Cell Physiology

Dynamic changes in the TRPA1 selectivity filter lead to progressive but reversible pore dilation

T. G. Banke, S. R. Chaplan, A. D. Wickenden


TRPA1 is a nonselective cation channel belonging to the transient receptor potential (TRP) family that is expressed in peripheral sensory neurons and may play important roles in pain perception and inflammation. We found that agonist stimulation of TRPA1, along with other members of the TRP family (TRPV1–4 and TRPM8), can induce the appearance of a large pore permeable to large organic cations such as Yo-Pro (YP) and N-methyl-d-glucamine, in an agonist and divalent cation-dependent manner. YP uptake was not inhibited by a panel of putative gap junction/pannexin blockers, suggesting that gap junction proteins are not required in this process. Our data suggest that changes in the TRP channel selectivity filter itself result in a progressive but reversible pore dilation process, a process that is under strong regulation by external calcium ions. Our data suggest that calcium plays a novel role in setting the amount of time TRPA1 channels spend in a dilated state providing a mechanism that may limit sensory neuron activation by painful or irritating substances.

  • transient receptor potential channels
  • gap junction proteins
  • pannexins
  • sensory neurons
  • pain

trpa1 is a nonselective cation channel that belongs to the transient receptor potential (TRP) superfamily (13, 37). TRPA1 was first identified as a transformation-sensitive mRNA in cultured human lung fibroblasts (26). Subsequent studies indicated that TRPA1 was also highly expressed in sensory neurons of the dorsal root, trigeminal and nodose ganglia, and in hair cells of the inner ear (14, 17, 38, 47). In sensory neurons, TRPA1 expression is most prevalent in small-diameter neurons where it colocalizes with markers of peptidergic nociceptors such as TRPV1, CGRP, and substance P but not TRPM8 (5, 17, 38, 41, 47).

A primary function of TRPs, including TRPA1, is the detection of chemical irritants. For example, TRPA1 can be activated by a range of pungent or irritant compounds such as mustard oil (allyl isothiocyanate, AITC), cinnamaldehyde, and acrolein. These compounds are highly reactive electrophiles that activate TRPA1 by reversible covalent modification of cysteine and lysine residues in the NH2 terminus of the channel (23, 33). Agonist activation of TRPA1 results in activation of a nonselective cation current and sensory neuron activation. Activation of these ionotropic chemoreceptors on sensory neurons probably underlies the acute physiological and behavioral responses to reactive chemicals such as pain, neurogenic inflammation, and reflex withdrawal.

Previous data with other chemosensitive ion channels such as TRPV1, and some members of the P2X family, e.g., P2X7, indicate that prolonged agonist stimulation can result in an increased permeability to larger cations such as tetraethylammonium and N-methyl-d-glucamine (NMDG), cationic dyes such as Yo-Pro (YP) and FM1–43 (12, 36, 40, 42), and small organic chemicals such as the sodium channel blocker QX-314 (7). This phenomenon is generally referred to as “pore dilation,” although whether it is the channel itself that dilates or whether a secondary accessory protein(s) (e.g., gap junction or/and pannexins) is required is currently unclear. Recently, evidence has also been presented to indicate that TRPA1 may also undergo pore dilation after agonist exposure (10); however, the mechanism is not well understood. Hence, the purpose of the present study was therefore to investigate the underlying mechanism(s) of large dye uptake after prolonged activation of TRPA1. We show that stimulation of TRPA1, along with TRPV1–4 and TRPM8, with various agonists does indeed trigger uptake of a large cationic dye. In all cases, dye uptake was unaffected by a range of gap junction/pannexin blockers, suggesting that these proteins are not involved in this process. Rather, dye uptake was temporally associated with changes in the biophysical properties of TRPA1, suggesting that pore dilation may be responsible. Dye uptake and the changes in biophysical properties are highly sensitive to divalent and trivalent cations. These observations suggest that the pore properties of TRPA1 and related TRPs are highly dynamic. Our findings also identify a novel mechanism by which calcium may regulate the function of TRP channels. Preliminary accounts of these findings have appeared previously in abstract form (3, 4).


Cell cultures and clones.

Human TRPA1 (sequence identical to accession no. NM_007332) was cloned into pcDNA4/TO. CHO-T-REx cells (Invitrogen, Carlsbad, CA) were stably transfected with pcDNA4/TO-TRPA1 using standard techniques to generate a clonal cell line that expressed human TRPA1 in a tetracycline-inducible manner. The culture medium was Ham's F-12 supplemented with 10% fetal bovine serum, 2 mM l-glutamine, 5 μg/ml blasticidin, and 200 μg/ml zeocin (34). Accession numbers for rat (r)TRPV1, human (h)TRPV1, hTRPV2, hTRPV3, hTRPV4, canine (c)TRPM8, and rPrX7 are DQ015702, NP_542436, NM_016113, NM_145068.2, NM_021625, DQ273165, and NM_019256, respectively. hTRPA1(D915A) (Biomyx, San Diego, CA) was transiently transfected into HEK-293 cells using Lipofectamine 2000 (Invitrogen).

Calcium fluorescence.

Intracellular Ca2+ concentrations were monitored using a fluorometric imaging plate reader (FLIPR, Molecular Devices, Sunnyvale, CA). Cells were seeded in black-walled clear-bottom 96-well plates at a density of 50,000 cells per well and cultured overnight at 37°C/5% CO2 in culture medium supplemented with 1 mg/ml tetracycline if appropriate. On the day of the experiment, cells were washed three times with HEPES-buffered saline (in mM: 137 NaCl, 0.5 MgCl2, 2 KCl, 5 dextrose, 2 CaCl2, and 10 HEPES; pH 7.4). Cells were then loaded with calcium-sensitive fluorescent dye by incubation in the presence of 4 μM Fluo-3 AM at room temperature in the dark for 60 min. After incubation with dye, cells were washed in assay buffer and, if appropriate, antagonists were added at this time. Following a further 30 min of incubation, cells were assayed in FLIPR. Changes in fluorescence were monitored for 3 min after the addition of test agonist.

Calcium response curves were generated by measuring calcium inflow at time (t) = 3 min after agonist application.

Yo-Pro influx assay.

Cells were prepared as described above. Cells were washed and the extracellular solution was replaced with a divalent-free HEPES-buffered solution containing Yo-Pro-1 (YP, Invitrogen) at a final concentration of 2 μM. Following a 15-min incubation period, agonists or antagonists were added and the time course of YP uptake was recorded using a plate reader (Spectra Max Gemini) at 490/520 nm. To test for involvement of gap junctions, pannexins, and kinases in the YP-uptake process, TRPA1-expressing cells were prepared as described above and pretreated with various pharmacological inhibitors (see Chemicals) After incubation for 30 min at 37°C, agonists were applied and the uptake of YP was measured. To test for possible cell death in our cultures, we tested AITC and farnesyl thiosalicylic acid (FTS) in a lactate dehydrogenase (LDH) release assay. Samples of extracellular solution were taken at 15, 30, 45, and 60 min after addition of FTS (3, 10, 30, and 100 μM), AITC (3, 10, 30, and 100 μM), or 0.1% Triton X-100 (as a positive control) to cell cultures, Significant cell death was only observed 15 min after addition of 100 μM FTS (data not shown).

To test for involvement of changes in cell volume, the osmolarity of the extracellular solution was adjusted (range: 200–400 mosM). Under such conditions, no significant uptake of Yo-Pro was observed (data not shown).

YP-uptake responses were normalized to the maximum response (typically the response elicited by the highest agonist concentration) measured at t = 20 min.


Cells for use in electrophysiological assays were plated at low density onto glass coverslips 24–36 h before recording and maintained in appropriate media. On the day of the experiment, glass coverslips were placed in a bath on the stage of an inverted microscope and perfused (approximately 1 ml/min) with extracellular solution of the following composition (in mM): 140 NaCl, 2 KCl, 0.5 MgCl2, 5 glucose, and 10 HEPES, pH 7.4. Pipettes were filled with an intracellular solution of the following composition (in mM): 140 CsCl, 4 NaCl, 10 HEPES, and 5 BAPTA, pH 7.4; pipettes when filled with solution had a resistance of 2 to 4 MΩ. The osmolarity was adjusted to 300 mosM with mannitol. For NMDG recordings, the extracellular solution contained the following (in mM): 150 NMDG, 10 HEPES, and 10 glucose, pH 7.4 (HCl); osmolarity was adjusted with mannitol (∼300 mosM).

To calculate PNMDG/PNa, we used variations of the Goldman-Hodgkin-Katz equation: PNMDG/PNa=[NMDG]o/[Na]i×exp(ΔErev×F/RT)

All recordings were made at room temperature (22–24°C) using a Multiclamp 700A amplifier and pClamp 9 software (Axon Instruments). Current records were acquired at 5 KHz and filtered at 2 KHz. Agonists were applied using an SF-77B Fast-Step Perfusion device (Warner Instruments). Voltage commands were not corrected for liquid junction potentials.

Statistical analysis.

Data are presented as means ± SE. Statistical significance was assessed using Student's t-test unless stated otherwise.


AITC, FTS, and URB597 (3′-carbamoyl-biphenyl-3-yl-cyclohexylcarbamate cyclohexylcarbamic acid-3′-carbamoyl-biphenyl-3-yl ester) were purchased from Sigma-Aldrich (St. Louis, MO), Biomol (Plymouth Meeting, PA), and Calbiochem (San Diego, CA), respectively. Yo-Pro-1 was purchased from Invitrogen. 4α-Phorbol 12,13-didecanoate (4αPDD), 2-aminoethoxydiphenyl borate (2APB), capsaicin, icilin, gap junction blockers [carbenoxolone (CBX), quinine (Quin), flufenamic acid (FFA), 5-nitro-2-(3-phenyl-propylamino)benzoic acid (NPPB), niflumic acid (NFA), probenecid (Prob), and mefloquine (Mef)] and kinase inhibitors [mitogen-activated protein kinase inhibitors (MAPK): PD98059, SB202474, SB202190, SB203580, U0124, U0125, and U0126], protein kinase A inhibitor (H-89) or protein kinase C inhibitors (GO6976 and GO6983) were all purchased from Sigma-Aldrich.


The aim of this study was to investigate the underlying mechanism(s) of uptake of large dyes in TRPA1 expressing cells. To confirm channel activation, we first determined whether AITC, FTS, and URB597, three known TRPA1 agonists, could increase intracellular calcium in hTRPA1-expressing CHO cells, using a fluorometric imaging plate reader (FLIPR). As previously reported (34, 39), all three agonists evoked a robust increase in intracellular calcium concentration with EC50 values of 2.1 ± 0.6 μM (n = 8), 4.2 ± 0.2 μM (n = 8) and 0.7 ± 0.1 μM (n = 8) for AITC, FTS and URB597, respectively. The normalized maximum fluorescence for AITC, URB, and FTS was 1:∼0.6:∼1, respectively (Fig. 1, BD). Using the same three compounds, we then tested for agonist-induced uptake of YP, a large non-cell-permeable organic cation with a molecular mass of 629 Da, into hTRPA1-expressing cells. As shown in Fig. 1, A and B, and Supplemental Table S1 (the online version supplemental data for this article can be found online at the American Journal of Physiology-Cell Physiology website), AITC induced YP uptake into hTRPA1-expressing cells in a concentration-dependent manner with an EC50 value of 6.8 ± 1.8 μM (n = 8). Similarly, FTS and URB597 also evoked YP uptake with EC50 values of 20 ± 2 μM (n = 8) and 42 ± 10 μM (n = 8), respectively (Fig. 1, C and D, and Supplemental Table S1). In general, the AITC, FTS, and URB597 EC50 values measured in the YP-uptake assay were higher than those measured in the calcium fluorescence assay, most noticeably for URB597 (Fig. 1C and Supplemental Table S1). In control experiments, AITC also induced ethidium bromide (2 μM) uptake in TRPA1-expressing cells (n = 14, data not shown), suggesting that agonist-induced uptake of fluorescent dyes is not a YP-dependent process. The TRPA1 antagonists HC030031 and AP-18 blocked AITC-evoked YP uptake with IC50 values of 5.2 ± 0.9 and 3.6 ± 0.7 μM (n = 6; Fig. 1E and Supplemental Table S2), respectively. AITC-induced YP uptake in hTRPA1-expressing cells was also strongly blocked by La3+ (Fig. 1E and Supplemental Table S2). Although the nonselective TRP channel blocker ruthenium red (RR) completely blocked agonist-evoked YP uptake in TRPA1-expressing cells, we found that this compound was capable of bleaching the YP signal (Supplemental Fig. S1) and thus was not useful for characterizing the mechanism of YP uptake. In control experiments, no YP uptake was observed in TRPA1-expressing cells in the absence of agonists. Neither was any uptake observed in the parental cell line, CHO-T-REx, after application of any of the three agonists (data not shown). No LDH release was observed from TRPA1 cells following agonist exposures sufficient to cause YP uptake (see materials and methods; data not shown). These results suggest that YP uptake is TRPA1 dependent and occurs through a mechanism distinct from cell lysis.

Fig. 1.

Activation of TRPA1 induces Yo-Pro (YP) uptake in a divalent cation-dependent manner. A: representative example of allyl isothiocyanate (AITC)-induced YP uptake in human (h)TRPA1-expressing cells. BD: agonist-induced YP uptake in TRPA1-expressing cells was concentration dependent as shown by the black symbols for AITC (B), URB597 (URB; C), and farnesyl thiosalicylic acid (FTS; D). The concentration dependence of agonist-induced calcium influx in hTRPA1 expressing cells is also shown for comparison (red circles). E: AITC (30 μM)-induced YP uptake was inhibited by the TRPA1 antagonists HC030031 (black circles), AP-18 (red triangles), and lanthanum (open blue squares). AITC (30 μM)-induced YP uptake in hTRPA1-expressing cells was also inhibited by divalent cations. F and G: representative examples of AITC-induced YP uptake in the presence of increasing concentrations of either magnesium (F) or calcium (G). H: for both divalent cations, the inhibition was concentration dependent for calcium (red circles) and magnesium (blue circles). Superimposed on the graph is the concentration-response curve for block of AITC-induced YP uptake by calcium in mutant D915A TRPA1-expressing cells (open squares, dotted line).

To determine whether large dye uptake was a unique feature of TRPA1 or a common feature of sensory TRP channels, we studied YP uptake in cells stably expressing hTRPV2, hTRPV3, hTRPV4, and cTRPM8. We also studied human and rat TRPV1, and rat P2X7-expressing cells, since dye uptake in response to prolonged agonist stimulation has previously been described for these ion channels (12, 42). Our results show that stimulation with a relevant agonist induced YP uptake in cells expressing TRPV1, TRPM8, and P2X7 (Fig. 2). In addition, stimulation of TRPV2–4 also induced YP uptake (Supplemental Table S1). EC50 values for YP uptake are similar to published EC50 values for channel activation in either fluorescent or electrophysiological assays (Fig. 2D and Supplemental Table S1) (11, 12, 22, 53). La3+ strongly blocked capsaicin-induced YP uptake in TRPV1-expressing cells and icilin-induced YP uptake in TRPM8-expressing cells with IC50 values of 74 ± 10 μM and 247 ± 10 μM, respectively (Supplemental Table S2), whereas Cu2+ blocked BzATP-induced YP uptake with an IC50 value of 4 ± 1 μM, (n = 9) in agreement with previous findings (51). In P2X7-expressing cells, dye uptake has been shown to be highly dependent on the extracellular divalent cation concentration (48). Therefore, we studied the effects of extracellular divalent cations on agonist-induced YP uptake in cells expressing TRP channels. The data shown in Fig. 1, FH, clearly show that YP uptake in TRPA1-expressing cells strongly depends on extracellular calcium and magnesium. Similar results were obtained from TRPV1-, TRPM8-, and P2X7-expressing cells (Supplemental Table S2). Closer inspection of the raw data indicated that divalent cations acted to delay the onset of YP uptake (not shown).

Fig. 2.

Activation of TRPV1, TRPM8, and P2X7 induces YP uptake. Shown are representative examples of YP uptake in hTRPV1 (A)-, canine (c)TRPM8 (B)-, and rat (r)P2X7 (C)-expressing cells following exposure to capsaicin (CAP), icilin, or BzATP, respectively. In each case, agonist-induced YP uptake was concentration dependent, as indicated by the concentration-response curves shown in D. Ctrl, control; norm, normalized.

Our findings show that agonist stimulation of TRPA1, other distantly related TRP channels, and the unrelated P2X7 channel results in YP uptake. The finding that YP uptake is strongly inhibited by divalent cations in all channels tested may suggest that a common mechanism underlies YP uptake. Furthermore, under conditions where YP uptake is strongly inhibited, i.e., in the presence of high divalent cation concentrations, these proteins are still able to function as calcium-permeable ion channels, suggesting that distinct pathways may be responsible for calcium and YP permeability.

We next asked whether additional proteins (e.g., gap junctions and/or pannexins) were required for large dye uptake, as shown by some investigators for P2X7 channels (42), or whether the TRPA1 channel protein itself is capable of allowing entry of larger molecules via a mechanism often referred to as pore dilation. To address the role of gap junctions and pannexins, we studied the ability of a panel of gap junction/pannexin blockers to modulate agonist-induced calcium entry and YP uptake in TRPA1-expressing cells. The gap junction/pannexin blockers used were CBX (16), Quin (46), FFA, NPPB (20), NFA (32), Prob (45), and Mef (15). In calcium fluorescence studies, CBX (1–30 μM) was without effect on intracellular calcium or AITC-induced calcium fluorescence. At higher concentrations, CBX produced a concentration-dependent increase in basal calcium fluorescence and occluded the response to AITC (Fig. 3, A and B). Similar calcium mobilizing effects of ≥100 μM CBX were observed in cells expressing other TRP channels (TRPV1 and TRPM8) and in cells expressing P2X7, suggesting that this compound may exert nonspecific effects on intracellular calcium levels at high concentrations. Indeed, the CBX-induced increase in calcium fluorescence was not blocked by preincubation with RR (Fig. 3C), and no TRPA1 agonist activity could be detected for CBX in whole cell patch-clamp experiments (Fig. 4D). Similar RR-insensitive, nonspecific increases in calcium fluorescence were observed for Quin, Mef, and Prob (Fig. 3, B and C), although these effects were only observed at very high concentrations of Prob (>300 μM) and Quin. NFA, NPPB, and FFA, on the other hand, increased calcium fluorescence in a RR-sensitive manner (Fig. 3, B and C) and direct RR-sensitive TRPA1 agonist activity was confirmed for FFA (EC50 = 2.8 ± 0.1 μM; n = 7), NPPB (EC50 = 5.8 ± 0.2 μM; n = 7), and NFA (EC50 = 25 ± 4 μM, n = 7) in whole cell patch-clamp experiments (Fig. 4) (see also Ref. 24). In YP-uptake studies, CBX (up to 100 μM), Prob (up to 1 mM), Mef (up to 100 μM), and Quin (up to 100 μM) did not stimulate dye uptake in the absence of agonist and had no significant effect on AITC-induced YP uptake (Fig. 5, A and B). Consistent with their TRPA1 agonist properties, NFA, FFA, and NPPB all stimulated YP uptake in the absence of other agonists (Fig. 5C). The rank order of agonist potency (NPPB>FFA>NFA) was the same for agonist-induced increases in calcium fluorescence, activation of inward currents in whole cell patch clamp, and for stimulation of YP uptake. The level of YP uptake ranged from 30 ± 4% AITC for FFA to 69 ± 9% for NPPB (Fig. 5C).

Fig. 3.

Effect of gap junction blockers on basal and AITC-induced increases in intracellular calcium in TRPA1-expressing cells. A: the effects of carbenoxolone (CBX) on intracellular calcium concentrations were measured before and after exposure of TRPA1-expressing cells to AITC (50 μM). Concentrations <100 μM did not evoke any significant effects. Higher concentrations of CBX induced an increase in resting intracellular calcium concentration and occluded the response to AITC. Changes in intracellular calcium concentrations were measured using the FLIPR-based Ca2+ assay. The timing of the CBX and AITC applications is indicated by gray and white bars, respectively. RFU, relative fluorescence units. B: 5-nitro-2-(3-phenyl-propylamino)benzoic acid (NPPB; open circles), niflumic acid (NFA; crossed hexagons), CBX (open squares), mefloquine (Mef; upward open triangles), probenecid (Prob; downward open triangles), FFA (closed squares), and quinine (Quin; closed stars) all induced concentration-dependent increases in resting intracellular calcium concentration in TRPA1-expressing cells. C: the increases in intracellular calcium concentration caused by NFA (100 μM), FFA (100 μM), NPPB (100 μM), and AITC (100 μM) were blocked by ruthenium red (RR), suggesting involvement of transient receptor potential (TRP) channels in this response. The increases in intracellular calcium concentration caused by Quin (100 μM), Prob (1 mM), CBX (100 μM), and Mef (100 μM), however, were not blocked by RR, suggesting a non-TRP-mediated effect.

Fig. 4.

RR-sensitive activation of TRPA1 by putative gap junction blockers. A, C, and E: NFA (50 μM; A), NPPB (10 μM; C), and FFA (10 μM; E) induced inward currents at a holding potential of −60 mV in whole cell voltage-clamp experiments, confirming direct activation of TRPA1 by these agents. Coapplication of 5 μM RR (indicated by a black bar at the top of each trace) completely blocked the inward current (left trace and shown on an expanded time axis in the middle traces of A, C, E). Right trace: current-voltage relationship from the traces shown to the left at time point before (black trace) and just after (green trace) RR application. Red trace is the current-voltage relationship from before channel activation with AITC. B, D, and F: application of NFA, NPPB, or FFA did not evoke any currents in control cells (T-REx CHO cells). G, left: pulses of NPPB-, FFA-, and AITC (concentration used as indicated)-evoked inward currents in a TRPA1-expressing cell. G, right: normalized (to 20 μM AITC) concentration-response curves for NFA, Quin, CBX, NPPB, and FFA as measured in whole cell voltage-clamp experiments. Quin and CBX did not activate TRPA1 in voltage-clamp studies.

Fig. 5.

YP uptake in hTRPA1-expressing cells is not prevented by gap junction blockers. A: representative traces showing AITC-evoked uptake of YP in TRPA1-expressing cells before and after preincubation with CBX (10–100 μM). Neither CBX, Quin, Prob, nor Mef had any significant effect on AITC (30 μM)-evoked YP uptake in TRPA1-expressing cells. B: average effect of 100 μM CBX, 100 μM Quin, 1 mM Prob, or 100 μM Mef on AITC (30 μM)-evoked YP uptake in TRPA1-expressing cells, measured 20 min after AITC application. C: in contrast, FFA and NPPB or NFA evoked uptake of YP in TRPA1-expressing cells. In each case, these effects were concentration dependent, but the maximum increases in YP uptake were lower than that seen with AITC (30 μM). D: similar to the findings in TRPA1-expressing cells, 100 μM CBX did not affect capsaicin (1 μM), icilin (100 μM), or BzATP (30 μM)-induced YP uptake in TRPV1 (V1)-, TRPM8 (M8)-, or P2X7-expressing cells, respectively.

In parallel studies, CBX had no effect on agonist-induced YP uptake in cells expressing TRPM8, TRPV1, or P2X7, respectively (Fig. 5D).

The finding that YP uptake occurs in the presence of a range of gap junction/pannexin blockers suggests that these proteins are not essential for agonist-induced dye uptake in TRP/P2X7-expressing cells. Furthermore, a wide range of kinase inhibitors (inhibitors of MAPK, PKA, and PKC; see materials and methods) were without effect on AITC-induced Yo-Pro uptake in TRPA1-expressing cells, suggesting that these kinases are not involved in the formation of the large pore (data not shown). These observations raise the possibility that dye uptake occurs via the channel itself, and thus we next focused on seeking evidence for agonist-induced changes in TRPA1 pore properties.

Previous studies have demonstrated time-dependent changes in ionic selectivity following agonist-induced activation of P2X7 and TRP channels (10, 12, 27), and it has been proposed that this pore dilation may be responsible for large dye uptake via this channel. We conducted ion substitution experiments in CHO cells expressing hTRPA1, using the whole cell voltage-clamp technique. In these experiments, standard extracellular cations were replaced with the large monovalent cation NMDG (molecular weight = 195) and Na+ was used as the major internal cation (150 mM Na+). Under these conditions a small, outwardly rectifying current could be recorded in the absence of TRPA1 agonists, consistent with a degree of constitutive activation of TRPA1 in our cell line (Fig. 6B; see also Ref. 10). Application of the TRP channel blocker La3+ (1 mM) completely shifted the reversal close to zero (Fig. 6B, bottom), strongly suggesting that the constitutive currents were TRPA1 in nature. The reversal potential of this pre-drug control current (Erev) was −85 ± 3.8 mV (n = 16) (Fig. 6, B and C). At a holding potential of −60 mV, application of 20 μM AITC or 30 μM URB597 evoked a slowly developing inward (NMDG) current (Fig. 6A). Similar currents were activated by 5 μM FTS (data not shown). Transient outward currents were not observed. The AITC-, FTS-, and URB-induced currents reversed at −47.4 ± 3.7 mV (n = 17), −40.0 ± 2.5 mV (n = 13), and −52.8 ± 2.6 mV (n = 9), respectively (Fig. 6, B and C). No time-dependent changes in Erev were apparent for URB, FTS, or AITC. The change in Erev from baseline after application of FTS, AITC, and URB597 corresponded to a 5.8 ± 0.9-, 7.6 ± 1.3-, and 3.9 ± 0.3-fold increase in the TRPA1 Na:NMDG permeability ratio, respectively (Fig. 6C).

Fig. 6.

AITC, FTS, and URB597 evoke N-methyl-d-glucamine (NMDG) currents in hTRPA1-expressing cells. A: representative traces showing 20 μM AITC (left)- and 30 μM URB597 (right)-induced inward NMDG currents in two different TRPA1-expressing cells. Cells were held at −60 mV, and ramps from (−100 mV to +100 mV) were evoked periodically. The timing of compound application is shown by the gray bars. Note that transient outward currents were not observed following agonist application. B: representative recordings from the cells shown in A (see arrow) illustrating ramp currents in the absence and presence of either AITC (left) or URB597 (right). In the absence of agonist, small outward currents are observed at all voltages positive to −80 mV, but following agonist exposure, outward currents are only observed at voltages positive to −40 mV. At the end of the experiment, application of 1 mM La3+ completely shifted the reversal potential (Erev) close to zero. Currents at top are shown on expanded axis at bottom. C, top: calculated reversal potentials for control (c), 20 μM AITC (A)-, 5 μM FTS (F)-, and 30 μM URB597 (U)-evoked currents in hTRPA1-expressing cells and for capsaicin-evoked currents in rTRPV1-expressing cells (V1). C, bottom: derived Na+:NMDG+ permeability (P) ratios. ***P < 0.001.

To further explore agonist- and time-dependent changes in the TRPA1 current-voltage (IV) relationship, we applied voltage ramps (−100 to +100 mV) every 5 s to TRPA1-expressing cells during application of AITC (20 μM) (Fig. 7A). In the absence of extracellular calcium, the AITC-evoked currents were at first strongly outwardly rectifying (Fig. 7B). However, the degree of rectification progressively declined over time and by the approximately third IV ramp, the IV relationship was almost linear (Fig. 7B). In the presence of 2 mM external calcium, AITC evoked an inward current that initially increased in exponential fashion followed by almost complete inactivation (Fig. 7, C and D). During the early phase of current activation, the degree of rectification changed from strongly outwardly rectifying to almost linear (compare 2 with 3 in Fig. 7E). However, as inactivation progressed and the current amplitude declined, the degree of rectification was restored (compare 3 with 6 in Fig. 7E). The changes in rectification properties are illustrated by plotting the normalized −100/+100 mV ratios measured for AITC-induced currents in the absence and presence of extracellular calcium. Figure 7H clearly shows the progressive reduction in the rectification ratio for TRPA1 in the absence of calcium, and the transient reduction in the ratio observed when calcium was present.

Fig. 7.

Agonist-induced, calcium-dependent changes in channel rectification properties. A: representative whole cell recording from a TRPA1-expressing cell. In the recording shown, the cell was voltage clamped at −60 mV in the absence of extracellular calcium and AITC-induced inward currents evoked by brief agonist exposure (black bars). The gray bar at top illustrates application of 1 mM La3+. During the recording, every 5 s a voltage ramp was applied. B: current-voltage (IV) plots derived from the ramps in A. The timing of the ramps is indicated by the numbers 15. As illustrated, the TRPA1 IV relationship is initially outwardly rectifying (sweep 2, red) but becomes linear with continuous exposure to AITC (sweeps 35, bright green, blue, muted green). C and D: similar experiments were conducted in the presence of extracellular calcium (2 mM) and a representative trace is shown (C) and on an expanded scale for clarity (D). E: IV plots from the recording shown in D at time points indicated with numbers. Similar to the recordings made in the absence of extracellular calcium, the IV plots are initially outwardly rectifying (sweep 2, red). However, during the continuous application of agonist, the IV relationship becomes linear at the time of peak inward current (sweeps 3, bright green and 4, blue) and then reverts to an inwardly rectifying current as the current desensitizes (sweeps 5, black and 6, muted green). F: similar recordings were made from cells expressing hTRPA1 (D915A) channels held at −60 mV in the presence of 2 mM external calcium. The timing of AITC application is indicated with black bars at top. The open bar at top illustrates application of 1 mM La3+. G: IV plots at the indicated times. In cells expressing hTRPA1 (D915A), the IV relationship is initially outwardly rectifying (sweep 2, red), but the degree of rectification decreases with continuous exposure to AITC, even in the presence of extracellular calcium (sweeps 35, bright green, blue, muted green). H: −100/+100 mV rectification ratios from wild-type (WT) and hTRPA1 (D915A)-expressing cells in the presence (WT, red circle; D915A, bright green circle) or absence (WT, blue square) of extracellular calcium. The zero time point was defined as first significant change in the rectification ratio (i.e., time point 3 in B, E, and G). Pt = 0 (WT, 0 mM Ca2+) = 0.024, n = 11; Pt = 0 (WT, 2 mM Ca2+) = 0.002, n = 7; Pt = 0 (D915A, 2 mM Ca2+) = 0.03, n = 5.

We also investigated whether other known TRPA1 agonists were capable of changing the rectification properties of TRPA1. For this we compared a single pulse of AITC with a single pulse of FFA (Fig. 8A). Application of AITC to TRPA1-expressing cells under calcium-free conditions evoked an outwardly rectifying current similar to that described in Fig. 7B. In contrast, a single application of FFA (30 μM) elicited a large, essentially linear current (Fig. 8, A and B). Similar results were obtained using other nonelectrophilic TRPA1 agonists, FTS (34) and NPPB (Fig. 8C). Interestingly, when FFA was washed off the cell and the same cell was then exposed to AITC, the IV relationship of the resulting current was once again outwardly rectifying (Fig. 8, A and C), indicating that any changes in rectification properties induced by FFA, FTS, or NPPB were fully and rapidly reversible.

Fig. 8.

Changes in rectification properties are reversible. A: representative traces showing 20 μM AITC- and 30 μM FFA-induced currents in a TRPA1-expressing cell, recorded in the absence of extracellular calcium. Inward currents were evoked with 15-s pulses (gray bar at top) of TRPA1 agonists. In the experiment shown, the order of agonist application was AITC (red trace), followed by FFA (blue trace), followed by AITC again (black trace), with washout between agonist applications. B: current-voltage relationships from the recordings shown in A. From the plots it can be seen that AITC-induced currents are outwardly rectifying in both instances. FFA-induced currents, however, are essentially linear. C: rectification ratios (−60/60) measured during the first application of AITC; during a subsequent application of either 30 μM FFA, 5 μM FTS, or 10 μM NPPB; and finally during the last application of AITC. These data show that the rectification ratio for AITC is significantly greater than for FFA, NPPB, and FTS. PFTS = 0.036; PFFA = 0.0001; PNPPB = 0.005.

These results provide further evidence for agonist-induced changes in the pore properties of TRPA1. Both changes in pore properties and YP uptake appear to be strongly regulated by extracellular divalent cations, suggesting that these two events may be related.

To explore the mechanism of calcium regulation further, we measured YP uptake and channel rectification properties in a TRPA1 mutant in which a putative calcium-binding residue in the pore-loop domain [aspartate 915, which is equivalent to aspartate 918 in mouse TRPA1 (54); aspartate 646 in TRPV1 and aspartate 682 in TRPV4 (53)], was replaced with an alanine (D915A). AITC, FTS, and URB597 potency, as measured in FLIPR, was reduced by four- to sixfold when tested on the D915A mutant (not shown).

As for wild-type (WT) channels, AITC induced YP uptake in D915A-expressing cells. However, in contrast to WT TRPA1, YP uptake in D915A-expressing cells was relatively insensitive to inhibition by extracellular calcium. The IC50 for inhibition of YP uptake by calcium was 1.9 ± 0.1 mM, (n = 6) for D915A compared with 75 ± 15 μM (n = 14) for WT TRPA1 (Fig. 1H and Supplemental Table S2). In whole cell patch-clamp experiments in the presence of calcium, neutralization of aspartate 915 largely removed calcium-induced inactivation, permitting repeated agonist applications without noticeable decline in current amplitude. As for WT currents, D915A currents were initially strongly outwardly rectifying, but, interestingly, the rectification ratio quickly diminished in D915A currents, and the reduction was sustained, even in the presence of extracellular calcium (Fig. 7, FH).


Here, we show that prolonged agonist application can evoke uptake of large dye molecules in TRPA1 expressing cells, in agreement with recent findings of Chen et al. (10). We also show that dye uptake following agonist stimulation is not unique to TRPA1, but also occurs following agonist exposure in cells expressing other TRP channels, including TRPV1–4 and TRPM8 (and P2X7). These findings confirm several previous reports of agonist-induced dye uptake. For example, dye uptake has been described as a result of agonist stimulation of several members of the ATP-gated ion channel, especially P2X2, P2X4, and P2X7 (27, 28, 52). Dye uptake also occurs following agonist stimulation of TRPV1 (10, 12). We were also able to show robust dye uptake in response to icilin in TRPM8-expressing cells, whereas Chen and colleagues were unable to evoke dye uptake in cells expressing this channel in response to menthol (34). The reasons for this discrepancy are not clear, but likely reflect minor differences in experimental reagents and/or protocols (10). The present study also extends earlier observations by demonstrating dye uptake in TRPV2- and TRPV4-expressing cells. In all cases studied (TRPA1, TRPV1, and TRPM8), dye uptake is strongly inhibited in the presence of physiological concentrations of divalent cations. Block of agonist-induced dye uptake has also been demonstrated to be highly dependent on the extracellular divalent cation concentration in P2X7-expressing cells (48).

On the basis of the findings presented here and in earlier publications, it would seem that dye uptake is a common, divalent cation-sensitive event that occurs following agonist stimulation of numerous ion channels. Given that TRPA1, TRPV, TRPM, and P2X channels share little sequence homology, we wondered whether dye uptake could be explained by the existence of accessory/partner proteins common to the expression systems commonly used in these studies.

Potential accessory/partner proteins implicated in dye uptake include gap junctions and related hemichannels (30, 42) and kinases (19). To test whether gap junctions/pannexins or kinases were involved, we studied the effects of a series of putative gap junction blockers, a series that is known for blocking not only a wide range of gap junctions but also pannexin proteins (15, 20, 32, 45, 46) and kinase inhibitors on dye uptake in cells expressing TRP channels. Our findings argue against a significant involvement of pannexins or kinases in agonist-induced dye uptake. First, agonist-induced YP uptake in TRPA1, TRPV1, and TRPM8 channels was not blocked by carbenoxolone or any of the other known gap junction blockers or kinase inhibitors tested here (see also Ref. 12). The ability of some of these compounds to evoke YP uptake, despite their gap junction/pannexin blocking activity, lends further weight to the argument that these partner/accessory proteins are not required for dye uptake. In contrast to previous reports (42), we could not block dye uptake in P2X7-expressing cells with carbenoxolone, suggesting that under our experimental conditions, neither TRP channels nor members of ATP-gated channels require carbenoxolone-sensitive proteins for dye uptake. We cannot, however, rule out the possibility that pannexins or kinases may play a role under different experimental conditions, and/or at different time points after agonist exposure or under scenarios where the pharmacology of the accessory proteins changes after being recruited by TRP channels.

Since recruitment of gap junctions/pannexins does not seem to explain dye uptake under the conditions of this study, we considered an alternative hypothesis whereby dye uptake occurs via the TRP channel itself, as a result of a process that has been termed “pore dilation.” Examination of TRPA1 pore properties under conditions conducive to YP uptake provide several pieces of evidence in support of this hypothesis. We show that TRPA1 agonist exposure leads to altered NMDG permeability and changed rectification properties. The changes in pore properties are reduced or delayed by extracellular divalent cations and are reversible. Similar changes in pore properties following agonist stimulation have been described for P2X7 (48), TRPV1 (12), TRPV3 (11), and TRPA1 (10).

TRPA1 channels are regulated in a complex manner by external calcium ions. Hence, on one hand, extracellular calcium potentiates TRPA1 activity through a mechanism involving calcium entry though TRPA1 and interaction with an intracellular site on the channel protein (18, 54, 57). On the other hand, extracellular calcium can induce strong TRPA1 inactivation (38, 54). The findings of the present study identify a third mechanism by which calcium can regulate TRPA1 activity, i.e., by preventing or delaying agonist-induced pore dilation. To investigate the mechanism of this novel aspect of calcium regulation, we examined the properties of a TRPA1 mutant in which a putative calcium-binding residue in the pore-loop domain (D915A) was neutralized. In keeping with an important role for this residue, dye uptake via the mutant channel was relatively insensitive to inhibition by extracellular calcium. The mutant channel also resembled the “dilated” WT channel in its reduced rectification, even in the presence of calcium, further supporting a key role for this residue in calcium regulation of channel pore properties. Whether the effects of neutralizing D915 are a direct consequence of reduced calcium binding within the pore domain or are secondary to reduced calcium permeability [as described for effects of D918A on potentiation and inactivation in mouse TRPA1 (54)] is not known. Interestingly, we found that the pore mutant (D915A) also exhibited modest reductions in agonist potency, suggesting that this residue may also play an important role in the transduction of agonist-induced conformation changes.

The finding that agonist stimulation of several distantly related TRP channels and the unrelated P2X7 channel leads to dye uptake by what appears to be a common, divalent-sensitive mechanism, despite there being minimal sequence homology among these channels, raises an obvious question as to the structural basis of this effect. One possible explanation is that the pore region of these channels exhibits a degree of flexibility that is normally constrained by binding of calcium ions within the pore region. Since TRP channels are structurally related to voltage-gated potassium channels, it is interesting to note that the pore region of Kv3.1 and Kv3.2b may be flexible and able to permit permeation of NMDG under potassium free conditions (55). Pore region flexibility has also been proposed for voltage-gated sodium channels (6, 50) and inward rectifier potassium channels (31). As a working hypothesis then, we propose a model where the pore regions of various calcium permeable channels are flexible but stabilized by binding of calcium to acidic residues including D915. Activation in the absence of calcium leads to increased flexibility of the pore, changing the dimensions of the selectivity filter and a progressive but reversible pore dilation.

Given that pore dilation and YP uptake appear to be highly regulated by divalent cations, it is interesting to speculate on the possible physiological relevance of pore dilation. In this respect, it is worth noting that although physiological concentrations of divalent cations block dye uptake to a considerable extent, the block is not complete and a small but consistent pool of TRP channels remain capable of YP uptake even in the presence of high concentrations of divalent cations. Furthermore, our data show that transient changes in pore properties/(rectification ratio) occur even in the presence of physiological concentrations of calcium, suggesting that the inhibitory properties of divalent cations requires channel activation and possibly divalent permeation. Some evidence also exists in the literature to suggest that pore dilation may occur under physiological conditions, For example, Binshtok et al. (7) found that in the presence of physiologically relevant divalent ion composition, coadministration of QX-314 with capsaicin promoted QX-314 uptake to its site of action on the intracellular side of DRG neurons, probably via a TRPV1 pore dilation mechanism. In addition, native TRPV1 channels in trigeminal ganglion neurons also undergo pore dilation after prolonged agonist application (12). These observations suggest that native TRP channels are capable of entering the dilated state even in the presence of normal extracellular calcium concentrations, making this process a potentially important aspect of TRPA1 signaling. What then is the role of pore dilation in TRP channel signaling? One possibility is that pore dilation adds an important temporal aspect to TRPA1 signaling, providing a mechanism for transient signal amplification during continuous exposure to potentially harmful chemicals. In this case, calcium would play a key role in setting the amount of time TRPA1 channels spend in a dilated state providing a mechanism that may limit sensory neuron activation by painful or irritating substances. By analogy with P2X7 channels, a second possibility is that large pore formation may be involved in various processes that follow TRP channel activation, such as cytokine and/or ATP release (2, 8, 9, 29, 43, 44, 49, 56), apoptosis/necrosis (1, 25), and cell differentiation and growth (26). Even though the relevance of our findings to drug discovery remains to be determined, it is likely that the channel protein(s) goes through a series of dramatic molecular rearrangements during the dilation process. During such a process, it is possible that the channel(s) may reveal new binding/regulatory sites normally hidden (deep) within the protein. Targeting such a site(s) could potential create interesting new state-dependent TRP channel blockers with greater specificity for channels under pathophysiologically relevant conditions (i.e., prolonged exposure to high agonist concentrations).


No conflicts of interest, financial or otherwise, are declared by the authors.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 22.
  22. 23.
  23. 24.
  24. 25.
  25. 26.
  26. 27.
  27. 28.
  28. 29.
  29. 30.
  30. 31.
  31. 32.
  32. 33.
  33. 34.
  34. 36.
  35. 37.
  36. 38.
  37. 39.
  38. 40.
  39. 41.
  40. 42.
  41. 43.
  42. 44.
  43. 45.
  44. 46.
  45. 47.
  46. 48.
  47. 49.
  48. 50.
  49. 51.
  50. 52.
  51. 53.
  52. 54.
  53. 55.
  54. 56.
  55. 57.
View Abstract