion selectivity are accounted for by the heterologous expression of V1. It should be noted that the biophysical/electrophysiological properties of V3 channels do not differ dramatically from V1 with regards to these properties, and V1/V3 heterotetramers may form channels with similar biophysical properties as V1 homotetramer. Fluorescence resonance energy transfer (FRET) experiments using V1 that has a yellow variant of green fluorescent protein attached to the C-terminus (V1-YFP) and V3-CFP (cyanin variant of green fluorescent protein) expressed in HEK293 cells demonstrated very low levels of FRET efficiencies between the CFP and YFP portion of the proteins (Hellwig et al., 2005). In contrast, coexpression of V1-CFP and V1-YFP or V3-CFP and V3YFP resulted in a high FRET efficiency, suggesting that homomultimers were highly favored

140 over heteromultimers. It should be noted that the tetramerization of TRP channels is determined by cooperative interactions between transmembrane segments as well as between N- and C-terminal regions of adjacent subunits that make up TRP tetrameric channels, and tagging these proteins may alter the selectivity of partners (Hellwig et al.,

2005). In addition, tagging the C-terminus of V3 will remove any specificity toward

PDZ domain containing proteins that V3 may have affinity to, and this could also influence binding partners by altering localization.

Experiments in Chapter 2, 3 and 4 have demonstrated that V3 responds to many stimuli. In addition, Chapter 6 will discuss additional stimuli that modulate V3 channels.

Although some of these stimuli are similar to those that activate V1, such as acid, heat and 2APB, the efficacy of these stimuli toward V1 and V3 differ. Additionally, some chemical activators of V3 do not modulate V1 and vice versa. Phosphorylation by PKA and PKC also differs between these proteins. In fact, although V3 shares a high level of sequence homology with V1 in the transmembrane segments, the intracellularly localized

N- and C-terminal domain of V1 and V3 differ dramatically in amino acid sequence, suggesting that these channels could form heterotetramers with similar biophysical properties upon activation, but differ in the way in which the resultant channels are regulated by chemicals, phosphorylation events and various other signaling pathways within cells (Smith et al., 2002; Xu et al., 2002; Peier et al., 2002). Considering the medical and/or economical impact that V1 related disorders present to society, determining whether or not V1 and V3 form heterotetramers should be considered an

141 important endeavor. In this chapter, we used a pharmacological as well as a molecular biological approach to demonstrate strong support that V1 and V3 form functional heterotetramers.

5.2 Material and Methods

5.2.1 DNA constructs, cell culture, and transfections

All compounds and reagents were purchased from Fisher or Sigma unless otherwise stated. 2APB was from Tocris or Cayman Chemical Company. All receptor cDNAs were cloned into pcDNA3. The cloning strategy for V1-V3 concatamer has been published (Wang et al., 2004).

5.2.2 For all experimental procedures, see Section 2.2.

5.3 Results

5.3.1 Coexpression of TRPV1 and TRPV3 increases Ca2+-influx and decreases Ca2+- release in response to capsaicin

2+ Coexpression of V1 and V3 in HEK293 cells results in a larger rise of [Ca ]i in response to capsaicin than when V1 is expressed alone (Smith et al., 2002). Since V3 is not activated by capsaicin, it has been hypothesized that this result is due to

142 heterotetramerization of V1 and V3. In addition, application of capsaicin to V1 expressing cells bathed in a Ca2+-free buffer has revealed that capsaicin can activate V1 located on the ER and/or other intracellular Ca2+ storage organelles (Fig. 5.1). Although

2APB can activate V3 located on intracellular Ca2+ storage organelles of HEK293 cells

(Section 2.3.3), localization studies have shown that V3 is more predominantly located to the plasma membrane than V1 when expressed in these cells (Hellwig et al., 2005). We therefore cotransfected different ratios of V1 and V3 [V1:V3 (1:0), (1:4), (1:10); with V1 cDNA constant] into HEK293 cells in order to determine if the coexpression of V3 could affect the ratio of Ca2+-influx to Ca2+-release in these cells (Fig. 5.1 A - C). In the presence of extracellular Ca2+, application of 30 µM capsaicin resulted in a larger rise of

2+ [Ca ]i in cells cotransfected with V1 and V3 than in cells transfected with V1 alone (Fig,

5.1A & C; also see Fig. 5.4B). For Figure 5.1A, only traces for V1:V3 (1:0) and (1:4) are shown. These data are in agreement with previously reported results (Smith et al., 2002).

In their studies, it was suggested that the increased sensitivity of V1 to capsaicin when coexpressed with V3 was due to either heterotramerization of V1 and V3 or a competition of inhibitory factors between V1 and V3. Interestingly, performing the same experiments with a nominally Ca2+-free solution resulted in a reduction of Ca2+-release as the ratio of V3 to V1 increases (Fig. 5.1A & C). To confirm these results, we cotransfected V1 and V3 at three different ratios (1:0, 1:1, 1:3) and sequentially added 10

µM capsaicin to the cells in a nominally Ca2+-free solution followed by the same concentration of capsaicin with 1.8 mM extracellular Ca2+ (Fig. 5.1C). This allowed us to examine the effect on both Ca2+-release and Ca2+-influx on the same sample of cells.

143 As with the experiments in Figure 5.1A & B, as the ratio of V3 to V1 increases, a reduction in Ca2+-release is accompanied by an increase in Ca2+-influx in response to capsaicin. These results suggest that V3 causes the fraction of V1 localized on the ER to relocalize to the plasma membrane. This could occur by two different mechanisms; either V1 is transported to the plasma membrane by interacting with the cellular components utilized by V3 for localization to the plasma membrane, or V1 forms heterotetramers with V3, and the fraction of V1 on the ER is transported to the plasma membrane via direct interaction with V3. Although this result alone cannot distinguish between the two scenarios, it does suggest that V3 does not increase the activity of V1 by titrating inhibitory cellular components away from V1, since if this were the case, both

Ca2+-release and Ca2+-influx would be expected to increase.

Since the above results are consistent with heterotetramerization of V1 with V3, we next forced heterotetramerization of V1 and V3 by constructing a V1-V3 concatamer linked by a small amino acid linker (CQQQQFCSRAQASNSAVDD) between the C- terminus of V1 and the N-terminus of V3. This forces heterotetramerization by dramatically increasing the “effective” concentration of V1 and V3 between the two linked subunits. Although the observation that concatamers are functional does not guarantee that heterotetramerization between V1 and V3 occurs in vivo, it will allow a comparison of properties observed when V1 and V3 are coexpressed to those of the V1-

V3 concatamer (forced heterotetramer). Performing the equivalent experiments as in

Figure 5.1A with the V1-V3 concatamer in either the presence or absence of extracellular

144 Ca2+ resulted in levels of Ca2+-release and Ca2+-influx that were more similar to those of

V1:V3 (1:3) than V1:V3 (1:0) transfected cells, suggesting that coexpression of V1 and

V3 results in heterotetramers (compare Fig. 5.1A with Fig. 5.1D).

2+ 5.3.2 The capsaicin-induced rise of [Ca ]i is similar for TRPV1-TRPV3 transfected or TRPV1/TRPV3 cotransfected cells

We next decided to compare the response of cells transfected with either V1, both

V1 and V3 (1:3), or the V1-V3 concatamer to 30 µM capsaicin. Interestingly, cells transfected with V1-V3 or cotransfected with V1 and V3 displayed a larger yet slower

2+ rise in [Ca ]i in response to 30 µM capsaicin than V1 transfected cells (Fig. 5.2).

Typically in our hands, V1 rapidly reaches the maximum response to a particular capsaicin concentration following application. V3 on the other hand is usually slowly activated to stimuli, only reaching maximal response over an extended period of time, with the response to high concentrations of 2APB being the exception. The slow activation kinetics of V3 is due to a time-dependent sensitization process (or time- dependent relief of inhibition) (Section 2.4, 3.4, 4.4). These results demonstrate not only that the concatamer of V1-V3 is functional, but that expression of V1-V3 or coexpression of V1 and V3 both result in capsaicin-sensitive channels with properties that are specific to both V3 (time-dependent sensitization) and V1 (capsaicin responsiveness), a phenomenon that is best explained by heterotetramerization of V1 and V3.

145 5.3.3 Pharmacological comparison of cells transfected with TRPV1, TRPV3, or TRPV1/TRPV3

Both V1 and V3 are activated by 2APB, with V3 (57.7 µM) having a lower EC50 for 2APB than V1 (114 µM). In addition, the 2APB-induced activation of V3 is inhibited by high concentrations of capsaicin (Fig. 2.11). We therefore compared the response of cells transfected with V1, V1 and V3 at a 1 to 3 ratio [V1:V3 (1:3)], V3, and pcDNA3 to either 30 µM capsaicin followed by 166 µM 2APB or 166 µM 2APB only (Fig. 5.3).

Point 1 of Figure 5.3 confirms that coexpression of V1 and V3 results in a larger but

2+ slower rise of [Ca ]i in response to 30 µM capsaicin than V1 expressing cells. As mentioned previously, Smith et al. (2002) suggested that this phenomenon is either the result of heterotramerization of V1 and V3 or an increased sensitivity of V1 to capsaicin due to a competition of inhibitory factors between V1 and V3. Point 3 of Figure 5.3 is in conflict with the latter suggestion, as the coexpression of V1/V3 results in a smaller rise

2+ of [Ca ]i in response to 166 µM 2APB than when V3 is expressed alone. Interestingly, the response of V1/V3 transfected cells is intermediate between that observed when V1

(smaller than V1/V3) or V3 (larger than V1/V3) are expressed alone. If separate homotetramers of V1 and V3 were competing for inhibitory factors, not only should the response of V1/V3 to capsaicin be larger than that of V1 (observed), but the response of

V1/V3 to 166 µM 2APB should be larger than that observed for V3 alone (not observed).

In addition, the 2APB response of V3 has been shown to be slightly inhibited by 30 µM capsaicin, a phenomenon that is also shown in point four (Fig. 2.11; Fig. 5.3, point 4).

Point four of Figure 5.3 also demonstrates that the 2APB-only response of V1 is not as

146 pronounced as that to 30 µM capsaicin, a result that is expected since 30 µM capsaicin maximally activates V1, whereas 166 µM 2APB is a submaximal dose of 2APB with regards to V1 activation (Fig. 2.10). Interestingly, point 2 of Figure 5.3 demonstrates that the response of cells coexpressing V1 and V3 to 166 µM 2APB is neither inhibited (a property of V3) nor potentiated by 30 µM capsaicin (a property of V1), suggesting that coexpression of V1 and V3 results in channels with properties intermediate between V1 and V3 with regards to the effect of 30 µM capsaicin on the response to 166 µM 2APB.

5.3.4 Pharmacological comparison of cells transfected with TRPV1, TRPV1-TRPV3 or TRPV1/TRPV3

We next compared the response to either 3.7 µM capsaicin followed by 166 µM

2APB or 166 µM 2APB-only of cells transfected with either V1, the V1-V3 concatamer, or V1 and V3 at the following ratios (1:1), (1:3), or (1:9) (Fig. 5.4). Point 1 of Figure 5.4

2+ demonstrates that the rise of [Ca ]i in response to 3.7 µM capsaicin increases with increasingly larger ratios of V3 to V1 (Fig. 5.4, point 1; Fig. 5.4, lower). In addition, the response of the V1-V3 concatamer is intermediate between that of the response by V1/V3

(1:1) and V1/V3 (1:3), but is closer to that of V1/V3 (1:3). Furthermore, the capsaicin-

2+ induced rise of [Ca ]i (a property of V1) for the V1-V3 concatamer displays a more sustained time-dependent increase (a property of V3) than V1. Taken together, points 2,

3 and 4 of Figure 5.4 also demonstrate that the V1-V3 concatamer has some properties that resemble V1 and some that resemble V3. V1-V3 is similar to V1 in that it is fully activated by 3.7 µM capsaicin, as demonstrated by the inability of 2APB to potentiate the capsaicin response (Fig. 5.4, point 2). Unlike V1, but like V3 (see Fig. 5.3, point 3), V1- 147 V3 is fully activated by 166 µM 2APB (Fig. 5.4, point 3), as the 2APB only response is just as large as the response to both capsaicin and 2APB (Fig. 5.4, point 4). Point 3 of

Figure 5.4 also demonstrates that as the ratio of V3 to V1 increases so does the magnitude of the response to 2APB only. This is consistent with a population of heterotetrameric channels with a progressively larger portion consisting of V3 subunits.

This point is illustrated in Figure 5.5. The subunit composition of heterotetramers will be based on two factors: 1) The relative concentrations of “free” V1 compared to V3. 2)

The relative affinities of V1 for other V1 subunits compared to V3 and vice versa. The example in Figure 5.5A assumes both equal affinity between V1 and V3 and equal concentrations of V1 and V3 subunits, a situation that probably does not exist.

Nonetheless, assuming that V3 and V1 have slightly higher affinities for their own respective protein subunits (Fig. 5.5B), and that much higher concentrations of V3 are present as the ratio of V3 to V1 is increased in the transfections (Fig. 5.5C), it is safe to assume that as the ratio of V3 to V1 increases, a higher number of channels will either contain V3 only, or contain only one V1 subunit (Fig. 5.4B & C). This accounts for the increase in 2APB response (Fig 5.4, points 2 & 3) as the ratio of V3 to V1 increases, but not the increased capsaicin response (Fig. 5.4, point 1). The increase in capsaicin response can continue to rise as the V3 to V1 ratio increases if activation of heterotetrameric channels containing V1 only requires binding of capsaicin to one V1 subunit in heterotetrameric complexes. In this scenario, increasing the total V3 subunit concentration while maintaining the same total V1 subunit concentration will result in less V1 homotetramers and V13V31 heterotetramers, as the higher concentration of V3

148 will allow the V3 subunits to overcome the higher affinity of V1 for other V1 subunits, ultimately resulting in more channels that have at least one V1 subunit (compare Fig.

5.5B to 5.5C). In addition, a comparison of Figure 5.4 and 5.3 demonstrates that capsaicin inhibits the 2APB response to a greater extent, as either the concentration of capsaicin is larger (Fig. 5.3), or the proportion of V3 subunits increases (Fig. 5.4).

5.3.5 Coexpression of TRPV3 with TRPV1 increases the sensitivity of TRPV1 to acid

We next decided to compare the acid response (pH 5.6) of cells transfected with

V1, to that of cells cotransfected with V1 and V3 at a 1:3 ratio. To accomplish this, V1,

2+ V3, V1:V3 (1:3), or pcDNA3 were transfected in HEK293 cells and increases of [Ca ]i in response to acid application were monitored via fluo4 fluorescence at 32 oC, a temperature in which V3 is not activated by acid (Fig. 5.6A). As expected, the acid response of V3 transfected cells is equivalent to those of the pcDNA3 controls, confirming that V3 is not activated by acid at this temperature. Conversely, V1 transfected cells were activated by acid at this temperature. Although it appears as if the acid response of V1 displays a time-dependent desensitization, the time-dependent decrease is roughly equivalent to that of the endogenous response. Like with capsaicin,

2+ the acid-induced rise of [Ca ]i in cells cotransfected with V1/V3 (1:3) is larger than that of cells transfected with V1 alone. In addition, cells transfected with the V1-V3

2+ concatamer also displayed a larger rise of [Ca ]i than those of V1 transfected cells (Fig.

2+ 5.6B). Although the capsaicin-induced rise of [Ca ]i for V1-V3 as well as V1/V3

149 cotransfected cells displayed an extended time-dependent rise, the acid response for these cells appear to display an extended time-dependent decline. Like with the V1 acid response, this is explained by the time-dependent decline of the endogenous response. In fact, since the acid response of V1-V3 and V1/V3 (1:3) declines at a slower rate than the endogenous response, the V1-V3 or V1/V3 (1:3) portion of the acid response displays an extended time-dependent increase similar to the capsaicin response. Also of relevance, the increased acid response of cells transfected with V1-V3 compared to cells transfected with V1 (Fig. 5.6B) was similar in magnitude and shape as that seen when comparing cells cotransfected with V1/V3 (1:3) to those of V1 transfected cells (Fig. 5.6A), further showing support for the formation of V1/V3 heterotetramers.

5.3.6 The effect of 2APB on the acid response of TRPV1/TRPV3 heterotetramers

We next compared the affect of 15 µM 2APB on the acid response of V1, V3, V1-

V3 or pcDNA3 transfected cells (Fig. 5.6C). Like with the capsaicin response, 2APB did not further activate the V1 response, a result that is expected since lowering the pH below

6.0 has been shown to fully activate V1 at room temperature (Tominaga et al., 1998;

Tominaga and Tominaiga, 2005). As expected, V3 was not activated by acid alone, but was strongly activated by the combination of acid and 2APB. Interestingly, the acid response of V1-V3 was potentiated by 2APB, indicating that the V1-V3 concatamer is not fully activated by acid. The potentiation of the acid response by 2APB in V1-V3 transfected cells differed from that of V3 transfected cells in both the overall magnitude

150 (V1-V3 smaller) and rate (V1-V3 slower) of activation. The ability of 2APB to further potentiate the acid response of V1-V3 transfected cells is different then that observed when capsaicin is used as the initial activator (compare Fig. 5.4 to Fig. 5.6C), where no further activation of V1-V3 was observed when 166 µM 2APB (an even higher concentration than used in this experiment) application followed treatment with capsaicin. The reason underlying this difference is not clear, but could be a result of the difference in the mechanism in which capsaicin and acid activates V1 channels. As a generalization, the capsaicin dependent activation mechanism depends on the interaction of capsaicin with a vanilloid-sensitive binding site within the transmembrane portion of

V1 (Tominaga and Tominaga, 2005). In contrast, the activation mechanism used by protons involves protonation of acidic residues on the P-loop of V1 channels (Tominaga and Tominaga, 2005). The difference in the mechanism of activation is apparent when considering avian V1 channels. The avian ortholog is insensitive to capsaicin while still being acid-sensitive (Jordt and Julius, 2002). It is possible that capsaicin binding provides enough binding energy so that binding of one or two V1 subunits results in activation of the entire channel, whereas protonation of acidic residues on V1 only provides enough binding energy for activation when all V1 channels are protonated, with only partial activation taking place when only two V1 subunits are protonated at key acidic residues. Of course, this is only a speculative mechanism, as the data provided in this chapter did not address the activation mechanism.

151 5.3.6 Coexpression of TRPV3D641N with TRPV1 results in capsaicin-sensitive channels with a reduced sensitivity to RR

So far, the experiments performed in this chapter have focused on the analysis of pharmacological differences between the response of cells transfected with V1, V3 or various combinations of V1 and V3 to capsaicin, 2APB or acid. Although the combined observations strongly support the hypothesis that V1 and V3 form functional heterotetrameric channels, the polymodal nature of V1 and V3 highlights the possibility that several unknown variables within the expression system could be altered by coexpression, and that the unknown variables could be responsible for the differences in the pharmacology of individual homotetrameric V1 and V3 channels observed in these experiments. In fact, the strongest evidence of V1/V3 heterotetramers should come from a demonstration of changes in the pore properties of capsaicin-sensitive channels due to coexpression of V3 with V1. This in itself is not easy because V1 and V3 are very similar with regards to cation selectivity, rectification, and other readily observable electrophysiological properties. In addition, even if a large enough difference in a particular pore property between V1 and V3 is identified, an enormous amount of work would be necessary using electrophysiological techniques to address this issue. We therefore reasoned that mutating Asp641 of V3 should result in V3 channels with a reduced affinity for RR, and that if V3 formed heterotetrameric channels with V1, coexpression of V1 and V3D641N should result in capsaicin-sensitive channels that are less sensitive to RR than wild type V1 (see Section 2.3.1 for an explanation of V3D641N).

Figure 2.2 shows that V3D641N has a reduced sensitivity to RR compared to wild type V3.

152 We therefore compared the ability of RR to block the capsaicin response of cells transfected with V1, V1/V3 (1:3) and V1/V3D641N (1:3). The dose response curve for each is shown in Figure 5.7B, with only the traces for V1/V3 (1:3) shown in Figure 5.7A.

As demonstrated, coexpression of V3D641N with V1 resulted in capsaicin-sensitive channels with a reduced sensitivity to RR. This result can only be obtained if the P-loop of at least one V3D641N subunit contributes to the formation of the pore of capsaicin- sensitive channels (V1 subunits). Figure 5.7C illustrates that the pore of TRPV channels is formed by the contribution of four P-loops from each subunit in the tetrameric channel complex. By demonstrating that pore properties of V3D641N are exhibited by capsaicin- sensitive channels, these data show unequivocally that V1 and V3 are capable of forming heterotetrameric channels.

5.4 Discussion

V1 plays a role in many physiological processes including temperature sensation, pain sensation and proper bladder function (Szallasi, 2001). In fact, hypersensitive V1 channels are at least partially responsible for pathophysiological conditions such as inflammatory and neurogenic pain, arthritis, migraine headaches, asthma, bladder overactivity, psoriasis and muscle cramps (Szallasi, 2002). In addition, the ubiquitous expression of V1 suggests that V1 plays a role in other physiological and cellular processes that have not been identified to date. In this chapter, we have shown that V3 is capable of forming functional heterotetramers with V1, and that the resulting channels

153 have altered responsiveness to capsaicin, acid and 2APB. In addition, the coexpression of V3 with V1 altered the cellular localization of V1, as determined by a reduction in the

2+ rise of [Ca ]i in response to capsaicin when comparing V1-only transfected cells with those of V1/V3 cotransfected cells bathed in a nominally Ca2+-free solution.

The ability of V1 and V3 to form heterotetramers could have a profound impact on the regulation of CRs in vivo and several lines of evidence highlight the possibility that V1/V3 heterotetramers make up at least part of native capsaicin receptors in DRG neurons. First, ectopically expressed V1 has a different sensitivity to vanilloid ligands, such as capsaicin and resiniferatoxin, than the native channels expressed in sensory neurons (Szallasi, 2002). Although differences in signaling factors and/or expression of regulatory components could underlie this phenomenon, especially considering the polymodality of V1 channels, heterotetramerization with V3 is also a possible explanation considering that coexpression with V3 results in changes in the pharmacology of capsaicin-sensitive channels in our studies and others. Second, V1 and

V3 are not only coexpressed in a subset of cultured rat DRG neurons (Fig. 2.4), but serial sections of human DRGs show that V1 and V3 are coexpressed in a subset of human

DRG as well (Smith et al, 2002). Finally, V3 expression is upregulated following neuronal injury and in patients experiencing breast cancer pain, two scenarios in which

V1 is proposed to have important function (Smith et al., 2002; Gopinath et al., 2005).

Interestingly, both cancer and neuronal injury are associated with inflammation, and

Chapter 3 and 4 has demonstrated that, like with V1, V3 is regulated by many factors associated with inflammation. Although highly variable in the intracellular N- and C-

154 terminal regions, the transmembrane and pore forming regions of V1 and V3 are highly conserved, consistent with the idea of heterotetramers forming channel/receptors with similar pore properties, yet differing with regard to regulation by kinases, phosphatases, calmodulin, endogenous and exogenous chemical activators, as well as other signaling components, which ultimately should result in capsaicin-sensitive channels that respond to a broader array of signals in the inflammatory environment.

Capsaicin-sensitive thermal receptor/channels are the target of many studies aimed at the discovery of therapeutic pain management strategies including drug development. In addition to pain management, the development of therapeutic drugs that target capsaicin-sensitive channel/receptors could lead to strategies for the treatment of the other pathophysiological conditions that V1 plays a role in. In fact, capsaicin therapies are used to treat neurogenic and chronic inflammatory pain, bladder overactivity, and muscle cramps (Szallasi, 2002). The development of drugs aimed at capsaicin sensitive channels/receptors often requires the identification of chemical compounds that either directly interacts with the channel or indirectly modulates the capsaicin receptors. These drugs are often identified by the observation that they modulate heterologously expressed V1 in high-throughput screening assays. This strategy is based on the assumption that the capsaicin receptors of DRG are composed of

V1 homotetramers. Since V1/V3 heterotetramers form in heterologous expression systems, it is probable they exist in vivo. Using V1-only expression systems for the identification of capsaicin receptor specific drugs could limit the number of drugs identified. Therefore, in addition to using V1, coexpression of V1 and V3 may be

155 necessary in the screening process for a comprehensive search of drugs capable of modulating capsaicin sensitive channels/receptors. With the enormously negative socioeconomic impact that the pathological disorders involving capsaicin receptors have on society, and with the observation that V3 is capable of forming heterotetrameric channels with V1, the role that V3 plays in the disorders involving capsaicin receptors should be elucidated.

156

Fig. 5.1. Coexpression of TRPV3 with TRPV1 alters capsaicin responsiveness. A-D, Cells were transfected with V1 and V3 or V3D641N at different ratios as indicated. The amount of V1 cDNA was kept constant and total amount of DNA was kept constant by adding pcDNA3. V1:V3 ratios are indicated on graphs. For all data, traces were obtained by averaging duplicates of the 2+ experiments. A, Representative traces for capsaicin-induced [Ca ]i increases in the absence or presence of 1.8 mM extracellular Ca2+ for cells transfected with different V1:V3 ratios. B, Ratio of the peak values obtained for Ca-influx in the presence of 1.8 mM extracellular Ca2+ over Ca2+- release in a nominally Ca2+-free solution for each of the V1:V3 ratios indicated. These data were taken from the experiments equivalent to those shown in (A). Shown are averages ± ranges of duplicated samples. C, Coexpression of V1 with V3D641N at the ratios indicated. V3D641N caused a decrease in capsaicin-induced Ca2+-release and an increase in the agoinst-induced Ca2+-influx as in (A). These experiments differ from (A) in that the agonist-induced activation in the absence or presence of 1.8 mM extracellular Ca2+ was performed on the same group of cells. D, Same experiment as in (A), but the V1-V3 concatamer was used in order to force V1 and V3 subunits into the same channel complex, thus mimicking heterotetramerization.

157

A 30 µM Capsaicin 4 V1/V3 (1:3)

) 3

-4

(X 10 2 o TRPV1

F - F 1

pcDNA3 0 0 20406080100120140 Time (sec)

B 30 µM Capsaicin 12

10 V1-V3

) 4 8 TRPV1

(x10- 6 0

F-F 4

2 pcDNA3 0 01 23 Time (m)

Fig. 5.2. Coexpression of TRPV3 with TRPV1 mimics the capsaicin responsiveness of cells expressing the TRPV1-TRPV3 concatamer. A, HEK293 cells were transfected with pcDNA3, TRPV1, or V1/V3 (1:3) at a 1 to 3 ratio. B, HEK293 cells were transfected with pcDNA3, TRPV1, or the V1-V3 concatamer. A&B, 30 µM capsaicin 2+ was added to cells seeded in 96-well plates and [Ca ]i was monitored with fluo-4 fluorescence. Traces represent the average of duplicate experiments.

158

166 µM 2APB 166 µM 2APB 30 µM Capsaicin

16 16 V3 { 16 ) -4 12 12 V1/V3 (1:3) 12 4

1

(X 10 8 8 2 3 8 o 4 4 V1 4 F - F pcDNA3 0 0 0 3 01 2 34 4 Time (min)

Fig. 5.3. Pharmacological comparison of cells transfected with V1, V3, or V1/V3 at a 1:3 ratio. HEK293 cells were transfected with pcDNA3, V1, V3 or V1/V3 (1:3) at a 1 to 3 ratio. 30 µM capsaicin followed by 166 µM 2APB or 166 µM 2APB-only were added 2+ to cells seeded in 96-well plates and [Ca ]i was monitored with flou-4 fluorescence. Traces represent the average of triplicate experiments. Bold circles with 1 (Point 1) highlight the response of each specific set of transfected cells to 30 µM capsaicin-only. Bold circles with 2 (Point 2) highlight the effect that 166 µM 2APB has on the response of each specific set of transfected cells to 30 µM capsaicin. Bold circles with 3 (Point 3) highlight the response of each specific set of transfected cells to 166 µM 2APB-only. Bold circles with 4 (Point 4) highlight the difference that 30 µM capsaicin has on the 2APB response for each specific set of transfected cells.

159

3.7 µM Capsaicin 166 µM 2APB 16 16 16 14 14 V1/V3 (1:9) 14 )

-4 12 12 12 4 10 10 V1/V3 (1:3) 10 (X 10 8 8 8 o 2 3 6 6 V1-V3 6 F - F 4 4 V1 4 2 2 2 1 0 0 0 135135 Time (min)

3.7 µM Capsaicin 7

6 V1/V3 (1:9)

) 5 V1/V3 (1:3) -4 4 V1-V3

(X 10 o 3 V1/V3 (1:1) V1 F - F 2

1 0 0 1 2 3 Time (min)

Fig. 5.4. Pharmacological comparison of cells transfected with TRPV1, or different ratios of TRPV1/TRPV3. HEK293 cells were transfected with pcDNA3, V1-V3, V1/V3 (1:1), V1/V3 (1:3) or V1/V3 (1:9). 3.7 µM capsaicin followed by 166 µM 2APB or 166 2+ µM 2APB-only were added to cells seeded in 96-well plates and [Ca ]i was monitored with fluo-4 fluorescence. Traces represent the average of triplicate experiments. Bold circles with 1 (Point 1) highlight the response of each specific set of transfected cells to 3.7 µM capsaicin only. This portion of the graph is enlarged below. Bold circles with 2 (Point 2) highlight the effect that 166 µM 2APB has on the response of each specific set of transfected cells to 30 µM capsaicin. Bold circles with 3 (Point 3) highlight the response of each specific set of transfected cells to 166 µM 2APB. Bold circles with 4 (Point 4) highlight the difference that 30 µM capsaicin has on the 2APB response for each specific set of transfected cells.

160

A

V34 V33 V11 V32 V12 V31 V13 V14 3 3 1 3 1 1 1 3 1 1 1 1 1 1 3 3 3 3 3 3 3 1 3 1 1 1 3 3

144V1-V32 4 1

B C 3 1 3 1 1 1 3 3 3 1 3 1 3 1 3 1 1 1 1 1 4 X 3 4 X 3 3 3 3 3 3 3 1 1 1 1 +

3 3 3 3 3 3 8 X 3 3

3 3 1 3 1 1 1 3 1 1 1 1 3 3 1 3 1 1 1 3 1 1 1 1 3 3 3 3 3 3 3 1 3 1 1 1 3 3 3 3 3 3 3 1 3 1 1 1 4 3 113 4 13 10 422 1

Fig 5.5. Model of TRPV1/TRPV3 heterotetramerization. A-C, V1 and V3 subunits are represented by circles with the number 1 or 3 in the center respectively. A, Theoretical distribution of V1 and V3 subunits in V1/V3 heterotetramers when assuming that an equal number of V1 and V3 subunits are present and that V1 and V3 have equal affinities for each other. This distribution was determined according to the binomial theorem: (A + B)n, where n is equal to 4 and represents the total number of subunits in V1/V3 heterotetramers and A and B represent V1 and V3 subunits respectively. The V1-V3 concatamer is at the right with half circles representing the amino acid linker between V1 and V3 subunits. B, A theoretical distribution of V1/V3 heterotetramers if V1 and V3 have higher affinities for their own respective subunits and equal concentrations of V1 and V3 subunits are present. C, A theoretical distribution of V1/V3 heterotetramers if V1 and V3 have higher affinities for their own respective subunits as in (B), but the number of V3 subunits has increased while the number of V1 subunits is the same as in (B). A comparison of (B & C) shows that a larger number of total heterotetrameric complexes will have at least one V1 subunits in (C) compared to (B), explaining the increased capsaicin responsiveness as the V3:V1 ratio increases. This occurs because the increased concentration of V3 will allow V3 to overcome the higher affinity that V1 has for other V1 subunits compared to V3 subunits. Also, a comparison of (B & C) reveals that more V34 and V33V11 channels will exist as the V3:V1 ratio increases, explaining the larger 2APB response as the V3:V1 ratio increases.

161

A B pH 5.6 pH 5.6 3.5 2.5 V1/V3 (1:3) 3.0 2.0 ) ) -4 -4 2.5 V1-V3 1.5 2.0 (X 10 (X 10 1.5 V1 1.0 V1 o o 1.0 V3 0.5 V3 0.5 F - F F - F 0.0 pcDNA3 0.0 pcDNA3 -0.5 -0.5 01 23 4 5 01 23 45 Time (min) Time (min)

152 µAPBM 2APB C pH 5.6 9 8 V1-V3 V1 ) 7

-4 6 V3 pcDNA3 5 (X 10

o 4 3

F - F 2 1 0 01234567 Time (min)

Fig. 5.6. The acid response of (TRPV3 and TRPV1) mimics the acid response of the TRPV1-TRPV3 concatamer. A, HEK293 cells were transfected with pcDNA3, V1, V3 or V1/V3 (1:3) at a 1 to 3 ratio. B, HEK293 cells were transfected with pcDNA3, V1, V3 or the V1-V3 concatamer. These traces are the acid-only portion of (C). A&B, The solution was acidified at 20 seconds. C, The potentiation of the acid response by 15 µM 2+ 2APB for the cells in (B). A-C, Cells seeded in 96-well plates and [Ca ]i was monitored with fluo-4 fluorescence. Traces represent the average of duplicate experiments.

162

A 1 µM Capsaicin RR 6 5

) 4 0 -4 0.37 3 1.11 (X 10 2 o 3.33 1 10

F - F 0 30 -1 -2 0123456 Time (min)

C B Individual TRPV Subunit

1.0 TRPV1 only V1:V3 = 1:3

0.8 2 3

V1:V3D = 1:3 1 4 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 641N 6 5 6 0.6 4 6 1 0.4 3 5 6 2

0.2 2 66 5 3

1 6 4 0 5 6

Normalized Response 0.1 1 10 100 4 1 P-loop 3 2 [RR] (µM)

Fig. 5.7. Coexpression of TRPV3D641N with TRPV1 results in capsaicin-sensitive channels with a reduced sensitivity to RR. A, dose-dependent inhibition of the response to 1 µM capsaicin of cells cotransfected with V1 and V3 at a 3 to 1 ratio [V1/V3 (1:3)]. Different concentrations of RR were applied to fluo-4-loaded cells seeded in 96- well plates at 20 seconds and then stimulated with 1 µM capsaicin at 160 seconds. B, Inhibition dose-response curve from equivalent experiments in (A). Shown are the least square fits of data points for V1-only (red triangles) , V1/V3 (1:3) (open circles) or V1/V3D641N (1:3) (filled circles). Data are averages of triplicate samples +/- SD 60 seconds after capsaicin application. C, Illustration of tetrameric V1, V3 or V1/V3 heterotetramers. Importantly, the characteristics of the pore (orange half-circles) are determined by contributions from each TRPV subunit (blue ovals). Circles with numbers represent transmembrane domains 1-6 of all TRP subunits.

163

CHAPTER 6

POSSIBLE MECHANISMS OF ACTION FOR THE 2APB INDUCED ACTIVATION OF TRPV3

6.1 Introduction

The thermosensitive TRP channels (thermoTRPs) V1, V2, V3, V4, M8 and A1 are activated by an extraordinary number of chemical compounds (Calixto et al., 2005).

Although subsets of chemical activators for each particular thermosensitive channel/receptor bear structural similarities, others bear no apparent resemblance, causing a great deal of confusion as to the mechanism in which these chemical compounds utilize for activation of these receptors/channels. Further confusion stems from the fact that some of the chemical activators show cross-efficacy for other thermoTRPs, while others do not (see Table 6.1). Although the emphasis of this chapter is to describe several possible mechanisms in which 2APB activates V1, V2 and V3, with an emphasis on V3, this chapter will begin with an overview of other chemical activators shown to modulate each of the thermoTRPs. This information will then be used to help support or discredit potential mechanisms of 2APB action that are proposed in this

164 chapter. To illustrate the diversity of chemicals involved, a brief review of the chemical activators specific for individual thermoTRPs will be provided in this section.

6.1.1 Chemical activators of TRPV1

The many pathological disorders that V1 plays a role in, coupled to the observation that capsaicin therapies successfully alleviate symptoms of some of these disorders, has led to a massive search for V1 specific chemical agonist/antagonists. This search has identified an enormous number of compounds that can activate V1. These compounds can be classified as capsaicinoids, gingerols, unsaturated dialdehydes terpenes, triprenyl phenols, gengenosides, phorbol esters (independent of PKC), camphor, boron containing compounds such as 2APB, or protons (Calixto et al., 2005; Bhave et al.,

2003; Xu et al., 2003; Tominaga & Tominaga 2005; Hu et al., 2004). The first group of compounds, the capsaicinoids, includes capsaicin and resiniferatoxin, the most well known V1 agonists. In fact, V1 is generally accepted as the protein that makes up capsaicin receptors of native peripheral sensory neurons (Caterina et al., 1997).

Interestingly, compounds in this group share a structural moiety referred to as a vanilloid group (Fig. 6.1A). Both the observation that radiolabled resiniferatoxin binds to V1, as well as V1 mutagenisis studies, have resulted in the identification of a loosely defined capsaicin-binding domain on V1, demonstrating that the activation mechanism for compounds of the capsaicinoid group activate V1 by a traditional ligand-protein interaction that results in an open channel conformation of the V1 channel (Tominaga &

165 Tominaga, 2005). Further support for a vanilloid-binding domain on V1 comes from the observation that the capsaicin analogue capsazipine competitively inhibits the ability of capsaicin to activate V1. In addition to capsaicin, resiniferatoxin, guaiacol, eugenol (Fig.

6.1A), capsiate, [8]-gingerol, [6]-gingerol, zingerone, and piperine are capsaicinoid compounds shown to activate V1, demonstrating that the vanilloid moiety is an important functional group for V1 activation (Calixto et al., 2005). Several non-capsaicinoid compounds that activate V1 are shown in Figure 6.1B. Figure 6.1 is not a comprehensive list of V1 agonists, with the compounds in this figure being chosen to emphasize the great deal of structural diversity amongst V1 agonists. Clearly, activation of V1 is not limited to compounds with a vanilloid moiety. Whether or not these compounds activate V1 by interacting with the same binding site as capsaicinoid compound is not clear, but the possibility that either additional agonist binding sites exist for V1 or that these compounds activate V1 by a completely unrelated mechanism is an important question considering the great demand for therapeutic treatments of the many known pathologies that V1 plays a role in. The mechanism in which V1 is activated by protons was discussed in Section 4.1.

6.1.2 Chemical activators of TRPV2

To date, the only known chemical activators of V2 are the boron containing compounds 2APB and DPBA (Hu et al., 2004, Xu et al., 2005) (Fig. 6.2A). The mechanism of action for the activation of V2 by these compounds is not known.

166 6.1.3 Chemical activators of TRPV4

To date, 4α phorbol 12, 13-didecanoate (4α-PDD), 5’, 6’-epoxyeicosatrienoic acid

(an epoxygenase product of arachidonic acid) and bisandrographalide A (BAA) have been shown to activate V4 (Fig. 6.2B) (Watanabe et al., 2003; Smith et al., 2006).

Interestingly, none of these compounds share an obvious structural similarity, except that they are large lipophilic molecules. Although the mechanism of action for the activation of V4 by these compounds is not known, hypotonic solutions activate V4 by causing cell swelling (Liedtke, 2005). This activation process occurs in a membrane delimited fashion, most likely as a result of the stretching and/or curvature of the membrane due to the increase volume of the cell in the presence of hypotonic solutions. Therefore, a possible mechanism of action for these compounds is that insertion of large lipophilic molecules into the outer leaflet of the plasma membrane mimics the biophysical effects of hypotonic solutions on the plasma membrane.

6.1.4 Chemical activators of TRPM8

Like with V1, a large number of molecules that activate M8 have been identified, of which, menthol is the most well known (Calixto et al., 2005; Macpherson et al., 2006).

Figure 6.2C shows the structures of several chemical activators of M8, including menthol, icilin, geraniol, WS-23 and eugenol. Like with V1, only a select few structures are shown in order to display the structural diversity of these molecules. Mutagenesis

167 studies have identified a critical tyrosine for the menthol-induced activation of M8

(Tominaga and Tominaga, 2005; Bandell et al., 2006). Interestingly, this residue is at the equivalent position as a critical tyrosine residue for the capsaicin-induced activation of

V1 (Tominaga and Tominaga, 2005). Although it is tempting to conclude that this is a critical amino acid in the ligand-binding pocket of M8, it is also possible that these tyrosine residues are only required for the conformational transition from a closed to an open state in the presence of chemical agonists.

6.1.5 Chemical activators of TRPA1

Like with V1 and M8, a large number of molecules that activate A1 have been identified (Jordt et al., 2004; Bautista et al., 2005; Calixto et al., 2005). Figure 6.2D shows the structures of several chemical activators of A1, including allyl isothiocyanate, cinnamaldehyde, arachidonic acid, carvacrol and eugenol. Only a select few structures are shown in order to display the structural diversity of these molecules. The mechanism in which these chemical compounds use to activate A1 is not known.

6.1.6 Chemical activators of TRPV3

2APB was identified as the first known chemical activator of V3 by our lab. Soon after, two 2APB analogues, diphenylboroanhydride (DPBA) and 2,2- diphenyltetrahydrofuran (DPTH), were shown to activate and inhibit V3 respectively

168 (Chang et al., 2005). Interestingly, although DPTH inhibited the 2APB-induced activation of V3, it enhanced the heat-induced activation of V3 at low concentrations.

Since then, polyunsaturated fatty acids (PUFAs) such as AA and ETYA, CDC, NDGA, eugenol, ethyl vanillin, vanillin, carvacrol and protons have been shown to activate V3

(Hu et al., 2006, Xu et al., 2006). In addition, results in this chapter demonstrate that V3 activity is enhanced by menthol, U73122, U73433, diethylstilbestrol (DES) and the phorbol ester PDBU. Structures of the chemical activators of V3 are shown in Figure 3.

These molecules were grouped according to structural similarities that will be discussed throughout this chapter. Like with the chemical activators of other thermoTRPs, most of these molecules share very little or no structural similarities, making the determination of the activation mechanism challenging.

6.2 Material and methods

6.2.1 DNA constructs, cell culture, and transfections

All compounds and reagents were purchased from Fisher or Sigma unless otherwise stated. 2APB was from Tocris or Cayman Chemical Company. All receptor cDNAs were cloned into pcDNA3. For all experimental procedures, see Section 2.2.

169 6.3 Results

6.3.1 U73122 and U73433 are chemical activators of TRPV3

In Section 4.3.9 it was demonstrated that 15 µM U73122 and/or U73433 each have a potentiating effect on the acid-induced activation of V3. In addition, it was demonstrated in Section 3.3.5.2 that 15 µM U73433 potentiated the receptor-induced activation of V3 and V3D641N. Although U73122 is effective at blocking the activity of

PLC, U73433 is ineffective, suggesting that the potentiating effect of these compounds is independent of PLC-dependent pathways. We therefore reasoned that U73122 and

U73433 could directly activate V3. To test this V3D641N or pcDNA3 transfected HEK293 cells that were seeded in 96-well plates were challenged with differing concentrations of

U73122 (Fig. 6.4A) or U73433 (Fig. 6.4B) and fluo-4 fluorescence was monitored at 32 oC using the FLEX-Station. As shown, 5 – 50 µM U73122 or U73433 directly activated

V3D641N in a dose-dependent fashion, identifying another lipophilic activator of V3. It should be noted that the only structural difference between U73122 and U73433 is that the double bond of the nitrogen containing ring is reduced in U73433 (Fig. 6.3F).

6.3.2 PDBU is a chemical activator of TRPV3

Both V1 and V4 are directly activated by phorbol esters, with PMA and 4α-PDD activating V1 and V4 respectively (Bhave et al., 2003; Gao et al., 2003; Vriens et al.,

170 2004). Interestingly, studies have shown that the activation mechanism used by these phorbol esters is independent of the PKC pathway, leading to the speculation that they may have affinity for V1 and V4. We therefore reasoned that V3 could also be activated by phorbol esters. To test this V3D641N or pcDNA3 transfected HEK293 cells that were seeded in 96-well plates were challenged with PMA (10 µM), 4β-PDD (10 µM), or

PDBU (16.7 µM) and fluo-4 fluorescence was monitored at 32 oC using the FLEX-

Station (Fig. 6.5). As shown, PDBU, but not PMA or 4β-PDD, activates V3D641N at this temperature. Phorbol esters activate PKC at 0.1 µM, and none of the phorbol esters tested activated V3 at this concentrations (data not shown), although phorbol esters were shown to potentiate the AA response of V3 at 0.1 µM (Hu et al., 2006). In fact, PMA does not directly activate V1 at room temperature in the presence of PKC inhibitors, although it does potentiate the capsaicin response with inhibitors present (Bhave et al.,

2003). It is therefore believed that PMA is a weak agonist of V1, and that the combination of sensitization of V1 via the PKC-induced phosphorylation of V1 subunits, combined with the agonist affect of PMA are responsible for the direct activation of V1 at concentrations of PMA that are higher than that needed to induce PKC activation.

Whether or not PKC phosphorylation plays a role in the PDBU induced activation of V3 was not determined in these studies, but the lack of a response at 0.1 µM PDBU, PMA, or

PDD suggests that phosphorylation by PKC alone is not enough to activate V3 at 32oC.

We therefore concluded that PDBU acts as a chemical activator and/or potentiator of V3.

171 6.3.3 Menthol is a chemical activator of TRPV3

After we discovered that 2APB was a chemical activator of V3, we retested whether or not several of the compounds that were previously shown not to directly activate V3 in Section 2.3.1 could potentiate the 2APB response (data not shown). Of the molecules tested, menthol (60 µM) was the only one shown to potentiate the 2APB- induced activation of V3 (data not shown). 60 µM menthol was chosen because this concentration will activate M8. We therefore applied higher concentrations of menthol

o to V3D641N or pcDNA3 transfected cells and monitored fluo4 fluorescence at 32 C to determine if menthol could activate V3 (Fig. 6.6). As shown, 333 - 3000 µM menthol

2+ induced an increase in [Ca ]i that was not seen in control cells (Fig. 6.6A). To ensure that menthol activation was not unique to V3D641N, we applied 2 mM menthol to cells transfected with either V3, V3D641N or pcDNA3 and compared the level of activation at

o 32 C (Fig. 6.6B). As expected, V3D641N cells were activated to a greater extent than V3 cells, but V3 was activated at this concentration. The pcDNA3 control cells did not respond to any of these menthol concentrations. These data demonstrate that menthol is a chemical activator of V3, but that much higher concentrations of menthol are needed to activate V3 than M8, revealing the possibility that menthol activates M8 and V3 via a similar activation mechanism.

172 6.3.4 DES is a chemical activator of TRPV3

As discussed in Section 6.1.6, the 2APB analogues DPBA and DPTH modulate

V3 activity, with the former having an enhancing effect and the latter an inhibitory effect.

Before this was published, we attempted to activate V3 with several other 2APB analogues including diethylstilbestrol (DES), phenytoin, diphenhydramine (benadryl) and t-stilbene (Fig. 6.5). These compounds were chosen because they were nearly as effective as 2APB at blocking the thrombosis response, which is presumably a TRPC1 dependent process (Dobbrydneva & Blackmore, 2001). The inhibitory action of 2APB on most TRP channels is discussed in Section 2.1. Of these compounds, only DES enhanced V3 activity between the concentrations of 3 – 300 µM (3 – 100 µM for benadryl due to solubility) (Fig. 6.7A). DES potentiated the response of 20 µM 2APB at

3 - 300 µM, but also directly activated V3D641N at 300 µM. The structures of the 2APB analogues tested are shown in Figure 6.7B.

5.4 Discussion

5.4.1 Chemical activators of TRPV3 can be grouped into several catagories.

Figure 6.3 displays compounds shown to modulate V3. These compounds are grouped as follows: 1) 2APB and its analogs (Fig. 6.3A). This classification was based on the fact that the molecules either contained boron, or had 2 phenyl groups in close

173 proximity that are not constrained within a plane. Additionally, each of these molecules, with the exception of NDGA, has been shown to block the TRPC1 mediated thrombosis response (Dobrydneva & Blackmore, 2001). 2) Various derivatives of phenol (only one phenyl group with a hydroxide group) (Fig. 6.3B). Although it is tempting to assume that the phenol group is necessary for activation when looking at these molecules, a comparison of thymol and menthol (Fig. 6.3B; Fig. 6.3C) illustrates that the phenol group may not be necessary. 3) Small lipophilic molecules with a polar group (Fig. 6.3C).

Molecules such as these would be expected to accumulate at the membrane water interface in a manner similar as cholesterol (see Fig. 6.11). 4) Molecules which contain a vanilloid moiety (Fig. 6.3D). These molecules are discussed in Section 5.4.2. 5)

Polyunsaturated fatty acids (PUFAs), with the double or triple bonds beginning at the fifth position from the carboxylate head group (Fig. 6.3E). These molecules were discussed in more detail in Section 3.3.5.3 and Section 3.4. 6) Large bulky lipophilic molecules (Fig. 6.3F). These molecules would be expected to localize to cellular membranes and could work similarly to molecules that activate V4.

5.4.2 Possible vanilloid-like binding site on TRPV3 and other thermoTRPs.

Several observations suggest that V3 may contain a vanilloid-like binding site, of which the most obvious is that eugenol, which contains a vanilloid moiety, activates V3

(Xu et al., 2006). In addition, capsaicin blocks the acid-induced and 2APB-induced activation of V3, possibly by binding to a vanilloid-like binding site in a similar manner

174 as capsazapine does for V1. Interestingly, eugenol also activates V1, M8, and A1, suggesting that a vanilloid-like binding site may exist for each of the thermally gated

TRP channels (Table 6.1). Of course, none of the compounds shown to activate V2 or

V4 contain a vanilloid-like moiety, suggesting that a different activation mechanism exists for V2 and V4. Although, if an assumption is made that 2APB binds to a vanilloid-like binding site of V1 and V3, then the argument can be made that V2 also contains a vanilloid-like binding site, as 2APB and its analogue DPBA activates V2. In addition, phorbol esters have also been shown to activate V1, V3 and V4 in a PKC- independent manner. Therefore, if it is assumed that phorbol esters activate V1 and V3 by binding to a vanilloid-like moiety, the argument could be made that V4 also contains a vanilloid-like binding site. If in fact each of the thermoTRP channels contain a vanilloid- like binding site, a comprehensive analysis of the structures of the compounds shown to activate each of the different thermoTRPs suggests that the binding site is loosely conserved, so as to allow selective differentiation of chemical species, while still binding similar molecules. In fact, eugenol, cinamaldehyde, 2APB, carvacrol, menthol, camphor, phorbol esters and PUFAs all show cross-reactivity between certain members of the thermoTRPs, although specific phorbol esters and PUFAs are specific for individual thermoTRPs (Table 6.1). To confound things further, Table 6.1 also demonstrates that some compounds are specific for only one receptor, as capsaicin, BAA, icilin and allyl isothiocyanate activate only V1, V4, M8 and A1 respectively. It should be noted that, in addition to binding to a vanilloid-like binding site, alternative explanations for the efficacy of 2APB toward activating V3, as well as V1 and V2 exists. In the remainder of

175 this discussion, a critical analysis of the literature will be used to support or discredit various possible mechanism in which 2APB could activate these channels. This will begin with a review of the available literature with regard to the effect of 2APB on V1,

V2 and V3. Following this, a structural comparison of 2APB with other V3 activators will be used to illustrate several possible mechanisms in which 2APB can activate V3.

This discussion is not meant to conclude that a particular mechanism is correct, but to provide a critical analysis of the available literature, laying a foundation for a future hypothesis driven determination of the actual mechanism used by 2APB to activate V3, as well as V1 and V2.

6.4.3 2APB as an activator of TRPV1

In whole-cell recordings, 2APB dose-dependently activated currents in HEK293 cells transiently transfected with mouse V1 (Hu et al., 2004). Control cells did not show any response to 2APB at concentrations as high as 3 mM. Under similar conditions,

TRPC6 currents activated by OAG and M8 currents activated by menthol were inhibited by 2APB in a dose-dependent manner. The EC50 values of 2APB for V1 were 197 ±13

µM at –100 mV and 130 ± 17 µM at +100 mV. The current-voltage (I-V) relationship of the 2APB-evoked currents was not different from that elicited by 1 µM capsaicin. In order to confirm that the stimulatory effect of 2APB on V1 was not a unique property of

HEK293 cells, this effect was also studied in the context of Xenopus oocytes (Wang et al., 2004). At 300 µM, 2APB elicited an inward current at –40 mV in oocytes that were 176 injected with cRNA of mouse V1 but not the uninjected oocytes or those injected with the cRNA for an inactive form of V1. The current was completely blocked by 3 µM ruthenium red (RR) but only partially blocked by 30 µM capsazipine (apprx. 30%), confirming specificity toward V1. Interestingly, 30 µM of capsazipine completely blocked the currents evoked by 1 µM capsaicin in these same cells. Since capsazipine has been shown to only partially block the acid- and heat-induced currents through rat V1

(McIntyre et al., 2001), these data suggest that the site of action for 2APB and capsaicin is different.

A characteristic feature associated with the polymodality of V1 is that known activators such as heat, protons and capsaicin act in a synergistic manner. This was demonstrated for 2APB with respect to other known V1 activators. In HEK293 cells, co- application of 0.3 µM capsaicin and 100 µM 2APB, or 100 µM 2APB at pH6.5, greatly increased the V1 current at –100 mV more than 20 fold as compared to the stimulation with capsaicin, 2APB or the weak acid (pH 6.5) alone. The synergistic effect was more nicely illustrated in Xenopus oocytes when the dose-response curves for capsaicin and protons were compared in the absence and presence of 100 µM 2APB and those for

2APB determined in the absence and presence of 0.3 µM capsaicin or weak acid. It was shown that 100 µM 2APB left-shifted the dose-response curve for capsaicin 3.8 fold and the pH dependence 6.6 fold. Conversely, capsaicin and weak acid also caused left-shifts of the dose-response curve to 2APB 9.3 and 2.0 fold, respectively. Furthermore, about a

9-fold increase in current at –40 mV was obtained when 100 µM 2APB was applied at

40°C as compared to the same 2APB concentration at 22°C (Hu et al., 2004). Typically,

177 strong potentiation of one stimuli for another is the result of a cooperative effect caused by two stimuli acting at different sites (or by different mechanisms), also suggesting that

2APB and capsaicin activate V1 by a different mechanism. In conclusion, these data suggest that 2APB does not active V1 by binding to the capsaicin binding site. Other possible mechanisms include a separate binding site specific for 2APB and its analogues, interaction with the pore, or effects on the membrane.

6.4.4 2APB as an activator of TRPV2

2APB-evoked whole-cell currents have been observed in HEK293 cells transiently transfected with mouse V2 (Hu et al., 2004). At 22°C, this activation was very weak when using 1 mM 2APB, but became strong at 3 mM 2APB. The currents showed weak double rectification and were blocked by 3 µM RR. Recently, Chung et al.

(2005) also confirmed the effect of 2APB (320 µM) on eliciting an intracellular Ca2+ increase in HEK293 cells transiently expressing rat V2 at room temperature. On the other hand, for an endogenous channel encoded by mouse V2 in the F-11 hybridoma derived from rat dorsal root ganglia (DRG) and mouse nueroblastoma, the presence of

100 µM 2APB did not significantly change the temperature threshold of current activation at –60 mV (Bender et al., 2005), indicating that either there may be other requirement(s) for the activation of TRPV2 by 2APB, or that the lipid environment in which V2 is expressed influences the ability of 2APB to activate V2.

178 6.4.5 2APB as an activator of TRPV3

The effect of 2APB on V3 was discussed in detail in Chapter 2. Also of relevance, it has been shown using whole cell patch clamping of HEK293 cells transiently expressing V3, that infusion of 1 mM 2APB into the cell through the patch pipette for > 6 minutes failed to elicit any current while subsequent application of 2APB in the bathing solution elicited TRPV3 currents, indicating that the site of action for

2APB is extracellular (Hu et al., 2004). This observation is consistent with the site of action for 2APB not being a capsaicin-like binding site, as the determinants of capsaicin induced activation of V1 are located on the intracellular side of the membrane (Tominaga and Tominaga, 2005).

6.4.6 The effects of 2APB analogs on TRPV channels

2APB analogs were first studied in order to identify molecules that would inhibit

Ca2+-influx induced by thrombin in human platelets, a process that is believed to involve

TRPC1 (Rosado et al., 2002). In their original work, Dobrydneva and Blackmore (2001) showed that like 2APB, DPBA and DPTH (see Fig. 6.3A for structures) could inhibit the thrombin-induced Ca2+ signal with a similar efficacy as 2APB. This had led Chung et al.

(2005) to explore the possibility that these 2APB analogs would activate V3. Using Ca2+ imaging, they showed that 100 µM DPBA, but not 100 µM DPTH caused a rise in

179 intracellular Ca2+ in HEK293 cells transfected with V1, V2 or V3, but not those transfected with vector only or V4. Interestingly, 100 µM DPTH inhibited the response evoked by 100 µM 2APB and 100 µM DPBA by 73.2% and 93.2%, respectively, in V3- transfected cells. Unlike with V3, 100 µM DPTH did not inhibit the response of V1 and

V2 to 100 µM DPBA. Even at 500 µM DPTH, the inhibition was 25.2% and 33.2% for

V1 and V2, respectively. These results suggest that DPBA activates V1, V2, and V3 in a similar fashion as 2APB, but DPTH has an opposite action and may be more selective for

V3.

In whole-cell patch clamp studies of V3 expressed in HEK293 cells, Chung et al.

(2005) demonstrated that similar to 2APB, 32 µM DPBA evoked outwardly rectifying currents that became dually rectifying with successive application of the drug. In addition, DPBA-evoked currents were blocked by DPTH (58.9 and 90.8 % inhibition at

+80 and –80mV, respectively) or 10 µM RR (99% at –80 mV). A dose-dependent analysis of DPBA yielded EC50 values of 64.1 µM and 85.1 µM at +80mV and –80mV, respectively, with Hill coefficient values larger than 2. Under the same conditions, the

EC50 values obtained for 2APB were 90.6 and 165.8 µM (Hill coefficients ≈ 1.7) at +80 and –80mV, respectively. These values are greater than those obtained by the same group of investigators in an earlier study (Chung et al., 2004a). The data suggest that

DPBA has a slightly higher affinity and a further degree of cooperativity for activation of

V3 than 2APB. The authors also noted an inhibitory effect at high (>100 µM) DPBA and

2APB concentrations, which is characterized by a decline in current amplitude at 1 mM as compared to 0.3 mM DPBA, a desensitization in the continued presence of the drug,

180 and a strong rebound immediately after the washout. The IV relationship during the rebound appeared linear, indicative of a near maximal activation of the V3 channel. A possible explanation for this dual action phenomenon is that DPBA has two sites of action, where one is stimulatory and the other inhibitory. Although a single site of action being modulated by an intrinsic “desensitization” pathway is also possible, the rebound at the washout and the fact that V3 is sensitized but not desensitized upon repetitive stimulations make it unlikely.

The inhibitory action of DPTH on V3 also appeared to have two kinetic components. The IC50 values at –80mV were 6.0 µM and 151.5 µM and those at +80mV were 10.0 µM and 226.7 µM for the first and second components, respectively. Thus,

DPTH also has two or more sites and/or mechanisms of action on V3. In light of the fact that 2APB, DPBA, and DPTH all blocked SOCE in platelets, and that they each have the ability to inhibit V3 at high concentrations, it is possible that the low affinity site of

DPTH is shared by 2APB and DPBA at high (>100 µM) concentrations and is inhibitory for all three compounds. This accounts for the rebound during washout as the higher affinity stimulatory site may be fully occupied during the washout process. The stimulatory site may also be shared by the three compounds with similar, but nonetheless relatively high, affinities. Hence, they could compete for binding to the same site.

However, a structural feature important for activation may be lacking in DPTH, resulting in inhibition even though it is bound to the “stimulatory site”, especially in the presence of other stimulating compounds. Indeed, 100 µM DPTH was found to potentiate the heat-evoked response of V3 (Chung et al., 2005). Thus, the complex activation/inhibition

181 phenomenon observed with the 2APB analogs could be a result of dual bindings to separate stimulatory and inhibitory sites with different affinities. It should be noted that

100 µM DPBA or a 45°C heat challenge alone was unable to elicit V3-like currents in mouse keratinocytes, but superimposition of these two stimuli, like with 2APB, resulted in outwardly rectifying currents that sensitized upon repeated stimuli, suggesting that either other requirements are necessary for efficient activation by 2APB or that the lipid environment influences the ability of 2APB to activate V3. Additionally, we have shown that DES, but not phenytion, trans-stilbene, or diphenhydramine (benadryl) potentiates the response of V3 to 20 µM 2APB (Section 6.3.4). Clearly, 2APB and its analogs will be useful tools in determining the cellular and physiological functions of thermoTRPV family members.

6.4.7 Structural considerations for 2APB action

Due to the ability of 2APB to form an N
B coordinate bond, 2APB can exist in several different states (Fig. 6.8). Analyses on 2APB and several of its analogs by crystallography (Rettig and Trotter, 1976), pKb values in aqueous solution (Dobrydneva and Blackmore, 2001), and NMR (Dobrydneva et al., 2006) support the idea that 2APB exists predominantly in the monomer ring structure as shown in Figure 6.8, with the ethanolamine side chain forming a five membered boroxazolidine heterocyclic ring

(Strang et al., 1989; Dobrydneva and Blackmore, 2001). The fact that 2APB can block the intracellularly located IP3Rs supports the monomer ring structure because if 2APB 182 existed predominantly as the open chain form, the nitrogen of the ethanolamine side chain would most likely be protonated in order to neutralize the free electron pair. This positively charged form of 2APB would not be expected to pass through the membrane readily. 2APB can also form dimers (Fig. 6.8) (Nöth, 1970; van Rossum et al., 2000). It should be considered that the ability of 2APB to switch between these different forms may also be important for its functional ability to activate and/or block TRPV1-3 as well as other TRP channels.

The boron on 2APB allows for the formation of coordinate bonds between the electrophilic boron and nucleophiles. In fact, it has been proposed that 2APB and its boron containing analogs can form either N
B or O
B coordinate bonds with various amino acids (Dobrydneva et al., 2006). Boron trifluoride, a catalyst used in many organic chemical reactions, forms nitrogen adducts with amines, imidazoles, pyridines, and acetonitriles (Dobrydneva et al., 2006) and possibly carboxylates. These qualities of

2APB and its boron containing analogs allow for the potential that their stimulatory and/or inhibitory sites of action could be formed via coordinate bonds with amino acids that contain amines, imidazoles, and carboxyl groups on V1, V2, and V3 channels.

Interestingly, even though dimethyl 2APB (Fig. 6.8) blocked the thrombin-induced

SOCE in platelets (Dobrydneva et al., 2006), a non-boron analog with two methyl groups on the secondary amine nitrogen, diphenhydramine, at 100 µM was ineffective in blocking SOCE in platelets (Dobrydneva and Blackmore, 2001) and in activating V3 ectopically expressed in HEK293 cells (Chung et al., 2004a, Section 6.3.4). It would be interesting to test whether dimethyl 2APB activates V3. A positive effect would suggest

183 that boron and/or ring formation is necessary for the stimulatory action of 2APB analogs since the tertiary carbon and the secondary amine nitrogen of dipenyhydramine are unable to make the ring closure like the N
B coordinate bond of the 2APB monomer ring (Fig. 6.8). The blocking and potentiating effects of DPTH on V3 (Chung et al.,

2005) as well as the ability of several other non-boron analogs of 2APB to block SOCE in platelets (Dobrydneva et al., 2006), presumably mediated through TRPC1, suggest that the boron may not be necessary, at least for modulating TRP channels. However, without the boron, the compound may not be sufficient to activate the channel because heating appears to be necessary to reveal the stimulatory effect of the non-boron analog, DPTH, on V3 (Chung et al., 2005). In addition, 300 µM phenytoin (Fig. 6.7B) was also unable to activate V3. Even with the DES induced activation of V3, high concentrations (300

µM) were necessary to activate V3D641N, which is more sensitive than the wild type receptor. Although an analysis of the effects of 2APB analogs on the thrombosis response suggests that ring formation is beneficial, the potentiating effect of DES and

NDGA on V3 suggests that it is not necessary. The additional polarity of the boron or the ability of 2APB to convert between different forms could contribute to the activation mechanism.

It is worth mentioning that studies using several 2APB analogs to block the thrombin-induced SOCE in platelets have concluded that for the inhibitory effect, the minimal molecular requirements were two phenyl groups attached to either tetrahedral carbon or boron (Dobrydneva et al., 2006). However, an extensive modification of the five-membered ring could be detrimental (Dobrydneva and Blackmore, 2001) and under

184 certain conditions, convert the drug into an activator of Ca2+-influx, albeit only at a high concentration (Dobrydneva et al., 2006). It should be noted that the ability of DES to block the thrombosis response (Dobrydneva and Blackmore, 2001) as well as the ability of DES and NDMG, but not t-stilbene, to activate V3 suggest that the two phenyl groups do not have to necessarily be attached to a tetrahedral carbon or boron, but instead just need to be in close proximity and not be confined to a plane (note that the conjugation of t-stilbene holds this molecule in a planar confirmation). Although caution should be taken in imparting the effects of 2APB analogues on the inhibition of the thrombosis response to those of activation of V3, this method has proven to be very useful.

6.4.8. Possible site(s) of action for 2APB

There are several possible ways in which 2APB could activate V3 including (Fig. 6.9):

1) 2APB binds to an extracellular binding pocket that is typical for a ligand-protein interaction or forms a coordinate bond between amino acid side chains located on the extracellular side and the boron of 2APB. 2) The 2APB site of action is the pore, possibly exerting its effects similar to protons. 3) Disruption of lipid-protein interactions within the membrane, of which negatively charge phospholipids such as PIP2 are prime candidate. 4) 2APB activates V3 by affecting membrane properties. 5) 2APB activates a signaling pathway independent of V3 proteins, with activation of these pathways modulating V3. The following sections will explore these possibilities.

185 6.4.8.1 Does the activation of TRPV3 by 2APB involve a 2APB binding pocket?

Several lines of evidence favor the existence of at least two binding sites for

2APB and its analogs, with one being stimulatory and the other inhibitory. First, at low concentrations, 2APB has been shown to potentiate a native SOCC that is normally blocked by higher concentrations (Prakriya and Lewis, 2001). Second, at above 100 µM,

DPBA-evoked V3 currents tended to reach a maximum value and then declined in mid- response (Chung et al., 2005). This effect became more obvious at higher concentrations of DPBA and led to an apparent reduction in the maximal current amplitude at 1 mM. A similar response also occurred to 2APB, but at somewhat higher concentrations. Third, even though DPTH has a predominantly inhibitory effect, it potentiated the heat-induced

V3 currents (Chung et al., 2005). Fourth, the inhibition of DPBA-evoked V3 currents by

DPTH extended over several orders of magnitude and had two kinetic components, indicative of two or more sites of action. One of these inhibitory sites could result from competition with 2APB or DPBA for binding to the stimulatory site. This two-sites model could explain the concentration-dependent dual actions of the 2APB analogs. If the model holds true, modification of the 2APB structure may generate analogs with greater differences in the affinities to the stimulatory and the inhibitory sites and for different TRP subtypes, allowing for highly specific agonists and/or antagonists to be made for some TRP channels. This exciting possibility warrants an extensive modification of 2APB analogs and evaluation of their effects on multiple TRP channel types.

186 For TRP channels, the site of action of 2APB is most likely at the extracellular side of the plasma membrane. This is supported by the failure of intracellular injection of

2APB to activate any V1 current in Xenopus oocytes and intracellular infusion of 2APB and DPBA through patch pipettes to activate V3 expressed in HEK293 cells in whole-cell experiments (Hu et al., 2004; Chung et al., 2005). In HEK293 cells, this same manipulation also failed to inhibit TRPC3 and TRPC5 channels (Trebak et al., 2002; Xu et al., 2005a). In all cases, subsequent application of 2APB or DPBA in the bathing solution had elicited either stimulation or inhibitory responses of the TRP channels. In excised inside-out patches, 2APB also failed to inhibit TRPC5 channel activity whereas in outside-out patches, the same concentration of 2APB effectively blocked the channel

(Xu et al., 2005a). One exception is that V3 is activated by 2APB applied to the intracellular side of the inside-out patches (Chung et al., 2004b). This could be explained by the notion that 2APB is membrane permeable and can be accumulated at the pipette side (outside) even though it is applied to the exposed side of the membrane patch.

Similar accumulation of 2APB at the extracellular side will not occur in the outside-out or whole-cell configurations as the drug will be diluted by the bath solution or washed away by perfusion. However, this does not explain why 2APB failed to inhibit TRPC5 in the inside-out patches.

The available data also suggest that 2APB acts at a different site(s) from those of known V1 agonists. First, V2 and V3 are not activated by capsaicin but they are activated by 2APB. Second, while capsazipine, a competitive antagonist of capsaicin, completely inhibited the capsaicin-induced response, it only partially blocked the 2APB-

187 evoked activation of V1. Third, superimposition of 2APB and capsaicin invoked responses that were more than additive to those elicited by each drug alone. A similar synergistic effect was also observed between 2APB and weak acid, indicating that different mechanisms are involved for the activation of V1 by 2APB, capsaicin and protons. It is possible that a similar 2APB-binding pocket exists for V1, V2, and V3, but it is very different from the vanilloid-binding pocket, which is mostly intracellular (see

Tominaga and Tominaga 2005 for a review on vanilloid binding sites of V1). It is also unlikely that 2APB shares a common mechanism with protons as they are very different in structures (but see next).

6.4.8.2 Does the activation of TRPV3 by 2APB involve the pore?

The pore could be a site of 2APB action. Potential interactions could occur through either an electrostatic effect by the positively charged protonated 2APB open chain monomer or the formation of a coordinate bond between 2APB and amino acid residues that contain amines, carboxylates, and imidazoles. Indeed, it has recently been proposed that the protonated 2APB (Fig. 6.8) might inhibit the Gd3+-incuced potentiation of TRPC5 by competing for negatively charged amino acids on the pore (Xu et al.,

2005b). If 2APB were to become positively charged in the pore environment, this could explain why a large number of TRP channels are blocked by 2APB. In this model one could imagine a positively charged 2APB blocking the pore via electrostatic interactions similar to the way lanthanides and extracellular Ca2+ do. These cations have been shown

188 to either potentiate or inhibit TRP channels (Jung et al., 2003; Shi et al., 2004).

Interestingly, several glutamic residues near to pore helix are found to be critical for the potentiation of TRPC5 by cations (Jung et al., 2003). Likewise, protonation of acidic residues near the pore helix is also involved in the potentiation and inhibition of V1 and

V5, respectively, and it may do so by causing a rotation of the pore helix (Jordt et al.,

2000; Yeh et al., 2005). In this sense, there could be a similarity between 2APB and protons in regulating TRP channels. However, this model cannot explain the activity of

DBPA or DPTH since neither of these compounds is positively charged regardless of the conditions or environment. While DPBA could potentially form a coordinate bond with negatively charged amino acids, and in this way functionally mimic the electrostatic association of cations with the pore, DPTH would not.

6.4.8.3 Does the activation of TRPV3 by 2APB involve a signaling pathway?

An additional site of action could be outside of the channel complex, for instance signaling steps involved in the activation of TRP channels. Such an example has been presented in chicken B cells, where 2APB activates phospholipase γ2 and in turn causes activation of both endogenous SOCE and the ectopically expressed human TRPC3 (Ma et al., 2003). However, this would put the site(s) of action at the intracellular side of the plasma membrane, which would be unlikely in HEK293 cells, as discussed earlier. In addition, activation by signaling pathways would most likely entail a slower activation mechanism than is observed for V3, although the PDZ binding sequence on V3 could

189 result in the formation of a signaling complex, increasing the speed and efficiency of activation of V3 by signaling pathways that are part of the complex.

6.4.8.4 Does the activation of TRPV3 by 2APB involve changes in plasma membrane properties?

Until there is proof that 2APB analogs exert their actions through binding to channel subunits or other protein components of the regulatory complex, it cannot be ruled out that the mechanism of action involves changes in membrane properties. To fully illustrate how changes in membrane properties could affect thermoTRP channel activity, a very brief review of relevant membrane principles is in order. Figure 6.10A shows that as temperature is increased, membrane fluidity also increases. The fluidity increases because of the energetic cost of keeping the hydrocarbon fatty acid tails in an extended, rigid conformation also increases with increasing temperature; an effect that is attributed to entropy. This becomes important when considering that the total volume of membranes does not change. For instance, biophysical studies have revealed that if a membrane is stretched laterally, the membrane will become thinner so that the total volume of membrane is unchanged (Hamill and Martinac, 2001). This principle is important with regards to increased temperature because as the temperature is increased, the hydrocarbon chains will be less extended (more disordered). Also relevant is that the hydrocarbon chains of phospholipids at the higher temperatures will maintain the same number of hydrophobic contacts between neighboring phospholipids. Since the hydrophobic tails increase their disorder (entropy driven), while maintaining the same

190 number of Van der Wals contacts (enthalpy driven), the net result is a slight thinning of the membrane. Protein-membrane interactions are extremely important in determining the conformation of membrane proteins (Fig. 6.10B). This is again due to the hydrophobic affect; there is a large energetic cost in exposing hydrophobic amino acids on the exterior of the transmembrane portion of channel proteins. If the membrane

“thins”, the channel protein will also correspondingly “thin” unless the energetic cost of changing to the “thin” conformation is greater than the energetic cost of exposing hydrophobic residues to the polar water molecules at the protein-lipid-water interface, demonstrating that the conformation that a channel protein assumes is dependent not just on the thermodynamic constraints of the channel protein itself, but also on the lipid environment. In fact, the lipid-water interface may have profound effects on membrane protein conformation by determining the distribution of lateral pressures within the membrane (Fig. 6.10C). This occurs because lipid bilayers are anisotropic, having very different biophysical properties at different depths. The free-energy reduction in ordering waters and lipids at the lipid-water boundary results in a large surface tension between the polar head groups of lipids and the non-polar tails (Kung et al., 2005). Deformation of the lipid-water interface can provide the energetic drive for an open channel conformation if redistribution of the lateral pressure profile within the membrane favors the open conformation over the closed conformation for a particular channel protein. It should be noted that what is occurring at the interface is of particular importance. This is reflected in the role of cholesterol to membrane integrity, as cholesterol inserts itself in

191 the membrane in such a manner as to stabilize the first couple hydrocarbons of the fatty acid tail, thus making the membrane more rigid and resistant to deformation at the interface.

The manner in which small temperature changes activate thermoTRP channels is a mystery due the effects that a few degrees Celcius temperature change can have on transitioning these channels from a closed to an open conformation, a phenomenon that is reflected in their large Q10 values. When focusing on the channel protein itself, it is very difficult to hypothesize a mechanism in which a few degrees Celcius temperature change could significantly affect channel conformation. Conversely though, if one considers the global effect that small temperature changes can impart to membranes, in particular at the lipid-water interface, the concept of thermally-gated channels is not as perplexing. For example, it is easy to imagine that the open and closed conformations of thermoTRPs have small differences in Gibbs free energy if a “tilting” of the transmembrane domains to a “thinner” channel (open conformation) is of similar energy as the “thicker” conformation (closed conformation). In contrast, a large difference in Gibbs free energy should occur upon exposure of hydrophobic residues to the polar solvent molecules if the channel does not conform to the thickness of the membrane. Additionally, these channels could have energetically favored electrostatic interactions with the phosholipid head groups that are lost if they do not conform to the thickness of the membrane.

Interestingly, it has recently been proposed that thermally gated channels have a voltage- sensor, and that temperature and ligand/chemical activators regulate these channels by shifting the voltage-dependence to more physiological potentials (Voets et al., 2004;

192 Nilius et al., 2005b). Unfortunately an obvious voltage sensor within the channel complex of thermoTRPs does not exist, as thermoTRPs do not have a corresponding stretch of positively charged amino acids that acts as the voltage sensor of voltage-gated potassium, sodium and calcium channels. An important question is whether or not the lipid-water interface of plasma membranes can act as the voltage-sensor for the thermoTRPs. For example, the resting membrane potential of a typical animal cell is –70 mV. With an average thickness of the phospholipid bilayer of 3.5 nM, the resulting voltage across the membrane is 2 X 105 V/cm, 10,000 times larger than that of high voltage power lines. With such large voltages involved, it is be expected that changes in membrane potential will affect the lipid-water interface, as the head groups of membrane phospholipids are charged. Considering that charged and polar amino acids of channel proteins interact with the charged head groups of phospholipids, and that biophysical studies have revealed a complex interplay between various potentials on transmembrane proteins and cellular membranes (O’Shea, 2003), it should be considered that the voltage- sensor of thermoTRPs may very well be the lipid-water interface. In the remaining portion of this discussion, the manner in which these principles of membrane biophysics could be responsible for the complicated assortment of chemical activators shown to activate various thermoTRPs will be provided, although the emphasis will be on V3.

Several observations suggest that membrane properties strongly influence the activities of thermosensitive channels (Deol et al., 2004). First, the thermoTRPs are activated by a large number of lipophilic molecules, many of which bear no structural similarity (Calixto et al., 2005). The ability of so many chemical species to activate the

193 thermoTRPs could be explained if these lipophilic molecules accumulate in the membrane, particularly at the lipid-water interface. The manner in which this could occur will be illustrated in the next point. Second, increasing the cholesterol content in

HEK293 cells shifts the temperature threshold of V1 from 42 to 46°C (Liu et al., 2003).

As mentioned, cholesterol is known for its ability to decrease the fluidity of membranes.

The polar head group and rigid steroid-like portion of cholesterol fit into each of the leaflets of the membrane in such a manner as to stabilize the first few carbons of the adjacent hydrocarbon tails of membrane phospholipids. In addition, the hydrocarbon tail of cholesterol provides an extended chain of carbons which promotes efficient Van der

Waals contacts with the hydrocarbon tails of phospholipids, and thus insertion of cholesterol does not require an extensive reorientation of the hydrocarbon tails of phospholipids to maintain a maximal number of Van der Waals contacts (Fig. 6.11). In contrast, molecules such as ginsenosides or evodiamine (Fig. 6.1B), which also have steroid like regions, may insert themselves into the membrane in a similar fashion as cholesterol, but have a destabilizing effect either at the lipid-water interface or deep within the membrane where, unlike cholesterol, these molecules do not provide an extended hydrocarbon tail for efficient Van der Waals contacts between adjacent phospholipid hydrocarbons. The reorientation of the hydrocarbon tails of phospholipids that would follow insertion of these molecules into the plasma membrane could mimic temperature increases (increased disorder) in many aspects. Interestingly, U73122 and

U73433 (Fig. 6.3F) also contain a steroid-like region, but these molecules provide an extended hydrocarbon like chain that ginsenosides or evodiamine does not. Unlike with

194 cholesterol, the polar nitrogen-containing ring of U73122 and U73433 most likely localizes to the lipid-water interface of the membrane. Although molecules such as camphor, menthol and carvacrol do not have a steroid-like region, the polar oxygen on these molecules may act as head group similar to the hydroxyl group on cholesterol. If these molecules were to insert themselves into the membrane at the lipid-water interface, the rest of the molecule would cause a bulge at the critical first few carbons on the adjacent phospholipids, and instead of promoting a more ordered crystalline like membrane similar to cholesterol, promote a more disordered array as the membrane tries to maximize Van der Waals contacts (Fig. 6.11C). Similar to lipophilic molecules with polar bonds, aromatic molecules accumulate at the lipid-water interfaces (Gaede et al.,

2005). This occurs because the ring of conjugated double bonds has an inducible dipole that prefers to be associated with the polar solvent (Mecozzi et al., 1996; Dougherty,

1996). At the same time, the hydrophobic nature of carbons prefers the hydrophobic region of membranes. In fact, the recent determination of the structure of several channel proteins to atomic detail has revealed that aromatic amino acids prefer to localize to the lipid-water interface, and part of the activation mechanism of many channels entails a shift of buried aromatic amino acids to the lipid-water interface (Domene et al., 2003,

Domene et al., 2005). Interestingly, molecules such as 2APB would be expected to accumulate at the lipid-water interface based on both the presence of polar bonds, and the presence of phenyl groups. Third, PIP2, a negatively charged phospholipid, has been proposed to hold V1, and possibly other TRPVs, in an inhibitory state (Prescott and

Julius, 2003). Fourth, arachidonic acid and other unsaturated fatty acids activate and/or

195 potentiate the 2APB-induced activation of V3 (Hu et al., 2005). The great variability in the fatty acids used, to include triple bonded analogs may suggest a “loosely” specific activation mechanism that could be accounted for if these molecules cause a change in the membrane biophysical properties that are “sensed” by the channel. It is possible that

PUFAs insert themselves into the membrane in an orientation similar to that of the fatty acid tails on phospholipids. Since the ability of PIP2 to inhibit V1 (and probably V3) requires the receptor to bind to PIP2, free PUFAs may compete for this interaction and displace them form the PUFAs covalently linked to phospholipids such as PIP2.

Conversely, PUFAs have been proposed to accumalate at the lipid-water interface and act as crenators, molecules that push phospholipids apart in the lateral direction, causing the membrane to stretch and/or bend (Fink et al., 1998, Patel et al., 2001) (Fig 6.11D).

Polyunsaturated fatty acids have also been shown to regulate TRPV channels in C. elegans (Kahn-Kirby et al., 2004) and TRPC channels in Drosophila (Chyb et al., 1999).

Fifth, it has recently been proposed that mechanosensitive and thermal sensitive channels may be modulated by a common mechanism (Kung, 2005) and many mechanosensitive channels are modulated in a membrane delimited fashion (Hamill and Martinac 2001).

2APB is a lipophilic molecule that possibly could accumulate in the membrane at high concentrations (Fig 6.11C). There are several ways in which a lipophilic molecule such as 2APB could modulate TRP channels. First, if 2APB accumulates in the membrane at high concentrations, then 2APB and its analogs could disrupt the interaction between various inhibitory phospholipids, such as PIP2. Second, the observation that

2APB and its analogs affect so many ionic channels and other membrane proteins, most

196 in an inhibitory fashion, suggests that 2APB could act in a similar fashion as general anesthetics. A property of the anesthetics is that they usually affect the gating of many different ion channels by altering membrane properties (Antkowiak, 2001; Campagna et al., 2003). Usually this gating is inhibitory, but in the case of TREK-1, a thermosensitive

K+-channel, it is stimulatory. It is also interesting to know that TREK-1 is activated by unsaturated fatty acids and heat (Patel et al., 2001; Heurteaux et al., 2004). It has also been proposed that the best anesthetics accumulate at the membrane-water interface

(North and Cafiso, 1997). The high degree of lipophilicity along with the polarity of the

N
B coordinate bond could result in the accumulation of 2APB in this region.

Interestingly, the polyunsaturated fatty acids are believed to concentrate in this region.

Crenators, e.g. Trinitrophenol, activate TREK-1 by affecting the interface of the membrane (Patel et al., 2001). Additionally, lysophospholipids activate TREK-1 by inserting into the outerleaflet of membrane phospholipids (Maingret et al., 2000; Danthi et al., 2003). These molecules affect the membrane because their headgroup width is greater than the width of their hydrocarbon tails (cone shaped), causing a redistribution of the hydrocarbon tails of phospholipids in order maximize Van der Waals contacts (Fig

6.11 B&C). Additional support for this mechanism comes from the observation that as the width of the headgroup increases, so does the degree of activation of TREK-1. It would definitely be interesting to test if these molecules that activate TREK-1 also activate V3. In addition, application of an alcohol series to V3 expressing cells could reveal whether or not anesthetics activate V3 similarly to TREK-1 (typically, this is the first step in determining whether or not anesthetics affect channel function).

197 6.4.9 Concluding remarks

Numerous studies have documented the effects of 2APB and its analogs on membrane channels. However the mechanisms by which 2APB regulate ion channels remain a mystery. New evidence suggests that the action of 2APB on TRP channels is not universal. While several TRP channels are inhibited, at least three of them, TRPV1-

3, are stimulated by 2APB. Some TRP channels are unaffected by 2APB and many more remain to be tested. The findings that 2APB activates TRPV1-3, while its analog DPTHF shows some selectivity for V3 over V1 and V2, make it promising that specific ligands may be made for V2 and V3 through modification of various 2APB analogs. More specific drugs would certainly accelerate the discovery of the physiological functions and mechanisms of regulation of these amazing thermosensitive channels.

198

A

capsaicin eugenolguaiacol resiniferatoxin

B

12-HPETE camphor isovelleral

ginsenosides evodiamine polygodial

B B O B O

H2N 2APB DPBA

Fig 6.1. Chemical activators of TRPV1. A, capsaicinoid compounds each contain a vanilloid functional group. B, A selection of noncapsaicinoid compounds that activate V1 and otherwise are unrelated structurally.

199

A

B B O O B

H2N 2 APB DPBA

B

4a-PDD EESA BAA

C

mentholgeranoil eugenol WS-23 icilin

D CH3 OH

CH3 CH3 allyl isothiocyanate eugenol cinnamaldehyde carvacrol

Fig. 6.2. Chemical activators of various thermoTRP channels. A, Chemical activators of V2. To date only boron containing compounds have been shown to activate V2. B, Chemical activators of V4. These are the only known chemical activators of V4. The compounds are structurally unrelated except they are all large lipophilic molecules. C, Chemical activators of M8. D, Chemical activators of A1. For C&D, only a selection of known chemical activators was chosen in order to display the structural diversity of chemical activators for these receptors.

200

A

B B O NDGA O B CH2CH3 O HO H N OH 2 CH CH 3 2 2 APB DPBA DPTH DES

B CH3 CH3 OH

CH3 CH3 CH3 CH3 thymol carvacrol ethyl vanillin vanillin C D

menthol camphor eugenol capsaicin E

arachidonic acid ETYA

F O HN O HN CH CH3 3 N N O U73122 U73433 O O O O O CH3 CH3 H3C(H2C)2 O O (CH ) CH H C 2 2 3 3 H CH 3 H3C CH3 HO H H O HO OH PDBU

Fig. 6.3. Chemical activators of TRPV3. A, 2APB and its analogues. DPTH resembles the closed ring form of 2APB. NDGA and DES resemble 2APB in that they have two phenyl groups in close proximity that are not constrained within a plain. B, Small derivatives of phenol. C, Small lipophilic, non aromatic molecules with a polar functional group D, Capsaicinoid molecules. E, Polyunsaturated fatty acids. Both double and triple bonded fatty acids with the first nonsaturated carbon located at the fifth carbon from the carboxylate group. F, Large lipophilic molecules. U73122 and U73433 only differ by one double bond in the nitrogen containing ring. A comparison of groups (A-F) demonstrates a great deal of variability in the structures of molecules capable of activating V3.

201

BA U73122 9 8 [U73122] 50 µM ) 7 -4 30 µM 6 20 µM 5 10 µM 5 µM (X 10 4 o 0 µM 3

F - F 2

1 0 012345

Time (min)

CB U73433 14 12 [U73433] 50 µM ) -4 10 30 µM 20 µM 8 10 µM 5 µM (X 10

o 6 0 µM

4 F - F

2

0 012345 Time (min)

Fig. 6.4. Activation of TRPV3D641N by U73122 and U73433. A, dose-dependent activation of V3D641N by U73122. B, dose-dependent activation of V3D641N by U73433. U73433 is the inactive analogue of U73122, demonstrating that activation is independent of PLC. For (A&B), Fluo4-loaded cells were stimulated with different concentrations of the specified compound at 20 sec. Data represent the average of two repeat experiments.

202

A O O O O O O O H3C(H2C)2 O O H3C(H2C)8 O O H3C(H2C)12 CH3 O (CH2)2CH3 CH (CH2)8CH3 H3C 3 H3C H3C CH3 CH3 H H H CH3 H3C H3C CH3 H CH3 HO H H C H HO H H 3 HO H O HO O HO HO O OH OH OH

PDBu PMA PDD

B Phorbol ester 8

) 6

) -4 4

(X 10 (X 10 4 V3 D641N 10 µM PDD o o V3 D641N 10 µM PMA

F-F V3 D641N 16.7 µM PDBU F - F 2 pcDNA 16.7 µM PDBU

0 0 123

Time (min)

Fig. 6.5. Activation of TRPV3D641N by phorbol esters. A, structure of the different phorbol esters tested. B, PDBU, but not PDD or PMA activates V3D641N. For (B), fluo4- loaded cells were stimulated with PDBU at 20 sec. Traces represent the average of two repeat experiments.

203

A Menthol 22 20 18 16 ) -4 14 3 mM 12

(X 10 1 mM o 10 333 µM 8 F - F 6

4

2

0 012345 Time (min)

B 2mM menthol 13 12 11 10 V3 )

-4 9 8 7

(X 10 6 o 5

F - F 4 V3D641N 3 2 1 pcDNA3 0 0 20 40 60 80 100 120 140 Time (min)

Fig. 6.6. Activation of TRPV3 by menthol. A, dose-dependent activation of V3D641N by menthol. pcDNA3 transfected cells did not respond to menthol at any of these concentrations, therefore the traces from pcDNA3 transfected cells were subtracted from the data shown. B, 2 mM menthol activates both V3 and V3D641N. For A&B, Fluo4- loaded cells were stimulated with different concentrations of the specified compound at 20 sec. Data represent the average of two repeat experiments.

204

A 2) 1) 12

10 )

-4 8 1) 300 µM DES 2) 20 µM 2APB 1) 30 µM DES 2) 20 µM 2APB

(X 10 6 o 1) 3 µM DES 2) 20 µM 2APB 4 1) 20 µM 2APB only F - F 2 1) 166 µM 2APB only

0 012345 Time (min) B CH CH 2 3 HO OH CH CH 3 2 DES trans-stilbene DES

B phenytoin

2APB bendadryl

Fig. 6.7. Activation of TRPV3D641N by DES. A, dose-dependent enhancement of V3D641N by DES. B, structure of the different 2APB analogs tested. Only DES potentiated the 2APB response. For (A), fluo4-loaded cells transiently expressing V3D641N were stimulated with different concentrations of DES at 20 sec followed by 20 µM 2APB at 140 sec. For comparison, cells were stimulated with 20 µM 2APB-only and with 166 µM 2APB-only to represent maximal activation. Data represent the average of two repeat experiments with the pcDNA data subtracted from the V3D641N data.

205

B B B H2N O O O

+H N H N 3 2 2APB monomer 2APB monomer 2APB monomer ring protonated

B B B H2N O N O O B

dimethyl 2APB

O O

O NH2 B diphenylboronic anhydride

2APB dimer O

O N

2,2-diphenyltetrahydrofuran diphenhydramine

Fig. 6.8. Various forms of 2APB and several 2APB analogs. The nitrogen of the ethanolamine side chain on the 2APB monomer can become protonated (2APB monomer protonated) or form coordinate bonds with either an internal boron (2APB monomer ring), or the boron on another 2APB molecule (2APB dimer). Most data support the 2APB monomer ring as the predominant form of 2APB. The boron containing 2APB analog dipenylboronic anhydride (DPBA) cannot be protonated. The non-boron containing analog 2,2-diphenyltetrahydrofuran (DPTHF) is structurally related to the 2APB monomer ring. Diphenhydramine (Benadryl) is a non-boron containing antihistamine that is structurally related to the 2APB monomer with the exception that 2APB has a primary amine and diphenhydramine has a tertiary amine. Diphenhydramine is also structurally related to dimethyl 2APB with the exception that dimethyl 2APB should exist predominantly in the ring form, whereas diphenhydramine is unable to form a ring and could be protonated. All of these molecules have a tetrahedral geometry at the equivalent position to the boron of 2APB.

206

1

4 2 3

4 5 6 1

2 2 6 5 1 4

T T T T

T T T

T

A A A

A A C A A A A N C N

5

Fig. 6.9. Possible binding site or sites of action for 2APB and its analogs on TRP channels. Potential sites of action include: 1) extracellular binding sites, either in a hydrophobic pocket (typical ligand-protein interaction), or via formation of a coordinate bond between amino acid side chains and the boron of 2APB. 2) The pore, either by electrostatic interaction between the protonated form of 2APB and negatively charged amino acids, or by formation of a coordinate bond between amino acid side chains (most likely aspartate or glutamate) that line the pore and the boron of 2APB. 3) Disruption of lipid-protein interactions within the membrane, of which negatively charge phospholipids such as PIP2 are prime candidates. 4) Accumulation of 2APB either at the membrane- water interface or the membrane interior, resulting in a change in the physical properties of the membrane and a subsequent conformational change of the channel. This mechanism would be similar to that of anesthetics. 5) Intracellular sites of action are unlikely, but in certain cell types, signaling components required for TRP activation may be targeted (see text). Numbers in and above columns depict transmembrane segments. A and T indicate ankyrin-like and TRP motifs, respectively. Larger bolded numbers indicate the possible sites of action described above.

207

Fig. 6.10. Ways in which the membrane can affect membrane protein conformation. A, Dipicts the effect of increasing the temperature on the membrane. As the disorder in the fatty acid tails of phospholipids increase upon increasing the temperature, the membrane will become thinner in order to maintain an equal number of Van der Waals contacts. Stretching the membrane laterally will have similiar effects. B, The conformation of a transmembrane protein is influenced by the thickness of the membrane and vice versa. Based on several conditions, a membrane protein will assume a conformation that minimizes the energetic cost of exposing hydrophobic amino acids when the thickness of the membrane changes. If the energetic drive (for example caused from ligand binding) for a protein to assume a conformation the protein thickness does not correlate well with the membrane thickness, the membrane will compensate as shown. This figure shows, that the conformation of a particular transmembrane protein is dependent on energy consideration from both the membrane and the protein itself. In this illustration, 1, 2 and 3 represent different confirmations, with any one of them possibly being an open or closed confirmation. This picture was recreated based on Hamill & Martinac (2001). C, Variuos thermodynamic contraints, especially at the lipid-water interface create lateral pressures within the membrane that vary at different depths. These presures in effect push on the transmembrane domains of the channel protein, influncing the conformation and orientation of various transmembrane domains. This illustration was recreated based on Kung (2005).

208

lower higher

A temperature temperature

B 1 2

3

C

Figure 6.10

209

Fig. 6.11. The effect of dissolving lipohilic molecules into phospholipid bilayers. A, Cholesterol has a stabilizing effect on phospholipid bilayers. The polar hydroxyl group of cholesterol acts as a head-group, causing the orientation of cholesterol shown in the figure on the left and right. Left shows different molecular models of cholesterol including the 3-D representation. Right shows a 2-D representation of cholesterol being inserted into the membrane. A more detailed discussion is in the text. The left figure was taken from Alberts et al. (1994). B, The “shape” of phospholipids (cylindrical) and lysophospholipids (cone-shaped) is depicted on the left. Lysophospolipids have been shown to activate the thermosensitive channel TREK-1. Right shows several lysophospolipids that were tested by Patel et al. (2001). Interestingly, lysophospholipids with larger ratios (in the lateral direction) of head group to hydrocarbon tail were more potent, demonstrating that these molecules work by affecting the membrane. LPE: lysophosphatidylethanolamide; LPS: lysophosphatidylserine; LPC: lysophosphatidylcholine; LPI; lysophosphatidylinositol C, Illustration of the shapes of molecules in the membrane. Left is a bilayer made of pure phospholipids. Middle is the bilayer with lysophospholipids dissolved in it. Right represents a bilayer with small lipophilic molecules such as carvacrol, camphor, menthol and possibly 2APB dissolved in it. Note that in (A), the shape of cholesterol is mostly cylindrical and should stabilize the membrane. Lysophospholipids and small lipophilic molecules will leave gaps in the extended network of Van der Waals contacts, causing a restructuring of the membrane to compensate. D, Arachidonic acid (AA) and other crenators activate TREK-1 by accumulating at the lipid-water interface and disrupting the normal lipid-protein interactions. Note AA also causes negative charges to accumulate at the outer leaflet. This illustration was recreated based on Patel et al (2001).

210 Figure 6.11

A

B phos pholipid lysophospholipid LPE LPS LPC LPI

Potency toward : ++++++++++ C TREK-1 Phospholipid-only LPC carvacrol

D gaps

AA

211

compoundV1 V2V3 V4 M8 A1

+ n - n - o capsaicin

eugenol + n + n + +

+ n n n + + cinnamaldehyde

B O + + + n - o H N 2 APB 2 CH3 OH n n + n o + CH3 CH3 carvacrol

n n + n + n menthol + n + n - - camphor

n n n n + n icilin

allyl isothiocyanate n n n n n +

n n n + n n

BAA O O O H3C(H2C)R2 O (CHR 2)2CH3 H3C H CH3

H3C CH3 H + + + o o o HO H O HO OH phorbol esters + + + + n + PUFAs

Table 6.1. Promiscuity of molecules shown to modulate the thermoTRPs. Only a small selection of molecules that activate these channels were chosen for comparison. Some molecules are specific, while others are very promiscuous. +: activates the channel; -: inhibits the channel; n: no effect on the channel; o: the effect of these compounds is not reported.

212

BIBLIOGRAPHY

Albert B, Bray D, Lewis J, Raff M, Roberts K, Watson JD. (1994) Molecular Biology of The Cell, 3rd ed. Garland Publishing, Inc, New York & London.

Antkowiak B. (2001) How do general anaesthetics work? Naturwissenschaften. 88(5): 201-13.

Asakawa M, Yoshioka T, Matsutani T, Hikita I, Suzuki M, Oshima I, Tsukahara K, Arimura A, Horikawa T, Hirasawa T, Sakata T. (2006) Association of a Mutation in TRPV3 with Defective Hair Growth in Rodents. J Invest Dermatol. Epub ahead of print.

Bandell M, Story GM, Hwang SW, Viswanath V, Eid SR, Petrus MJ, Earley TJ, Patapoutian A. (2004) Noxious cold ion channel TRPA1 is activated by pungent compounds and bradykinin. 41(6): 849-57.

Bandell M, Dubin AE, Petrus MJ, Orth A, Mathur J, Hwang SW, Patapoutian A. (2006) High-throughput random mutagenesis screen reveals TRPM8 residues specifically required for activation by menthol. Nat Neurosci. 9(4): 493-500.

Bautista DM, Movahed P, Hinman A, Axelsson HE, Sterner O, Hogestatt ED, Julius D, Jordt SE, Zygmunt PM. (2005) Pungent products from garlic activate the sensory ion channel TRPA1. Proc Natl Acad Sci U S A. 102(34): 12248-52.

Bender F, Mederos Y, Schnitzler M, Li Y, Ji A, Weihe E, Gudermann T, Schafer M (2005) The temperature-sensitive ion channel TRPV2 is endogenously expressed and functional in the primary sensory cell line F-11. Cell Physiol Biochem. 15: 183-94.

Birder LA, Nakamura Y, Kiss S, Nealen ML, Barrick S, Kanai AJ, Wang E, Ruiz G, De Groat WC, Apodaca G, Watkins S, Caterina MJ. (2002) Altered urinary bladder function in mice lacking the vanilloid receptor TRPV1. Nat Neurosci. 5(9): 856-60.

Bhave G, Zhu W, Wang H, Brasier DJ, Oxford GS, Gereau IV RW (2002) cAMP- dependant protein kinase regulates desensitization of the Capsaicin Receptor (VR1) by direct phosphorylation. Neuron. 35: 721-31.

213 Bhave G, Hu HJ, Glauner KS, Zhu W, Wang H, Brasier DJ, Oxford GS, Gereau IV RW. (2003) Protein kinase C phosphorylation sensitizes but does not activate the capsaicin receptor transient receptor potential vanilloid 1. (TRPV1). Proc Natl Acad Sci U S A. 100(21): 12480-5.

Bootman MD, Collins TJ, Mackenzie L, Roderick HL, Berridge MJ, Peppiatt CM (2002) 2-aminoethoxydiphenyl borate (2-APB) is a reliable blocker of store-operated Ca2+ 2+ entry but an inconsistent inhibitor of InsP3-induced Ca release. FASEB J 16: 1145- 50.

Brash AR. (2001) Arachidonic acid as a bioactive molecule. J Clin Invest. 107: 1339-45.

Calixto JB, Kassuya CA, Andre E, Ferreira J. (2005) Contribution of natural products to the discovery of the transient receptor potential (TRP) channels family and their functions. Pharmacol Ther. 106(2): 179-208.

Campagna JA, Miller KW, Forman SA. (2003) Mechanisms of actions of inhaled anesthetics. N Engl J Med. 349(9): 909-10.

Carr MJ, Kollarik M, Meeker SN, Undem BJ. A role for TRPV1 in bradykinin-induced excitation of vagal airway afferent nerve terminals. J Pharmacol Exp Ther. 304(3): 1275-9.

Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, Julius D. (1997) The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature. 389(6653): 783-4.

Caterina M, Julius D. (1999) Sense and specificity: a molecular identity for nociceptors. Current Opinion in Neurobiology. 9: 525-30.

Caterina M, Leffler A, Malmberg A, Martin W, Trafton J, Peterson-Zeitz K, Koltzenburg M, Basbaum A, Julius D. (2000) Impaired nociception and pain sensation in mice lacking the capsaicin receptor. Science. 288: 306-13.

Cesare P, Moriondo A, Vellani V, McNaughton PA. (1999) Ion channels gated by heat. Proc Natl Acad Sci U S A. 96(14): 7658-63.

Chu C, Huang S, De Petrocellis L, Bisogno T, Ewing S, Miller J, Zipkin R, Daddario N, Appendino G, Di Marzo V, Walker J. (2003) N-oleoyldopamine, a novel endogenous capsaicin-like lipid that produces hyperalgesia. J Biol Chem. 278. 13633-39.

Chuang HH, Prescott ED, Kong H, Shields S, Jordt SE, Basbaum AI, Chao MV, Julius D. (2001) Bradykinin and nerve growth factor release the capsaicin receptor from PtdIns(4,5)P2-mediated inhibition. Nature. 411(6840): 957-62.

214 Chung MK, Lee H, Mizuno A, Suzuki M, Caterina MJ (2004a) 2-aminoethoxydiphenyl borate activates and sensitizes the heat-gated ion channel TRPV3. J Neurosci. 24: 5177-82.

Chung MK, Lee H, Mizuno A, Suzuki M, Caterina MJ (2004b) TRPV3 and TRPV4 mediate warmth-evoked currents in primary mouse keratinocytes. J Biol Chem. 279: 21569-75.

Chung MK, Guler AD, Caterina MJ (2005) Biphasic currents evoked by chemical or thermal activation of the heat-gated ion channel, TRPV3. J Biol Chem. 280: 15928- 41.

Chyb S, Raghu P, Hardie RC. (1999) Polyunsaturated fatty acids activate the Drosophila light-sensitive channels TRP and TRPL. Nature. 397(6716): 255-9.

Clapham DE, Runnels LW, Strubing C (2001) The TRP ion channel family. Nat Rev Neurosci. 2: 387-96.

Clapham DE. (2003) TRP channels as cellular sensors. Nature. 426(6966): 517-24.

Colledge M, Dean R, Scott G, Langeberg L, Huganir R, and Scott J. (2000) Targeting of PKA to Glutamate Receptors through a MAGUK-AKAP complex. Neuron. 27: 107- 19.

Coutaux A, Adam F, Willer JC, Le Bars D. (2005) Hyperalgesia and allodynia: peripheral mechanisms. Joint Bone Spine. 72(5): 359-71.

Craig AD. (2003) Pain mechanisms: labeled lines versus convergence in central processing. Annu Rev Neurosci. 26: 1-30.

Danthi S, Enyeart JA, Enyeart JJ. (2003) Modulation of native TREK-1 and Kv1.4 K+ channels by polyunsaturated fatty acids and lysophospholipids. J Membr Biol. 195(3): 147-64.

Davis J, Gray J, Gunthorpe M, Hatcher J, Davey P, Overend P, Harries M, Latcham J, Clapham C, Atkinson K, Hughes S, Rance K, Grau E, Harper A, Pugh P, Rogers D, Bingham S, Randall A, Sheardown S. (2000) Vanilloid receptor-1 is essential for inflammatory thermal hyperalgesia. Nature. 405: 183-7.

Delmas P, Wanaverbecq N, Abogadie FC, Mistry M, Brown DA (2002) Signaling microdomains define the specificity of receptor-mediated InsP(3) pathways in neurons. Neuron. 34: 209-20.

Deol SS, Bond PJ, Domene C, Sansom MS. (2004) Lipid-protein interactions of integral membrane proteins: a comparative simulation study. Biophys J. 87(6): 3737-49.

215

Diver JM, Sage SO, Rosado JA. (2001) The inositol trisphosphate receptor antagonist 2- aminoethoxydiphenylborate (2-APB) blocks Ca2+ entry channels in human platelets: cautions for its use in studying Ca2+ influx. Cell Calcium. 30: 323-9.

Dobrydneva Y, Blackmore P (2001) 2-Aminoethoxydiphenyl borate directly inhibits store-operated calcium entry channels in human platelets. Mol Pharmacol. 60: 541- 52.

Dobrydneva Y, Abelt CJ, Dovel B, Thadigiri CM, Williams RL, Blackmore PF (2006) 2- Aminoethoxydiphenyl borate as a prototype drug for a group of structurally related calcium channel blockers in human platelets. Mol Pharmacol. 69: 247-56.

Domene C, Bond PJ, Deol SS, Sansom MS. (2003) Lipid/protein interactions and the membrane/water interfacial region. J Am Chem Soc. 125(49): 14966-7.

Domene C, Vemparala S, Klein ML, Venien-Bryan C, Doyle DA. (2005) Role of aromatic localization in the gating process of a potassium channel. Biophys J. 90(1): L01-3.

Dougherty DA. Cation-pi interactions in chemistry and biology: a new view of benzene, Phe, Tyr, and Trp. Science. 271(5246): 163-8.

Estacion M, Sinkins WG, Schilling WP. (2001) Regulation of Drosophila transient receptor potential-like (TrpL) channels by phospholipase C-dependent mechanisms. J Physiol. 530(Pt 1): 1-19.

Fanning A, Anderson J. (1999) PDZ domains: fundamental building blocks in the organization of protein complexes at the plasma membrane. Journal of Clinical Investigation. 103: 767-72.

Fersht A. (1985) Enzyme Structure and Mechanism (2nd ed.). W. H. Freeman and Company, New York.

Fink M, Lesage F, Duprat F, Heurteaux C, Reyes R, Fosset M, Lazdunski M. (1998) A neuronal two P domain K+ channel stimulated by arachidonic acid and polyunsaturated fatty acids. EMBO J. 17(12): 3297-308.

Gaede HC, Yau WM, Gawrisch K. (2005) Electrostatic contributions to indole-lipid interactions. J Phys Chem B Condens Matter Mater Surf Interfaces Biophys. 109(26): 13014-23.

Gallegos AM, Storey SM, Kier AB, Schroeder F, Ball JM. (2006) Structure and cholesterol dynamics of caveolae/raft and nonraft plasma membrane domains. Biochemistry. 45(39): 12100-16.

216

Gao X, Wu L, O'Neil RG. (2003) Temperature-modulated diversity of TRPV4 channel gating: activation by physical stresses and phorbol ester derivatives through protein kinase C-dependent and -independent pathways. J Biol Chem. 278(29): 27129-37.

Garcia-Martinez C, Morenilla-Palao C, Planells-Cases R, Merino JM, Ferrer-Montiel A. (2000) Identification of an aspartic residue in the P-loop of the Vanilloid Receptor that modulates pore properties. J. Biol. Chem. 275: 32552-8.

Gopinath P, Wan E, Holdcroft A, Facer P, Davis JB, Smith GD, Bountra C, Anand P. (2005) Increased capsaicin receptor TRPV1 in skin nerve fibres and related vanilloid receptors TRPV3 and TRPV4 in keratinocytes in human breast pain. BMC Womens Health. 5(1): 2.

Gottschalk A, Smith DS. (2001) New concepts in acute pain therapy: preemptive analgesia. Am Fam Physician. 63(10): 1924.

Griffin J. (2004) Central thermosensitivity and the integrative response of hypothalamic neurons. Journal of Thermal Biology. 29: 327-31.

Groschner K, Hingel S, Lintschinger B, Balzer M, Romanin C, Zhu X, Schreibmayer W. (1998) Trp proteins form store-operated cation channels in human vascular endothelial cells. FEBS Lett. 437(1-2): 101-6.

Guler AD, Lee H, Iida T, Shimizu I, Tominaga M, Caterina M. (2002) Heat-evoked activation of the ion channel, TRPV4. J Neurosci. 22(15): 6408-14.

Gunthorpe MJ, Benham CD, Randall A, Davis JB. (2002) The diversity in the vanilloid (TRPV) receptor family of ion channels. Trends Pharmacol Sci. 23(4): 183-91.

Gu Q, Lin RL, Hu HZ, Zhu MX, Lee LY (2005) 2-aminoethoxydiphenyl borate stimulates pulmonary C neurons via the activation of TRPV channels. Am J Physiol Lung Cell Mol Physiol. 288: L932-41.

Hamill OP, Martinac B (2001) Molecular basis of mechanotransduction in living cells. Physiol Rev. 81: 685-740.

Hammarstrom S, Hamberg M, Samuelsson B, Duell EA, Stawiski M, Voorhees JJ. (1975) Increased concentrations of nonesterified arachidonic acid, 12L-hydroxy- 5,8,10,14-eicosatetraenoic acid, prostaglandin E2, and prostaglandin F2alpha in epidermis of psoriasis. Proc Natl Acad Sci U S A. 72: 5130-4.

Hanano T, Hara Y, Shi J, Morita H, Umebayashi C, Mori E, Sumimoto H, Ito Y, Mori Y, Inoue R. (2004) Involvement of TRPM7 in cell growth as a spontaneously activated Ca2+ entry pathway in human retinoblastoma cells. J Pharmacol Sci. 95: 403-4.

217

Hardie RC. (2006) TRP channels and lipids: from Drosophila to mammalian physiology. J Physiol. Epub ahead of print.

Hellwig N, Albrecht N, Harteneck C, Schultz G, Schaefer M. (2005) Homo- and heteromeric assembly of TRPV channel subunits. J Cell Sci. 118(5): 917-28.

Hesselager M, Timmermann DB, Ahring PK. (2004) pH Dependency and desensitization kinetics of heterologously expressed combinations of acid-sensing ion channel subunits. J Biol Chem. 279(12): 11006-15.

Heurteaux C, Guy N, Laigle C, Blondeau N, Duprat F, Mazzuca M, Lang-Lazdunski L, Widmann C, Zanzouri M, Romey G, Lazdunski M (2004) TREK-1, a K+ channel involved in neuroprotection and general anesthesia. EMBO J. 23: 2684-95.

Hope HR, Pike LJ. (1996) Phosphoinositides and phosphoinositide-utilizing enzymes in detergent-insoluble lipid domains. Mol Biol Cell. 7(6): 843-51.

Huang S, Bisogno T, Trevisani M, Al-Hayani A, De Petrocellis L, Fezza F, Tognetto M, Petros T, Krey J, Chu C, Miller J, Davies S, Geppetti P, Walker J, Di Marzo V. (2002) An endogenous capsaicin-like substance with high potency at recombinant and native vanilloid VR1 receptors. Proc Natl Acad Sci U S A. 99: 8400-5.

Helms JB, Zurzolo C. (2004) Lipids as targeting signals: lipid rafts and intracellular trafficking. Traffic. 5(4): 247-54.

Huber A. (2001) Scaffolding proteins organize multimolecular protein complexes for sensory signal transduction. European Journal of Neuroscience. 14: 769-76.

Hu H, Bhave G, Gereau IV R. (2002) Prostaglandin and Protein Kinase A-dependent modulation of vanilloid receptor function by metabotropic glutamate receptor 5: potential mechanism for thermal hyperalgesia. The Journal of Neuroscience. 22: 7444-52.

Hu HZ, Gu Q, Wang C, Colton CK, Tang J, Kinoshita-Kawada M, Lee LY, Wood JD, Zhu MX (2004) 2-aminoethoxydiphenyl borate is a common activator of TRPV1, TRPV2, and TRPV3. J Biol Chem. 279(34): 35741-8.

Hu HZ, Xiao R, Wang C, Gao N, Colton CK, Wood JD, Zhu MX. (2006) Potentiation of TRPV3 channel function by unsaturated fatty acids. J Cell Physiol. 208(1): 201-12.

Hur EM, Park YS, Lee BD, Jang IH, Kim HS, Kim TD, Suh PG, Ryu SH, Kim KT. (2004) Sensitization of epidermal growth factor-induced signaling by bradykinin is mediated by c-Src. Implications for a role of lipid microdomains. J Biol Chem. 279(7): 5852-60.

218

Hwang S, Cho H, Kwak J, Lee S, Kang C, Jung J, Cho S, Min K, Suh Y, Kim D. (2000) Direct activation of capsaicin receptors by products of lipoxygenases: endogenous capsaicin-like substances. Proc Natl Acad Sci U S A. 97: 6155-60.

Hwang S, Oh U. (2002) Hot channels in airways: pharmacology of the vanilloid receptor. Current Opinion in Pharmacology. 2: 235-42.

Issberner U, Reeh PW, Steen KH. (1996) Pain due to tissue acidosis: a mechanism for inflammatory and ischemic myalgia? Neurosci Lett. 208(3): 191-4.

Jones NG, Slater R, Cadiou H, McNaughton P, McMahon SB. (2004) Acid-induced pain and its modulation in humans. J Neurosci. 24(48): 10974-9.

Jung S, Muhle A, Schaefer M, Strotmann R, Schultz G, Plant TD (2003) Lanthanides potentiate TRPC5 currents by an action at extracellular sites close to the pore mouth. J Biol Chem. 278: 3562-71.

Liu B, Hui K, Qin F. (2003) Thermodynamics of heat activation of single capsaicin ion channels VR1. Biophys J. 85(5): 2988-3006.

Jordt SE, Tominaga M, Julius D. (2000) Acid potentiation of the capsaicin receptor determined by a key extracellular site. Proc Natl Acad Sci U S A. 97(14): 8134-9.

Jordt SE, Julius D. (2002) Molecular basis for species-specific sensitivity to "hot" chili peppers. Cell. 108(3): 421-30.

Jordt SE, Bautista DM, Chuang HH, McKemy DD, Zygmunt PM, Hogestatt ED, Meng ID, Julius D. (2004) Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTM1. Nature. 427(6971): 260-5.

Kahn-Kirby AH, Dantzker JL, Apicella AJ, Schafer WR, Browse J, Bargmann CI, Watts JL. (2004) Specific polyunsaturated fatty acids drive TRPV-dependent sensory signaling in vivo. Cell. 119: 889-900.

Kim E, Neithammer M, Rothschild A, Jan Y, Sheng M. (1995) Clustering of Shaker- type K+ channels by interaction with a family of membrane-associated guanylate kinases. Nature. 378: 85-8.

Koizumi H, Yasui C, Fukaya T, Ueda T, Ohkawara A. (1994) Substance P induces inositol 1,4,5-trisphosphate and intracellular free calcium increase in cultured normal human epidermal keratinocytes. Exp Dermatol. 3(1): 40-4.

219 Koulen P, Cai Y, Geng L, Maeda Y, Nishimura S, Witzgall R, Ehrlich BE, Somlo S. (2002) Polycystin-2 is an intracellular calcium release channel. Nat Cell Biol. 4:191- 7.

Kress M, Zeilhofer HU. (1999) Capsaicin, protons and heat: new excitement about nociceptors. Trends Pharmacol Sci. 20(3): 112-8.

Kung C (2005) A possible unifying principle for mechanosensation. Nature 436: 647-54.

Lee Y, Lee CH, Oh U. (2005) Painful channels in sensory neurons. Mol Cells. 20(3): 315-24.

Liedtke W, Choe Y, Marti-Renom M, Bell A, Denis C, Sali A, Hudspeth A, Friedman J, Heller S. (2000) Vanilloid Receptor-related Osmotically Activated Channel (VR- OAC), a candidate vertebrate osmosensor. Cell. 103: 525-35.

Liedtke W. (2005) TRPV4 as osmosensor: a transgenic approach. Pflugers Arch. 451(1): 176-80.

Lievremont JP, Bird GS, Putney JW Jr. (2005) Mechanism of inhibition of TRPC cation channels by 2-aminoethoxydiphenylborane. Mol Pharmacol. 68: 758-62.

Liu B, Hui K, Qin F (2003) Thermodynamics of heat activation of single capsaicin ion channels VR1. Biophys J. 85: 2988-3006.

Liu B, Qin F. (2005) Functional control of cold- and menthol-sensitive TRPM8 ion channels by phosphatidylinositol 4,5-bisphosphate. J Neurosci. 25(7): 1674-81.

Liu D, Liman ER. (2003) Intracellular Ca2+ and the phospholipid PIP2 regulate the taste transduction ion channel TRPM5. Proc Natl Acad Sci U S A. 100(25): 15160-5.

Ma HT, Patterson RL, van Rossum DB, Birnbaumer L, Mikoshiba K, Gill DL (2000) Requirement of the inositol trisphosphate receptor for activation of store-operated Ca2+ channels. Science. 287: 1647-51.

Ma HT, Venkatachalam K, Rys-Sikora KE, He LP, Zheng F, Gill DL (2003) Modification of phospholipase C-gamma-induced Ca2+ signal generation by 2- aminoethoxydiphenyl borate. Biochem J. 376: 667-76.

Macpherson LJ, Hwang SW, Miyamoto T, Dubin AE, Patapoutian A, Story GM. (2006) More than cool: promiscuous relationships of menthol and other sensory compounds. Mol Cell Neurosci. 32(4): 335-43.

220 Maingret F, Patel AJ, Lesage F, Lazdunski M, Honore E. (2000) Lysophospholipids open the two-pore domain mechano-gated K+ channels TREK-1 and TRAAK. J Biol Chem. 275(14): 10128-33.

Maruyama T, Kanaji T, Nakade S, Kanno T, Mikoshiba K (1997) 2APB, 2- aminoethoxydiphenyl borate, a membrane-penetrable modulator of Ins(1,4,5)P3- induced Ca2+ release. J Biochem (Tokyo). 122: 498-505.

McIntyre P, McLatchie LM, Chambers A, Phillips E, Clarke M, Savidge J, Toms C, Peacock M, Shah K, Winter J, Weerasakera N, Webb M, Rang HP, Bevan S, James IF (2001) Pharmacological differences between the human and rat vanilloid receptor 1 (VR1). Br J Pharmacol. 132: 1084-94.

McGee A, Bredt D. (1999) Identification of an Intramolecular Interaction between the SH3 and Guanylate Kinase Domains of PSD-95. JBC. 274: 17431-36.

McGinty D, Szymusiak R. (2000) The sleep-wake switch: A neuronal alarm clock. Nat Med. 6(5): 510-1.

Mecozzi S, West AP Jr, Dougherty DA. (1996) Cation-pi interactions in aromatics of biological and medicinal interest: electrostatic potential surfaces as a useful qualitative guide. Proc Natl Acad Sci U S A. 93(20): 10566-71.

Montell C. (2005a) TRP channels in Drosophila photoreceptor cells. J. Phsiol. 567(1): 45-51.

Montell C. (2005b) Drosophila TRP channels. Pflugers Arch. 451(1): 19-28.

Moqrich A, Hwang SW, Earley TJ, Petrus MJ, Murray AN, Spencer KS, Andahazy M, Story GM, Patapoutian A (2005) Impaired thermosensation in mice lacking TRPV3, a heat and camphor sensor in the skin. Science. 307: 1468-72.

Moran MM, Xu H, Clapham DE. (2004) TRP ion channels in the nervous system. Curr Opin Neurobiol. 14(3): 362-9.

Nilius B, Prenen J, Vennekens R, Hoenderop JG, Bindels RJ, Droogmans G (2001) Pharmacological modulation of monovalent cation currents through the epithelial Ca2+ channel ECaC1. Br J Pharmacol. 134: 453-62.

Nilius B, Voets T. (2005a) TRP channels: a TR(I)P through a world of multifunctional cation channels. Pflugers Arch. 451(1): 1-10.

Nilius B, Talavera K, Owsianik G, Prenen J, Droogmans G, Voets T. (2005b) Gating of TRP channels: a voltage connection? J Physiol. 567(Pt 1): 35-44.

221 Nix S, Chishti A, Anderson J, Walther Z. (2000) hCASK and hDLG associate in epithelia, and their Src homology 3 and guanylate kinase domains participate in both intramolecular and intermolecular interactions. JBC. 275: 41192-200.

North C, Cafiso DS (1997) Contrasting membrane localization and behavior of halogenated cyclobutanes that follow or violate the Meyer-Overton hypothesis of general anesthetic potency. Biophys J. 72: 1754-61.

Nöth H (1970) Some recent developments in boron-nitrogen chemistry. In Brotherton RJ, Steinberg H (eds) Progress in Boron Chemistry. Pergamon Press, New York, pp 211- 311.

Numazaki M, Tominaga T, Toyooka H, Tominaga M. (2002) Direct phosphorylation of Capsaicin Receptor VR1 by Protein Kinase Cε and identification of two target serine residues. JBC. 277: 13375-8.

O'Shea P. (2003) Intermolecular interactions with/within cell membranes and the trinity of membrane potentials: kinetics and imaging. Biochem Soc Trans. 31(Pt 5): 990-6.

Parekh AB, Putney JW Jr. (2005) Store-operated calcium channels. Physiol Rev 85:757- 810.

Park KA, Vasko MR. (2005) Lipid mediators of sensitivity in sensory neurons. Trends Pharmacol Sci. 26(11): 571-7.

Patapoutian A, Peier A, Story G, Viswanath V. (2003) ThermoTRP channels and beyond: mechanisms of temperature sensation. Nature Rev. 4: 529-39.

Patel AJ, Lazdunski M, Honore E (2001) Lipid and mechano-gated 2P domain K+ channels. Curr Opin Cell Biol. 13: 422-8.

Peier A, Reeve A, Anderson D, Moqrich A, Earley T, Hergarden A, Story G, Colley S, Hogenesch J, Mcintyre P, Bevan S, Patapoutian A. (2002) A heat-sensitive TRP channel expressed in Keratinocytes. Science. 296: 2046-9.

Planells-Cases R, Garcia-Sanz N, Morenilla-Palao C, Ferrer-Montiel A. (2005) Functional aspects and mechanisms of TRPV1 involvement in neurogenic inflammation that leads to thermal hyperalgesia. Pflugers Arch. 451(1): 151-9.

Prakriya M, Lewis RS. (2001) Potentiation and inhibition of Ca2+ release-activated Ca2+ channels by 2-aminoethyldiphenyl borate (2-APB) occurs independently of IP3 receptors. J Physiol. 536: 3-19.

Premkumar L, Ahern G. (2000) Induction of vanilloid receptor channel activity by protein kinase C. Nature. 408: 985-90.

222

Prescott ED, Julius D. (2003) A modular PIP2 binding site as a determinant of capsaicin receptor sensitivity. Science 300: 1284-8.

Rathee PK, Distler C, Obreja O, Neuhuber W, Wang GK, Wang SY, Nau C, Kress M. (2002) PKA/AKAP/VR-1 module: A common link of Gs-mediated signaling to thermal hyperalgesia. J Neurosci. 22(11): 4740-5.

Rettig SJ, Trotter J (1976) Crystal and molecular structure of B, B-bis(p- tolyl)boroxazolidine and the orthorhombic form of B,B-diphenylboroxazolidine. Can J Chem. 54: 3130-41.

Rohacs T, Lopes CM, Michailidis I, Logothetis DE. (2005) PI(4,5)P2 regulates the activation and desensitization of TRPM8 channels through the TRP domain. Nat Neurosci. 8(5): 626-34.

Rosado JA, Brownlow SL, Sage SO. (2002) Endogenously expressed Trp1 is involved in store-mediated Ca2+ entry by conformational coupling in human platelets. J Biol Chem. 277: 42157-63.

Ruissen FV, Van De Kerkhof P, Schalkwijk J. (1995) Singnal transduction pathways in epidermal proliferation and cutaneous inflammation. Clinics in Dermatology. 13: 161-90.

Reeh PW, Steen KH. (1996) Tissue acidosis in nociception and pain. Prog Brain Res. 113: 143-51.

Runnels LW, Yue L, Clapham DE. (2002) The TRPM7 channel is inactivated by PIP(2) hydrolysis. Nat Cell Biol. 4(5): 329-36.

Shi J, Mori E, Mori Y, Mori M, Li J, Ito Y, Inoue R (2004) Multiple regulation by calcium of murine homologues of transient receptor potential proteins TRPC6 and TRPC7 expressed in HEK293 cells. J Physiol. 561: 415-432.

Shin J, Cho H, Hwang SW, Jung J, Shin CY, Lee SY, Kim SH, Lee MG, Choi YH, Kim J, Haber NA, Reichling DB, Khasar S, Levine JD, Oh U. (2002) Bradykinin-12- lipoxygenase-VR1 signaling pathway for inflammatory hyperalgesia. Proc Natl Acad Sci U S A. 99: 10150-5.

Shu X, Mendell LM. (2001) Acute sensitization by NGF of the response of small- diameter sensory neurons to capsaicin. J Neurophysiol. 86(6): 2931-8.

Smart D, Gunthorpe M, Jerman J, Nasir S, Gray J, Muir A, Chambers J, Randall A, Davis J. (2000) The endogenous lipid anandamide is a full agonist at the human vanilloid receptor (hVR1). Br J Pharmacol. 129: 227-30.

223

Smith G, Gunthorpe M, Kelsell R, Hayes P, Reilly P, Facer P, Wright J, Jerman J, Walhin J, Ooi L, Egerton J, Charles K, Smart D, Randall A, Anand P, Davis J. (2002) TRPV3 is a temperature-sensitive vanilloid receptor-like protein. Nature. 418: 186- 190.

Smith PL, Maloney KN, Pothen RG, Clardy J, Clapham DE. (2006) Bisandrographolide from Andrographis paniculata activates TRPV4 channels. J Biol Chem. 281(40): 29897-904.

Songyang Z, Fanning A, Fu C, Xu J. (1997) Recognition of unique carboxyl-terminal motifs by distinct PDZ domains. Science. 275: 73-7.

Steen KH, Steen AE, Kreysel HW, Reeh PW. (1996) Inflammatory mediators potentiate pain induced by experimental tissue acidosis. Pain. 66(2-3): 163-70.

Strang CJ, Henson E, Okamoto Y, Paz MA, Gallop PM (1989) Separation and determination of alpha -amino acids by boroxazolidone formation. Anal Biochem. 178: 278-86.

Suzuki M, Mizuno A, Kodaira K, Imai M. (2003) Impaired pressure sensation in mice lacking TRPV4. J Biol Chem. 278(25): 22664-8.

Szallasi, A. (2001) Vanilloid receptor ligands: hopes and realities for the future. Drugs and Aging. 18: 561-73.

Szallasi, A. (2002) Vanilloid (Capsaicin) receptors in health and disease. Am. J. Clin. Pathol. 118: 110-21.

Stucky CL, Gold MS, Zhang X. (2001) Mechanisms of pain. Proc Natl Acad Sci U S A. 98 (21): 11845-6.

Tominaga M, Caterina MJ, Malmberg AB, Rosen TA, Gilbert H, Skinner K, Raumann BE, Basbaum AI, Julius D. (1998) The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron. 21(3): 531-43.

Tominaga M, Wada M, Masu M. (2001) Potentiation of capsaicin receptor activity by metabotropic ATP receptors as a possible mechanism for ATP-evoked pain and hyperalgesia. Proc Natl Acad Sci U S A. 98: 6951-6.

Tominaga M, Tominaga T. (2005) Structure and function of TRPV1. Pflugers Arch. 451(1): 143-50.

224 Trebak M, Bird GS, McKay RR, Putney JW Jr. (2002) Comparison of human TRPC3 channels in receptor-activated and store-operated modes. Differential sensitivity to channel blockers suggests fundamental differences in channel composition. J Biol Chem. 277: 21617-23. van Meer G. (2002) The different hues of lipid rafts. Science. 296(5569): 855-7. van Rossum DB, Patterson RL, Ma HT, Gill DL (2000) Ca2+ entry mediated by store depletion, S-nitrosylation, and TRP3 channels. Comparison of coupling and function. J Biol Chem. 275: 28562-8.

Voets T, Droogmans G, Wissenbach U, Janssens A, Flockerzi V, Nilius B. (2004) The principle of temperature-dependent gating in cold- and heat-sensitive TRP channels. Nature. 430(7001): 748-54.

Vriens J, Watanabe H, Janssens A, Droogmans G, Voets T, Nilius B. (2004) Cell swelling, heat, and chemical agonists use distinct pathways for the activation of the cation channel TRPV4. Proc Natl Acad Sci U S A. 101(1): 396-401.

Waldmann R, Champigny G, Bassilana F, Heurteaux C, Lazdunski M. (1997) A proton-gated cation channel involved in acid-sensing. Nature. 386(6621): 173-7.

Wang C, Hu HZ, Colton CK, Wood JD, Zhu MX. (2004) An alternative splicing product of the murine trpv1 gene dominant negatively modulates the activity of TRPV1 channels. J Biol Chem. 279(36): 37423-30.

Watanabe H, Vriens J, Prenen J, Droogmans G, Voets T, Nilius B. (2003) Anandamide and arachidonic acid use epoxyeicosatrienoic acids to activate TRPV4 channels. Nature. 424(6947): 434-8.

Woolf CJ, Salter MW. (2000) Neuronal plasticity: increasing the gain in pain. Science. 288 (5472): 1765-9.

Xiao R, Tang J, Wang C, Colton CK, Tian J, Zhu MX (2006) Calcium plays a central role in the sensitization of TRPV3 channel to repetitive stimulations. In review.

Xu H, Ramsey S, Kotecha S, Moran M, Chong J, Lawson D, Ge P, Lilly J, Silos-Santiago I, Xie Y, DiStefano P, Curtis R, Clapham D. (2002) TRPV3 is a calcium-permeable temperature sensitive cation channel. Nature. 418: 181-6.

Xu H, Blair NT, Clapham DE. (2005) Camphor activates and strongly desensitizes the transient receptor potential vanilloid subtype 1 channel in a vanilloid-independent mechanism. J Neurosci. 25(39): 8924-37.

225 Xu H, Delling M, Jun JC, Clapham DE. (2006) Oregano, thyme and clove-derived flavors and skin sensitizers activate specific TRP channels. Nat Neurosci. 9(5): 628- 35.

Xu SZ, Zeng F, Boulay G, Grimm C, Harteneck C, Beech DJ (2005a) Block of TRPC5 channels by 2-aminoethoxydiphenyl borate: a differential, extracellular and voltage- dependent effect. Br J Pharmacol. 145: 405-14.

Xu H, Blair NT, Clapham DE (2005b) Camphor activates and strongly desensitizes the transient receptor potential vanilloid subtype 1 channel in a vanilloid-independent mechanism. J Neurosci 25: 8924-37.

Yeh BI, Kim YK, Jabbar W, Huang CL (2005) Conformational changes of pore helix coupled to gating of TRPV5 by protons. EMBO J 24: 3224-34.

Yue L, Peng JB, Hediger MA, Clapham DE (2001) CaT1 manifests the pore properties of the calcium-release-activated calcium channel. Nature. 410: 705-9.

Zhang P, Canessa CM. (2002) Single channel properties of rat acid-sensitive ion channel- 1alpha, -2a, and -3 expressed in Xenopus oocytes. J Gen Physiol. 120(4): 553-66.

Zhang Z, Okawa H, Wang Y, Liman ER. (2005) Phosphatidylinositol 4,5-bisphosphate rescues TRPM4 channels from desensitization. J Biol Chem. 280(47): 39185-92.

Zhu X, Jiang M, Peyton M, Boulay G, Hurst R, Stefani E, Birnbaumer L (1996) trp, a novel mammalian gene family essential for agonist-activated capacitative Ca2+ entry. Cell 85: 661-71.

Zhu X, Birnbaumer L. (1998) Calcium channels formed by mammalian Trp homologues. News Physiol Sci. 13: 211-7.

Zimmermann K, Leffler A, Fischer MM, Messlinger K, Nau C, Reeh PW (2005) The TRPV1/2/3 activator 2-aminoethoxydiphenyl borate sensitizes native nociceptive neurons to heat in wildtype but not TRPV1 deficient mice. Neuroscience. 135: 1277- 84.

Zygmunt PM, Petersson J, Andersson DA, Chuang H, Sorgard M, Di Marzo V, Julius D, Hogestatt ED. (1999) Vanilloid receptors on sensory nerves mediate the vasodilator action of anandamide. Nature. 400(6743): 452-7.

226