Analgesic transient receptor potential vanilloid-1-active compounds inhibit native and recombinant T-type calcium channels

Abstract

Background and Purpose

T-type calcium (Ca_v3) and transient receptor potential vanilloid-1 (TRPV1) channels play central roles in the control of excitability in the peripheral nervous system and are regarded as potential therapeutic pain targets. Modulators that either activate or inhibit TRPV1-mediated currents display analgesic properties in various pain models despite opposing effects on their connate target, TRPV1. We explored the effects of TRPV1-active compounds on Ca_v3-mediated currents.

Experimental Approach

Whole-cell patch clamp recordings were used to examine the effects of TRPV1-active compounds on rat dorsal root ganglion low voltage-activated calcium currents and recombinant Ca_v3 isoforms in expression systems.

Key Results

The classical TRPV1 agonist capsaicin as well as TRPV1 antagonists A-889425, BCTC, and capsazepine directly inhibited Ca_v3 channels. These compounds altered the voltage-dependence of activation and inactivation of Ca_v3 channels and delayed their recovery from inactivation, leading to a concomitant decrease in T-type current availability. The TRPV1 antagonist capsazepine potently inhibited Ca_v3.1 and 3.2 channels (K_D < 120 nM), as demonstrated by its slow off rate. In contrast, neither the TRPV1 agonists, Palvanil and resiniferatoxin, nor the TRPV1 antagonist AMG9810 modulated Ca_v3-mediated currents.

Conclusions and Implications

Analgesic TRPV1-active compounds inhibit Ca_v3 currents in native and heterologous systems. Hence, their analgesic effects may not be exclusively attributed to their actions on TRPV1, which has important implications in the current understanding of nociceptive pathways. Importantly, our results highlight the need for attention in the experimental design used to address the analgesic properties of Ca_v3 channel inhibitors.

Abbreviations

A-889425

1-(3-methylpyridin-2-yl)-N-(4-(trifluoromethylsulfonyl)phenyl)-1,2,3,6-tetrahydropyridine-4-carboxamide

BCTC

N-(4-tert-butylphenyl)-4-(3-chloropyridin-2-yl)piperazine-1-carboxamide

Ca_v

voltage-gated calcium channel

DRG

dorsal root ganglion

LVA

low voltage-activated

SSI

steady-state inactivation

TRP

transient receptor potential.

What is already known

• Capsaicin and capsazepine inhibit voltage-gated calcium channels in rat dorsal root ganglion neurons.

What this study adds

• A description of the modulation of rat and human T-type calcium channels by TRPV1-active compounds.

What is the clinical significance

• Some analgesic effects of TRPV1-active compounds may arise from inhibition of T-type calcium channels.

1 INTRODUCTION

T-type calcium channels, or Ca_v3 channels, are voltage-gated calcium selective channels that, compared to other Ca_v channels, activate at less depolarized potentials (Catterall, Goldin, & Waxman, 2003). This property makes Ca_v3 channels particularly important in the regulation of cell excitability (Cain & Snutch, 2010). Numerous studies have documented the critical roles played by T-type calcium channels in the control of neuronal firing modalities and demonstrate their involvement in neuronal low-threshold burst firing (Cain & Snutch, 2010). It has been shown that up-regulation of Ca_v3 expression is a common feature of various conditions that involve hyperexcitability such as chronic pain. For example, Ca_v3 channels are up-regulated in dorsal root ganglion (DRG) neurons of various animal models of pain (Jagodic et al., 2008; Kang et al., 2018; Wen et al., 2010; Yue, Liu, Liu, Shu, & Zhang, 2013), visceral hypersensitivity (Marger et al., 2011), and experimental models of diabetes (Cao, Byun, Chen, & Pan, 2011; Jagodic et al., 2007; Obradovic et al., 2014). Moreover, their abundance in peripheral nociceptors has attracted much attention to Ca_v3 channels as analgesic targets for the treatment of pain with numerous novel drugs being developed targeting Ca_v3 channels that show promising potent analgesic effects (for recent review, see Snutch & Zamponi, 2018).

Within the Ca_v3 family, Ca_v3.2 is the predominant isoform expressed in sensory neurons implicated in nociceptive signalling (Bourinet et al., 2005). Additionally, Ca_v3.2 knockin/flox studies marked their relevance in setting up the firing threshold of low-threshold mechanoreceptors (LTMRs) Aδ- and C-LTMRs (Francois et al., 2015); whereas genetic ablation of Ca_v3.2 in C-LTMRs validates their role in the noxious mechanical cold, light-touch, and chemical perception that underlie various forms of neuropathic pain. Although Ca_v3.1 channels have been less explored, this isoform was recently implicated in the pathophysiology of trigeminal neuropathic pain (Choi, Yu, Hwang, & Llinas, 2016).

T-type calcium channels (Ca_v3 channels) and transient receptor potential vanilloid 1 (TRPV1) channels co-localize in various neuronal types in the periphery (Cardenas, Del Mar, & Scroggs, 1995) and both have been implicated in pain signal generation. Capsaicin, the pungently painful compound found in hot chili peppers from the capsicum family, is the classical agonist of the TRPV1 receptor whose activation mediates Ca²⁺ influx and excitation in several neuronal types. This property guided the identification of TRPV1 channels (Caterina et al., 1997) and subsequently the development of the synthetic TRPV1 antagonist, capsazepine (Bevan et al., 1992). TRPV1 is a polymodal receptor abundantly expressed in sensory neurons where it responds to a variety of noxious stimuli including heat, low pH, and chemical irritants (Caterina et al., 1997; Jordt, Tominaga, & Julius, 2000). As a non-selective cation channel, TRPV1 activation leads to local membrane depolarization which initiates a cascade of events that trigger opening of voltage-gated sodium channels responsible for the upstroke of the action potential and propagation of the noxious signal. Therefore, inhibition of TRPV1-mediated currents is considered one of the mechanisms behind some forms of analgesia. Indeed, the classical TRPV1 agonist, capsaicin, and antagonist, capsazepine, both exhibit analgesic properties in various pain assays, despite of their opposing effects on this vanilloid receptor (see recent review of Moran & Szallasi, 2018). Capsazepine's analgesic action is conceptualized as a consequence of decreasing TRPV1-mediated currents that support hyperexcitability. Conversely, capsaicin promotes TRPV1 opening, which is initially perceived as heat or pain, and at a later phase analgesia. This apparent paradoxical effect has been explained by a capsaicin-induced desensitization of TRPV1 that causes quiescence in hyperactive pain neurons.

Capsaicin is used as a topical analgesic in “low-concentration” creams (0.1% or ~3 mM) that have poor efficacy in the treatment of neuropathic pain (Derry & Moore, 2012) or improved efficacy “high concentration” patches (8% or ~260 mM; Noto, Pappagallo, & Szallasi, 2009). These doses are several fold higher (>50,000, lower limit) than those required to activate the human isoform of TRPV1 (EC₅₀ ~0.05–0.3 μM; Li, Wang, Chuang, Cohen, & Chuang, 2011). At such high concentrations, off-target effects become significant and a clear mechanism for its analgesic effects is difficult to ascertain. Both capsaicin and capsazepine also modulate other membrane receptors and ion channels, particularly the voltage-dependent Ca_v channels (Castillo et al., 2007; Docherty, Yeats, & Piper, 1997; Hagenacker, Splettstoesser, Greffrath, Treede, & Busselberg, 2005). However, many reports have been based on native low voltage-activated (LVA) calcium currents without expression system verification, and therefore, a direct effect of these compounds over Ca_v3-mediated currents is missing. It has been suggested that the modulation of LVA and high voltage-activated calcium currents in DRG neurons induced by a low-concentration capsaicin occurs via a TRPV1-mediated mechanism (Comunanza, Carbone, Marcantoni, Sher, & Ursu, 2011; Kerckhove et al., 2014; Wu, Chen, & Pan, 2005). This has recently been shown to occur via an intracellular calcium-dependent mechanism (Cazade, Bidaud, Lory, & Chemin, 2017; Comunanza et al., 2011), leading to inhibition of Ca_v3-mediated currents, caused by the influx of Ca²⁺ through activated TRPV1. To date, the mechanism of direct capsaicin inhibition of Ca_v3 channels has not been addressed.

TRPV1 and Ca_v3 channels co-localize in various neuronal types in the periphery (Cardenas et al., 1995) and both channel families have been implicated in pain signalling. In the present study, we investigated Ca_v3 channels as a potential off target of TRPV1 modulators. Whole-cell patch clamp recording was used to assess native T-type calcium currents in DRG neurons and heterologously expressed Ca_v3 channels to determine their mode of action. Our results show that several but not all TRPV1 modulators are capable of direct inhibition of Ca_v3 channels in the absence of TRPV1.

2 METHODS

2.1 Cell culture and transfection

HEK293 cells containing the SV40 Large T-antigen (HEK293T, CRL-3216, ATCC, USA, RRID:CVCL_0063) were cultured in DMEM (Invitrogen Life Technologies, Australia), supplemented with 10% FBS (Bovigen, Australia), 1% penicillin and streptomycin (Pen/Strep, Invitrogen Life Technologies), and 1× GlutaMAX supplement (Invitrogen Life Technologies) at 37°C, 5% CO₂. In all experiments, HEK293T cells were transiently co-transfected using calcium phosphate with plasmid cDNAs encoding human Ca_v3.1 (provided by Dr G. Zamponi), human Ca_v3.2 (a1Ha-pcDNA3 was a gift from Dr E. Perez-Reyes, Addgene #45809; Cribbs et al., 1998), human Ca_v3.3 (a1Ic-HE3-pcDNA3 also from Dr E. Perez-Reyes, Addgene #45810; Gomora, Murbartian, Arias, Lee, & Perez-Reyes, 2002), or rat TRPV1 in combination with green fluorescent protein (GFP). Cells were transfected in solution and plated on 12 mm glass coverslips for patching within 24–72 hr. After 8–12 hr, transfection medium was replaced with fresh culture media and cells were incubated at 30°C, 5% CO₂ to allow adequate expression.

2.2 DRG isolation and culture

Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny, Browne, Cuthill, Emerson, & Altman, 2010) and with the recommendations made by the British Journal of Pharmacology. DRGs were isolated from neonatal (P0-P2) Sprague Dawley rats (ArcCrl:CD (SD)IGS, Animal Resources Centre, Australia, RRID:RGD_8553001) following decapitation, as approved by the University of Wollongong Animal Ethics Committee (AE17/23), in compliance with the United States NIH. Rats were housed in double decker, individually-ventilated cages containing corncob bedding, nesting material, chewing block, PVC tunnel, and a plastic or cardboard shelter. Rats were housed at 21°C on a 12 hour light/dark cycle, and fed Vella Rat Pellets and water ad libitum. Briefly, DRG ganglia were collected, by hemisecting the spinal cord and removing individual DRG ganglia, in ice-cold divalent free HBSS where dorsal root and peripheral nerve processes were carefully trimmed. DRG ganglia were transferred to dissociation media (divalent free HBSS with 2 mg·ml⁻¹ collagenase Type II, Worthington, USA) for 60 min at 37°C, 5% CO₂. Digested DRG ganglia were spun for 10 min at 200 g at 23°C and the supernatant discarded. The DRG neurons were washed with fresh DMEM (supplemented with 10% FBS, 1× GlutaMAX, and 1% penicillin/streptomycin) followed by trituration with progressively smaller diameter fire polished Pasteur pipettes. Dissociated DRG neurons were then filtered through a 160 μm nylon mesh (Millipore, Australia) to remove cell debris and non-dissociated material. The DRG neuron suspension was then plated on poly-L-lysine and laminin coated 12 mm cover glass (Sigma, Australia) and left to attach for ~3 hr at 37°C, 5% CO₂ after which ~1 ml of fresh DMEM (10% FBS, 1× GlutaMAX, and 1% pen/strep) was added and incubated overnight. Primary DRG cultures were held at 30°C and recorded within 48 hr. Electrophysiology of native LVA calcium currents

Whole-cell patch clamp recording of DRG neurons were carried out 1–2 days post isolation. DRG neurons were perfused with extracellular solution containing (mM): 140 TEA-Cl, 10 CaCl₂, 1 MgCl₂, 10 D-glucose, 10 HEPES, and pH 7.4 with TEA-OH (~320 mOsmol·kg⁻¹). Borosilicate patch pipettes (World Precision Instruments, USA) were fire-polished to a resistance of 1–3 MΩ and filled with intracellular solution containing (mM): 150 CsCl, 1.5 MgCl₂, 5 EGTA, 10 HEPES, and pH 7.2 with CsOH (~300 mOsmol·kg⁻¹). Depolarization-activated calcium currents (I_Ca) were recorded at room temperature (22–24°C) using a MultiClamp 700B amplifier (Molecular Devices, USA), digitalized via a Digidata 1440 controlled by pClamp10.7 acquisition system. LVA I_Ca were elicited from a holding potential (V_h) of −90 mV with a step depolarization of 100 ms to −40 mV at 0.2 Hz. Compounds were superfused to the neuron using a syringe pump (New Era Pump Systems Inc, USA) loaded with 1 ml syringe connected to a 50 cm MicroFil 28G (World Precision Instruments) at 2 μl·min⁻¹ allowing fast solution exchanges.

2.3 Electrophysiology of transiently transfected Ca_v3s

Whole-cell patch recordings of HEK293T cells were carried out 1–3 days post transfection at room temperature. Cells were constantly perfused with extracellular solution containing (mM): 100 NaCl, 10 CaCl₂, 1 MgCl₂, 5 CsCl, 30 TEA-Cl, 10 glucose, 10 HEPES, and pH 7.4 with TEA-OH (~320 mOsmol·kg⁻¹). Fire-polished borosilicate pipettes (1–3 mΩ) were filled with intracellular solution containing (mM): 140 K-Gluconate, 5 NaCl, 2 MgCl₂, 5 EGTA, 10 HEPES, and pH 7.4 with KOH (~300 mOsmol·kg⁻¹).

Calcium currents were elicited by 100 ms depolarizing test pulses to −20 mV or a 180 ms ramp from −90 to +70 mV from a holding potential of −90 mV at 0.2 Hz. Activation curves were generated from stimulation protocols consisting of a series of depolarizing test pulses from −90 to +20 mV (Δ10 mV, V_h − 90 mV) at 0.2 Hz before and after compound application. Similarly, steady-state inactivation (SSI) was assessed by the peak current amplitude at −20 mV (50 ms) achieved after 1 s pre-pulses (from −90 to +10 mV, Δ5 mV). Recovery from inactivation was recorded using a standard double pulse protocol (V_h − 90 mV) consisting of a 100 ms (hCa_v3.1 or hCa_v3.2) or 150 ms (hCa_v3.3) pulse to −20 mV (P1) that achieved complete inactivation, followed by a second 20 ms test pulse (P2) to −20 mV after progressively increasing inter-pulse intervals (from 0 to 10 s).

2.4 Data and statistical analysis

Data analysis and graphs were generated using OriginPro (Origin Lab Corporation, USA, RRID:SCR_014212). Concentration–response curves for each compound were constructed from ≥5 individual experiments per concentration tested. Fractional current, determined as the ratio between peak current in the presence of compound (I_Comp) over peak current determined prior to application (I_CTR), was plotted against the compounds concentration. Curves were fit with a Hill equation of following

where h is the n_H and IC₅₀ is the half-maximal inhibitory concentration. Activation and SSI curves were generated from 5 to 16 independent experiments and fit with a modified Boltzmann equation of the form:

where V_t is the test potential, V_0.5 is the half-maximal activation potential, and ka is the slope factor related to the voltage-dependence. Recovery from inactivation plots were fit using a single (Ca_v3.3) or double (Ca_v3.1 and Ca_v3.2) exponential of the following equations:

where τ is the fast and slow time constant and A is the amplitude of the fast and slow components.

Compound block and unblock kinetics (k_on and k_off) were measured by fitting the peak currents for successive depolarizations during toxin washin or washout with a single exponential to determine, τ_on and τ_off, of the following equation:

From the observed τ_on and τ_off, we calculated k_on and k_off with the following equations.

where [Comp] is the concentration of compound applied. These rate constants were used to calculate the equilibrium K_D from the following equation:

The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology (Curtis et al., 2015). As our experiments were conducted on isolated DRG neurons from a combined pool of DRGs from several rats, animals were not randomized, and each cell served as its own control. The nature of the experiments, which are for the most part perfusing known concentrations of a compound onto single cells, make blinding impractical. Statistical significance (P < 0.05) was determined using Student's unpaired t test. All data are presented as mean ± SEM (n).

2.5 Materials

All test compounds were dissolved in DMSO to generate stock solutions of 10 or 100 mM, according to the manufacturer's instructions. Capsaicin (Sigma-Aldrich, Australia), resiniferatoxin and capsazepine (Abcam, Australia), A-889425, BCTC, AMG9810, and N-palmitoyl-vanillamide (Palvanil, Alomone, Israel) were diluted in extracellular solution to the required final concentrations. The maximal DMSO concentration during recordings was 0.5% and did not cause any detectable changes in Ca_v3 Ca²⁺ currents (I_Ca) under our experimental conditions (data not shown).

2.6 Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (Alexander et al., 2017).

3 RESULTS

3.1 Capsaicin and capsazepine inhibit rat DRG neuron LVA calcium currents

At room temperature, the shallow voltage dependence and low open probability of TRPV1 channels allows the evaluation of the effects of capsaicinoid compounds, such as capsaicin and capsazepine, on native LVA I_Ca at physiological membrane potentials. High extracellular TEA concentrations were used to inhibit both endogenous K⁺- and TRPV1-mediated currents (Rivera-Acevedo, Pless, Schwarz, & Ahern, 2012). We recorded LVA I_Ca from neonatal (medium diameter, 30–40 μm) DRG neurons with a test pulse to −40 mV from a holding potential (V_h) of −90 mV (Figure 1a). Control LVA I_Ca (black trace) were recorded before 100 μM application of capsaicin (red, Figure 1a, i) or capsazepine (purple, Figure 1a, ii). Currents after a 5 min washout (shown in grey) qualitatively demonstrate differences in reversibility of both compounds. At 100 μM, capsaicin inhibited 93 ± 0.9% (n = 5) of the total LVA I_Ca elicited by the test pulse (Figure 1b), and its effect was completely reversed upon washout (Figure 1a, i). In contrast, capsazepine (100 μM) reduced the LVA I_Ca almost completely (98.3 ± 1.8%; n = 5; Figure 1b) whilst its effects were poorly reversed (<10%) during 5 min washout (Figure 1a, ii). These results show that capsaicin and capsazepine inhibit native T-type calcium currents in neonatal rat DRG neurons with distinct unblocking kinetics.

Figure 1

TRPV1 agonist (capsaicin) and antagonist (capsazepine) inhibit LVA calcium currents in rat DRG neurons. (a) Representative LVA calcium currents elicited by a 100 ms depolarizing pulse to −40 mV (V_h = −90 mV, 0.2 Hz, see voltage protocol bottom inset) in control (black), capsaicin (red, i) or capsazepine (purple, ii), and after a 5 min washout (grey). (b) Bar graph depicting % inhibition of LVA calcium currents in DRG neurons by 100 μM capsaicin (red, n = 5) and 100 μM capsazepine (purple, n = 5)

3.2 TRPV1-active compounds that inhibit recombinant Ca_v3 channels

A panel of TRPV1 agonists and antagonists commonly used in the study of TRPVs (Figure 2a) were screened against the human isoforms of the T-type calcium channels at a concentration of 10 μM (Figure 2b). hCa_v3.1, hCa_v3.2, and hCa_v3.2 were expressed in HEK293T cells, and whole-cell patch clamp recordings were used to monitor effects on I_Ca in the presence of TRPV1 agonists: capsaicin, resiniferatoxin, and Palvanil, and TRPV1 antagonists: capsazepine, BCTC, AMG9810, and A-889425 (Figure 2b). Neither Palvanil (n = 5) nor resiniferatoxin (n = 5) elicited any obvious effects on human Ca_v3-mediated current at 10 μM. Nevertheless, at the same concentration, capsaicin was able to modestly inhibit all three family members (Ca_v3.1: 22.5 ± 2.3%, n = 6; Ca_v3.2: 28.5 ± 3.5%, n = 5; Ca_v3.3: 32.6 ± 0.9%, n = 5). From the panel of TRPV1 antagonists, capsazepine, A-889425, and BCTC at 10 μM robustly inhibited (>50%) human Ca_v3.1, 3.2, and 3.3 channels (Figure 2b). At the same concentration, however, AMG9810 (n = 5) was ineffective at modulating the recombinant Ca_v3 currents.

Figure 2

TRPV1 agonists and antagonists inhibit heterologously expressed hCa_v3 isoforms in a concentration-dependent manor. (a) Chemical structures of TRPV1-active compounds comprising agonists and antagonists used in this study. (b) Bar graph showing % inhibition of hCa_v3.1, hCa_v3.2, and hCa_v3.3 by 10 μM capsaicin (n ≥ 5), capsazepine (n = 5), Palvanil (n = 5), BCTC (n ≥ 11), AMG 9810 (n = 5), A-889425 (n ≥ 13), and resiniferatoxin (n = 5). (c) Concentration–response relationships obtained for capsaicin, A-889425, and BCTC on hCa_v3.1 (left), hCa_v3.2 (middle), or hCa_v3.3 (right) calcium currents

Concentration–response curves for the inhibition of Ca_v3-mediated currents were generated for capsaicin, A-889425 and BCTC (Figure 2c, each concentration on the concentration–response relationship contains ≥5 individual points). Ca_v3.1 channels were most potently inhibited by A-889425 (IC₅₀: 3.2 ± 0.4 μM; n_H: 0.9 ± 0.1), followed by BCTC (IC₅₀: 10.2 ± 0.3 μM; n_H: 1.5 ± 0.1) and capsaicin (IC₅₀: 26.4 ± 1.4 μM; n_H: 1.4 ± 0.1; Figure 2c, left). In the case of Ca_v3.2, BCTC and A-889425 displayed comparable potencies (IC₅₀: 3.4 ± 0.3 μM, n_H: 1.0 ± 0.1; and IC₅₀: 4.8 ± 0.3 μM, n_H: 1.4 ± 0.1; respectively), whereas the IC₅₀ obtained for capsaicin was 23.6 ± 1.4 μM (n_H: 1.1 ± 0.1, Figure 2c, middle). Similarly, A-889425 and BCTC were equipotent at Ca_v3.3 (5.8 ± 0.4 μM; n_H: 1.5 ± 0.1; and 6.9 ± 0.4 μM; n_H: 1.1 ± 0.1, respectively), whereas capsaicin inhibited 50% of Ca_v3.3 currents at 16.3 ± 0.5 μM (n_H: 1.2 ± 0.04; Figure 2c, right). Thus, capsaicin, A-889425, and BCTC are capable of inhibiting all three hCa_v3s, whereas Palvanil, AMG9810, or resiniferatoxin do not inhibit.

Due to significantly slower blocking kinetics, application of low concentrations of capsazepine (<10 μM) would imply washins extending >30 min (Figure S1). Therefore, the potency of this compound was estimated from its blocking kinetics to Ca_v3 channels (described in Section 4.3) rather than concentration–response relationships.

3.3 Activation of TRPV1 by capsaicin indirectly inhibits Ca_v3 channels and depends strictly on Ca²⁺ influx

Previous studies have described T-type Ca²⁺ current inhibition by low doses of TRPV1-agonists in DRG neurons (Comunanza et al., 2011; Kerckhove et al., 2014). In the present study, this observation was verified and expanded in detail through co-expression experiments of TRPV1 and Ca_v3.1 in HEK293 cells.

In the absence of calcium, Ca_v3 channels can conduct Na⁺ efficiently (Khan, Gray, Obejero-Paz, & Jones, 2008; Senatore, Guan, Boone, & Spafford, 2014), thus robust inward and outward T-type currents can be elicited and measurement of channel block under calcium free conditions can be quantified (Figure 3).

Figure 3

Modulatory effects of Ca_v3.1 currents by permeant ions and co-expression of TRPV1. Representative ramp traces of hCa_v3.1 with Ca²⁺ (a) or Na⁺ (d) as the permeant ion or hCa_v3.1 co-expressed with TRPV1 with Ca²⁺ (b) or Na⁺ (e) as the permeant ion. Current traces obtained in the absence (control, black), in the presence of 1 μM capsaicin (red) or after washout (grey). In part (a) and (d), control, capsaicin, and washout traces are overlaid. Activation (triangle) and steady-state inactivation (square) curves when Ca²⁺ (c) or Na⁺ (f) is the permeant ion. When hCa_v3.1 is expressed alone (filled) or co-expressed with TRPV1 (open) in absence (black) or presence of 1 μM capsaicin (red)

When co-expressed with TRPV1 channels, Ca_v3.1 currents are easily distinguished by their voltage dependence and kinetics (Figure 3) and thus we can evaluate the effects of activating TRPV1 with capsaicin at a concentration (1 μM) that does not directly block Ca_v3.1 (Figure 2c, left panel). Ramp stimuli (−90 to +70 mV; 180 ms, 0.2 Hz; Figure 3a) were applied to cells expressing Ca_v3.1 alone (3A/D) and to cells where Ca_v3.1 and TRPV1 (3B/E) were co-transfected. Recordings in Na⁺-based solutions with 10 mM Ca²⁺ (Figure 3a/b) or zero external Ca²⁺ (Figure 3d/e) are shown. In the presence of 10 mM extracellular Ca²⁺, peak inward currents at −15 mV are elicited (Figure 3a/b, black traces). Application of capsaicin to cells expressing Ca_v3.1 did not alter the elicited current (Figure 3a) nor channel activation or SSI properties. Conversely, capsaicin applied to cells co-expressing Ca_v3.1 and TRPV1 caused a robust outwardly rectifying TRPV1 current that quickly desensitized (Figure 3b). The influx of Ca²⁺ through TRPV1 produced a leftward shift in the peak Ca_v3.1 current (Figure 3b, red trace) which was irreversible after a 5 min washout (Figure 3b, grey trace). This shift in the ramp current was mirrored by hyperpolarizing shifts in both channel activation and SSI from square pulse protocols (Figure 3c and Table 1).

Table 1. Comparison of activation and steady-state inactivation parameters for hCa_v3.1 and hCa_v3.1 + TRPV1 channels obtained with extracellular Ca²⁺ or Na⁺ in the absence and presence of capsaicin (1 μM)

Parameter	Cav3.1				Cav3.1 + TRPV1
	Ca^2+	Ca^2+ capsaicin	Na^+	Na^+ capsaicin	Ca^2+	Ca^2+ capsaicin	Na^+	Na^+ capsaicin
V0.5 Act (mV)	−27.0 ± 0.2	−20.9 ± 0.3	−2.6 ± 0.8	−4.0 ± 1.2	−24.6 ± 0.5	−40.8 ± 0.5	−6.2 ± 0.8	−2.8 ± 1.1
V0.5 Inact (mV)	−47.2 ± 0.1	−53.4 ± 0.5	−72.0 ± 0.3	−72.3 ± 0.2	−54.2 ± 0.5	−67.0 ± 0.1	−72.4 ± 0.2	−72.4 ± 0.4

A different response occurs in the absence of Ca²⁺ which allows Na⁺ to become the main charge carrier. Under these conditions, the currents elicited by the ramp protocol peak at −30 mV (Figure 3d/e, black traces). When Ca_v3.1 is expressed alone, Na⁺ currents elicited during the ramp protocol were unchanged (Figure 3d), similar to that seen when Ca²⁺ was the charge carrier. However, when TRPV1 is co-expressed with Ca_v3.1, TRPV1 activation by 1 μM capsaicin caused the similar non-desensitizing outwardly rectifying current as those observed in 10 mM Ca²⁺ (Figure 3e, red trace). The modest enhancement of inward peak current is likely due to the compounded TRPV1 and Ca_v3.1 currents that remain unaltered after washout (Figure 3e, grey trace), as reflected by the plots shown in Figure 3f. The co-expression experiments show that indirect inhibition of T-type currents by low capsaicin concentrations in cells where TRPV1 and Ca_v3s channels are present occurs via Ca²⁺ influx through activated TRPV1 channels.

3.4 Capsaicin and capsazepine display distinctive unblock kinetics

In contrast to the effects of capsaicin, inhibition of LVA I_Ca in DRGs by capsazepine was not reversible within the experimental constrains of patch clamp recordings and thus we studied their blocking kinetics using the heterologous system. Representative current traces elicited by a 100 ms depolarizing stimulus to −20 mV (V_h: −90 mV, 0.2 Hz) are shown in Figure 4a/b (red: 100 μM capsaicin; purple: 100 μM capsazepine; black: control; grey: washout). The time course of Ca_v3 current inhibition (washin) and its recovery (washout) upon exposure to capsaicin and capsazepine were examined (Figure 5). Representative diary plots are displayed and coloured accordingly (arrows correspond to current traces depicted in Figure 4). In agreement with the DRG data, >98% inhibition of Ca_v3 current was observed in the presence of 100 μM capsazepine whereas application of 100 μM capsaicin spares ~5–15% of the elicited I_Ca across the three hCa_v3's (Figures 4a/b and 5).

Figure 4

Representative calcium currents of Ca_v3 isoforms in the presence of TRPV1-active compounds. Representative currents mediated by hCa_v3.1 (top), hCa_v3.2 (middle), and hCa_v3.3 (bottom) in control conditions, after application of (a) 100 μM capsaicin, (b) 100 μM capsazepine, (c) 10 μM BCTC or (d) 10 μM A-889425 and washout elicited from V_h = −90 mV by a test pulse to −20 mV (100 ms, 0.2 Hz) and normalized to peak control current values (Scale bars are X: 0.2 nA; Y: 50 ms). (e) Percentage recovery after 5 min of washout

Figure 5

Comparison of capsazepine and capsaicin blocking and unblocking kinetics on Ca_v3 channels. Time course of representative whole-cell calcium current inhibition for (a) capsaicin or (b) capsazepine. Bar showing time course of capsaicin or capsazepine application. Coloured arrows depict time points of traces shown in Figure 4a/b

In the experiments shown, a clear difference in unblocking kinetics can be observed between capsaicin and capsazepine where the off rate, k_off, for capsazepine is significantly less than for capsaicin, but with comparable on rates (k_on; Table 2 and Figure 5). Accordingly, k_on's of capsaicin and capsazepine for both Ca_v3.1 and Ca_v3.2 are similar (~0.004 s⁻¹·μM⁻¹) and contrast with their interaction with Ca_v3.3 (Table 2). Capsaicin blocks Ca_v3.3 ~1.5× faster (k_on = 0.0057 ± 0.0003 s⁻¹·μM⁻¹), whereas capsazepine exhibits a ~2× slower blocking rate towards this channel (k_on = 0.0019 ± 0.0002 s⁻¹·μM⁻¹). Interestingly, capsazepine's off rates are radically slower than those of capsaicin as can be readily observed in the diary plots. Figure 5 shows that capsazepine washouts (I_Ca recovery) extended for several hundreds of seconds often reaching only partial recovery of the control current (<10%). Hence, the estimated k_off for capsazepine at Ca_v3.2 and Ca_v3.1 channels were ~400-fold and 180-fold slower than those of capsaicin to the same channels respectively (see Table 2). From these blocking rates, the apparent K_D for capsaicin were calculated for each channel isoform (19.6, 18.8, and 7.45 μM for hCa_v3.1, hCa_v3.2, and hCa_v3.3, respectively, Table 2) which are consistent with the IC₅₀'s determined from concentration–response relationships (Figure 2c). As expected from the observed slow k_offs, capsazepine K_D from Ca_v3.1 and Ca_v3.2 channels are in the low nanomolar range (120 and 46 nM, respectively) suggesting a much tighter interaction between this TRPV1 inhibitor and the nociceptive channels Ca_v3.2 and Ca_v3.1, than to their connate receptor. Interestingly, the observed blocking kinetics and, consequently, the K_D for capsazepine at Ca_v3.3 was 2.0 μM (Table 2) which may serve as a molecular tool to dissect the contribution of this isoform in native LVA I_Ca.

Table 2. Capsaicin and capsazepine kinetic rate constants for inhibition of Ca_v3.1, 3.2 and 3.3

Parameter	hCav3.1		hCav3.2		hCav3.3
	Capsaicin	Capsazepine	Capsaicin	Capsazepine	Capsaicin	Capsazepine
kon (s^−1·μM^−1)	0.0039 ± 0.0004 n = 6	0.0036 ± 0.0004 n = 9	0.0037 ± 0.0003 n = 6	0.0036 ± 0.0005 n = 5	0.0057 ± 0.0003 n = 6	0.0019 ± 0.0002a n = 6
koff (s^−1)	0.077 ± 0.003 n = 6	0.00043 ± 0.00004a n = 7	0.069 ± 0.007 n = 6	0.00017 ± 0.00002a n = 5	0.043 ± 0.003 n = 8	0.0037 ± 0.0008a n = 5
KD (μM)	19.6	0.12	18.8	0.046	7.45	2.0

• ^a Significant difference between capsaicin and capsazepine kinetics.

Taken together, these results support a strong inhibitory action of several TRPV1-active compounds on T-type calcium channel activity at a concentration range well below those commonly use to cause analgesia.

3.5 TRPV1-active compounds decrease Ca_v3 availability

3.5.1 Voltage-dependent activation and SSI

In order to understand the mechanism of inhibition of T-type calcium channels by capsaicin, BCTC, and A889425, we investigated their actions on the biophysical properties of Ca_v3-mediated currents. Figure 6 summarizes the effects of 30 μM capsaicin on T-type currents exhibiting a consistent depolarizing shift in activation for all three channels (Ca_v3.1: +5.5 mV; Ca_v3.2: +2.0 mV; and Ca_v3.3: +13.1 mV). Furthermore, hyperpolarizing shifts in SSI were also observed for all Ca_v3 isoforms (Ca_v3.1: −8.2 mV; Ca_v3.2: −2.0 mV; and Ca_v3.3: −11.6 mV; Figure 6a and Table 3). Interestingly, the slowest of the T-types, Ca_v3.3 displayed larger apparent shifts in activation and SSI than the other two family members.

Figure 6

Effects of TRPV1-active compounds on voltage-dependent activation, steady-state inactivation (SSI), and recovery from inactivation. (a) Activation (triangle) and steady-state inactivation (square) curves in control (black), 30 μM capsaicin (red), 10 μM BCTC (blue), or 10 μM A-889425 for hCa_v3.1 (top, n = 16 control activation, n = 9 control SSI, n = 6 capsaicin activation, n = 6 capsaicin SSI, n = 5 BCTC activation, n = 5 BCTC SSI, n = 6 A-889425 activation, and n = 5 A-889425 SSI), hCa_v3.2 (middle, n = 11 control activation, n = 10 control SSI, n = 8 capsaicin activation, n = 5 capsaicin SSI, n = 5 BCTC activation, n = 5 BCTC SSI, n = 5 A-889425 activation, and n = 5 A-889425 SSI), and hCa_v3.3 (bottom, n = 11 control activation, n = 5 control SSI, n = 7 capsaicin activation, n = 6 capsaicin, n = 6 BCTC activation, n = 5 BCTC SSI, n = 6 A-889425 activation, and n = 5 A-889425 SSI). (b) Recovery from inactivation in the absence (control), and presence of 30 μM capsaicin, 10 μM BCTC, or 10 μM A-889425 for hCa_v3.1 (top, n = 5), hCa_v3.2 (middle, n = 7 control; n = 5 capsaicin, BCTC, and A-889425), and hCa_v3.3 (bottom, n = 5)

Table 3. Comparison of activation, steady-state inactivation, recovery from inactivation, and macroscopic current activation and inactivation time constants parameters for hCa_v3.1, 3.2, and 3.3 channels obtained in the absence (control) and presence of capsaicin (30 μM), BCTC (10 μM), and A-889425 (10 μM)

Parameter	hCav3.1				hCav3.2				hCav3.3
	Control	Capsaicin	BCTC	A-889425	Control	Capsaicin	BCTC	A-889425	Control	Capsaicin	BCTC	A-889425
V0.5 Act (mV)	−26.96 ± 0.19	−21.48 ± 0.40a	−30.38 ± 0.34a	−27.53 ± 0.48	−28.38 ± 0.57	−26.39 ± 0.36a	−22.48 ± 0.60a	−25.27 ± 0.23a	−24.88 ± 0.36	−11.76 ± 0.50a	−18.86 ± 0.49a	−19.36 ± 0.55a
V0.5 Inact (mV)	−47.16 ± 0.14	−55.39 ± 0.14a	−59.07 ± 0.25a	−66.37 ± 0.33a	−51.46 ± 0.42	−53.49 ± 0.37a	−51.13 ± 0.42	−52.54 ± 0.32	−44.64 ± 0.49	−56.22 ± 0.36a	−56.56 ± 0.29a	−57.03 ± 0.38a
τfast (Afast) (s)	0.07 ± 0.01 (0.70 ± 0.07)	0.16 ± 0.01a (0.56 ± 0.05)	0.23 ± 0.01a (0.62 ± 0.03)	0.21 ± 0.03a (0.46 ± 0.05)	0.38 ± 0.03 (0.65 ± 0.04)	0.34 ± 0.04 (0.23 ± 0.03)	0.39 ± 0.03 (0.22 ± 0.07)	0.26 ± 0.61 (0.03 ± 0.05)	0.38 ± 0.01	0.52 ± 0.01a	0.72 ± 0.02a	1.47 ± 0.04a
τslow (Aslow) (s)	0.32 ± 0.06 (0.30 ± 0.07)	0.56 ± 0.04a (0.44 ± 0.05)	1.17 ± 0.08a (0.37 ± 0.03)	1.06 ± 0.09a (0.54 ± 0.06)	2.18 ± 0.32 (0.28 ± 0.04)	1.61 ± 0.06 (0.74 ± 0.03)	2.21 ± 0.21 (0.77 ± 0.07)	2.54 ± 0.17 (0.96 ± 0.05)	N/A	N/A	N/A	N/A
τact (ms)	2.45 ± 0.26	2.25 ± 0.33	2.01 ± 0.19	2.44 ± 0.29	4.41 ± 0.28	3.54 ± 0.15	5.18 ± 0.45	4.37 ± 0.42	14.35 ± 1.30	9.88 ± 1.41	11.65 ± 1.65	13.64 ± 1.28
τinact (ms)	11.40 ± 0.60	10.20 ± 1.16	10.71 ± 0.92	10.16 ± 1.25	14.62 ± 0.77	13.60 ± 0.72	15.88 ± 1.18	16.82 ± 0.57	22.02 ± 1.68	17.03 ± 1.19	17.63 ± 1.32	23.69 ± 1.39

• ^a Significantly different from control.

Compared to capsaicin, BCTC and A-889425 are more potent inhibitors of all three Ca_v3s (Figure 2); therefore, their effects on T-type channel voltage-dependent properties were investigated at a concentration of 10 μM (Figure 6). All three small molecules, capsaicin, BTCT, and A-889425, affected the voltage-dependent properties of Ca_v3.1 and 3.3 channels comparably. However, BCTC and A-889425 shifted their SSI further in the hyperpolarizing direction (>8 mV, Table 3). In contrast, the voltage dependence of Ca_v3.2 kinetics was the least affected by TRPV1-active compounds, highlighting this isoform's unique biophysical properties.

3.5.2 Recovery from inactivation

A double-pulse protocol was used to evaluate the influence of capsaicin, BCTC, and A-889425 on the recovery from inactivation of T-type channels. A control pulse of sufficient duration (P1, −20 mV, 100 ms, or 150 ms) was used to drive all channels into the inactivated state which was then followed by progressively longer inter-pulses at the holding potential, then a second test pulse P2 (see Figure 6b inset) was applied before and after application of the TRPV1-active molecules. In control conditions, the recovery from inactivation of Ca_v3.1 and Ca_v3.2 is better approximated by the sum of two exponential functions whereas Ca_v3.3 can be satisfactorily fit with a single exponential (Figure 6b). All three compounds significantly delayed the recovery of Ca_v3s from inactivation. The quantification of such effect revealed that in the presence of capsaicin, both recovery time constants (fast and slow) increased ~twofold, with a significant increase in the relative amplitude of the slow component (from 0.30 ± 0.07 control to 0.44 ± 0.05 in the presence of capsaicin; Figure 6b and Table 3). The major effect of capsaicin on Ca_v3.2 recovery from inactivation was a significant increase in the proportion of the slow component of the recovery (0.28 ± 0.04 vs. 0.74 ± 0.03). Accordingly, we observed a decrease in recovery from inactivation in Ca_v3.3 currents determined by the slowing of the time constant (τ = 0.38 ± 0.01 s for control vs. τ = 0.52 ± 0.01 s in the presence of capsaicin, Table 3). Examination of Ca_v3 channel recovery from inactivation in the presence of BCTC and A-889425 showed larger effects than those seen for capsaicin (Figure 6b and Table 3) but with overall similar trends. For Ca_v3.1, both BCTC and A-889425 increased both time constants and the proportion of τ_slow component (Table 3). For Ca_v3.2, the fast and slow time constants were not significantly altered but a large increase in the proportion of τ_slow component was observed (CTR: 0.28 ± 0.04, vs. BCTC: 0.74 ± 0.03, and A-889425: 0.96 ± 0.05). BCTC slowed the recovery from inactivation ~twofold for hCa_v3.3 whereas A-889425 demonstrated a larger ~fourfold slowing in recovery from inactivation (Figure 6b).

Across the Ca_v3s, the mechanisms of inhibition by capsaicin, BCTC, and A-889425 were consistent with a decrease in channel activity. The changes in voltage-dependent gating, activation rightward shift and inactivation leftward shift, together with the significant delay in recovery from inactivation of these TRPV1-active compounds act synergistically to decrease the available T-type current at membrane potentials where these channels are known to regulate excitability.

4 DISCUSSION

In this study, we describe the direct modulation of native and recombinant T-type calcium channels by TRPV1-active compounds that are considered analgesic. We evaluated molecules known to activate TRPV1 (capsaicin, Palvanil, and resiniferatoxin) as well as those that antagonize it (capsazepine, BCTC, A-889425, and AMG9810). Our results demonstrate that capsaicin, capsazepine, BCTC, and A-889425 are effective Ca_v3 channel inhibitors at concentrations well below those used clinically to produce analgesia (Brandt, Beyer, & Stahl, 2012). Interestingly, Palvanil, AMG9810, and resiniferatoxin did not modulate Ca_v3 channels despite being structurally and chemically related to the other TRPV1-active compounds tested (Figure 2a). In the TRPV1 cryo-EM structures, overlapping density for both capsaicin and resiniferatoxin describes a common vanilloid binding pocket within the voltage-sensing domain, enclosed by the S3 and S4 transmembrane segments (Cao, Liao, Cheng, & Julius, 2013). Given that resiniferatoxin does not inhibit Ca_v3 channels, this suggests a different binding site from that described as the vanilloid binding pocket in TRPV1.

4.1 Capsaicin and capsazepine modulate DRG LVA calcium currents

We have shown that 100 μM capsaicin or capsazepine inhibits >95% of the LVA I_Ca in rat DRGs neurons of 30–40 μm diameter which typically belong to the Aδ (thinly myelinated) high- and low-threshold mechanoreceptors (Fang, McMullan, Lawson, & Djouhri, 2005; Figure 1a). However, in contrast to capsaicin's effect, capsazepine inhibition of LVA I_Ca in primary sensory neurons essentially (Figure 1a, i,ii). Previous studies have proposed that capsazepine inhibits voltage-gated calcium channels (VGCC) directly (Docherty et al., 1997), and furthermore, it has been proposed that, in rat DRG neurons, capsaicin inhibits VGCCs in a TRPV1-dependent manner (Hagenacker et al., 2005; Wu et al., 2005). An earlier study showed ~10% inhibition of LVA I_Ca in rat DRG by 0.1 μM capsaicin; however, higher concentrations of this drug were not tested (Hagenacker et al., 2005). Similarly, capsazepine effects have been examined over high voltage-activated calcium currents in rat DRG neurons (IC₅₀ = 1.4–7.7 μM, Wu et al., 2005) with similarly slow off kinetics as reported in this study. In order to unequivocally ascertain the direct effects of these two compounds on T-type channels, we show their inhibitory actions on heterologously expressed Ca_v3s in mammalian cells.

4.2 TRPV1-active compounds inhibit Ca_v3 channels directly

Agonists and antagonists of TRPV1 channels negatively modulated recombinant Ca_v3.1-3 channel currents. Capsaicin, capsazepine, BCTC, and A-889425 were shown to directly inhibit Ca_v3-mediated currents (Figure 2). Furthermore, differential effects of other TRPV1-active molecules like Palvanil, resiniferatoxin, or AMG9810 that were unable to modulate Ca_v3 function can rule out major endogenous TRPV1-mediated effect in our heterologous system. Previous studies have described irreversible T-type Ca²⁺ current inhibition by low concentrations of TRPV1-agonists in DRG neurons (Comunanza et al., 2011; Kerckhove et al., 2014). In this study, we show that this inhibition depends strictly on Ca²⁺, likely via Ca²⁺ entry through capsaicin-activated TRPV1 triggering intracellular feedback mechanisms as proposed elsewhere (Cazade et al., 2017; Comunanza et al., 2011; Kerckhove et al., 2014). Therefore, in tissues where TRPV1 and Ca_v3 channels coexist, low concentrations of TRPV1 activators (AM404 and capsaicin) irreversibly decrease T-type current by a TRPV1-mediated intracellular Ca²⁺ increase. At high concentrations, capsaicin inhibits Ca_v3 currents directly and reversibly, as reflected by the apparent lower affinity estimated in the present study. Thus, the mechanisms of inhibition of Ca_v3 channels by capsaicin can be distinct but both lead to the loss of T-type current, which in turn supports their contribution in the analgesic effects of TRPV1 channels activators.

The kinetics of block by capsaicin, A-889425, and BCTC were amenable to IC₅₀ determination (Figure 2c), whereas capsazepine's kinetics were too slow for reliable determination of an IC₅₀, hence we estimated its K_D from on and off rates of block (Table 2). A-889425 was the most potent inhibitor across the Ca_v3 family, followed by BCTC and capsaicin. There were, however, differences in the selectivity patterns across Ca_v3s. Both capsaicin and BCTC exerted their stronger effects on Ca_v3.1 channels whereas A-889425 had the highest potency for Ca_v3.3.

Analyses of Ca_v3 biophysical properties in the presence of capsaicin, A-889425, and BCTC's shed light on the inhibitory mechanism of action of these small molecules. Consistently, all three compounds decreased the availability of the three T-type channel isoforms by a strong delay in their recovery from inactivation, a leftward shift of inactivation, and shifting activation to more positive potentials.

Endogenous cannabinoids, including anandamide, have been shown to directly inhibit T-type calcium channels and to activate TRPV1 (Chemin, Monteil, Perez-Reyes, Nargeot, & Lory, 2001). Anandamide has a long unsaturated fatty acyl chain and thus is chemically similar to many vanilloids. A detailed study of the molecular determinants of anandamide inhibition showed that potency increased with the level of unsaturation in the alkyl chain of various derivatives (Chemin, Nargeot, & Lory, 2007). Consistent with this report, we observed that unsaturated capsaicin (Figure 2a) blocks Ca_v3 channels, whereas fully saturated Palvanil (Figure 2a) did not exhibit any obvious modulatory effects on T-type currents. Other TRPV1-active compounds including N-arachidonoyl dopamine, N-arachidonoyl 5-HT, and AM404 have also been shown to inhibit Ca_v3 channels directly (Gilmore, Heblinski, Reynolds, Kassiou, & Connor, 2012; Kerckhove et al., 2014; Ross, Gilmore, & Connor, 2009). The novel ortho-phenoxylanilide derivative, MONIRO-1, also slows recovery from inactivation kinetics of Ca_v3 channels (McArthur et al., 2018) suggesting a common mechanism of action between structurally related small molecules on Ca_v3 channels.

4.3 TRPV1 actions and Ca_v3 inhibition

Direct comparison of IC₅₀s between TRPV1 and Ca_v3 channels, as measures of their potency, of the aforementioned compounds reveals a clear preference of these substances for TRPV1 channels. For example, A-889425 and BCTC display 10- to 100-fold higher affinity for TRPV1 channels than for Ca_v3s (Chaudhari et al., 2013; McGaraughty et al., 2008). With the caveat that our K_D estimate for capsazepine's inhibition of T-type channels does not necessarily represent an equilibrium reaction, the values obtained for Ca_v3.1 and 3.2 are threefold to sixfold lower than the reported concentration for half-maximal inhibition of TRPV1-mediated currents (Caterina et al., 1997).

Capsaicin activates TRPV1 channels at concentrations 30- to 1,000-fold (McIntyre et al., 2001; Voets et al., 2004) lower than those inhibiting Ca_v3 channels directly. Strikingly, when used as an analgesic, topically applied capsaicin doses range between ~3 and 260 mM (0.1–8%; Derry & Moore, 2012; Noto et al., 2009). At such high concentrations, it can be assumed that all the available T-type calcium currents may be inhibited by the drug, either due to indirect inhibition by Ca²⁺ influx through TRPV1 or directly. Considering that Ca_v3s, and in particular Ca_v3.2 channels, have been implicated as key regulators of neuronal excitability in the periphery and as bona fide pharmacological targets in the treatment of pain. It is tempting to infer that at least some of the analgesic effects attributed to capsaicin are mediated via the reduction of Ca_v3.2 channel currents in sensory neurons.

To our knowledge, this is the first systematic study of the modulation of analgesic TRPV1-active compounds over T-type calcium channels in native and recombinant systems. We demonstrate that several of these small molecules inhibit Ca_v3 channels by a common mechanism based on decreasing the availability of the T-type calcium channels at physiologically relevant potentials. Furthermore, our results propose an alternative explanation as to why TRPV1 antagonist and its agonist capsaicin can be analgesic in various pain assays. At clinically relevant doses, inhibition of Ca_v3s in sensory neurons may constitute a component of analgesia. Thus, these findings offer an alternative explanation to the apparent disparate analgesic effects of the TRPV1 agonist capsaicin.

4.4 A cautionary note on capsaicin-induced pain in animal models

Capsaicin and other small molecules have been used experimentally to induce and estimate hyperalgesia as well as allodynia. Following intradermal and topical application, capsaicin initiates nociceptive C-fibre activity and within seconds to minutes, the skin sensitivity is enhanced causing pain upon which desensitization may ensue. Thus, capsaicin is considered an important tool in the study of pain and it is used to probe the analgesic potential of various compounds, despite capsaicin-induced pain model's propensity to significant variability, in estimating allodynia and hyperalgesia. Here, we show that capsaicin both directly and indirectly inhibits Ca_v3 channels emphasizing that perhaps it is not surprising that selective Ca_v3 channel drugs, such as ABT-639, are not analgesic in the intradermal capsaicin pain model in which the routinely used capsaicin doses likely abolish all T-type currents (Wallace, Duan, Liu, Locke, & Nothaft, 2016). This may explain the apparent failure of potential T-type channel active drugs in alleviating capsaicin-induced pain and raises a cautionary note towards the inadequacy of this assay to evaluate the analgesic effects of T-type calcium channel drugs.

5 CONCLUSION

In this study, we demonstrated cross-reactivity between TRPV1-active analgesic compounds (agonists and antagonists) and Ca_v3 isoforms in rat DRG neurons and heterologously expressed human Ca_v3 isoforms. The findings raise the possibility that some of the analgesic effects elicited by the aforementioned compounds may arise through inhibition of Ca_v3 channels in the periphery and may explain why both TRPV1 agonists and antagonists can be analgesic in various pain assays. We provide a possible mechanism to explain why some T-type channel drugs fail to alleviate capsaicin-induced pain models and why this analgesic assay may not be a suitable for the study of T-type calcium channel-inhibiting drugs. Although capsaicin and BCTC are >100-fold more selective for TRPV1 over Ca_v3 channels, A-889425 is only 10- to 100-fold less potent, whereas capsazepine is more selective for hCa_v3.1 and hCa_v3.2 over TRPV1 largely due to an extremely slow off rate. At clinical concentrations of capsaicin (topical application in the mM range), T-type calcium channels may also be inhibited. These results highlight the potential of using TRPV1 channel drugs as templates from which to design more selective and potent inhibitors of Ca_v3 channels as novel analgesic therapeutic compounds.

ACKNOWLEDGEMENTS

We are grateful to Dr Alexander A. Harper for insightful discussions and comments of this manuscript. This work was supported by the National Health and Medical Research Council (NHMRC) Program Grant (APP1072113) to D.J.A., and the Illawarra Health and Medical Research Institute (IHMRI) career development grant to J.R.M.

AUTHOR CONTRIBUTIONS

J.R.M. and D.J.A. conceived and design the research. J.R.M. performed experiments. J.R.M. and R.K.F.U. analysed and interpreted the data. All authors reviewed, revised, and approved the final paper.

CONFLICT OF INTEREST

The authors declare no conflicts of interest.

DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for Design & Analysis, and Animal Experimentation, and as recommended by funding agencies, publishers and other organizations engaged with supporting research.