5 Dec 2013: Andersson DA, Gentry C, Light E, Vastani N, Vallortigara J, et al. (2013) Correction: Methylglyoxal Evokes Pain by Stimulating TRPA1. PLOS ONE 8(12): 10.1371/annotation/e707d50a-13b3-4cc3-b507-7d8360d8f048. doi: 10.1371/annotation/e707d50a-13b3-4cc3-b507-7d8360d8f048 View correction
Diabetic neuropathy is a severe complication of long-standing diabetes and one of the major etiologies of neuropathic pain. Diabetes is associated with an increased formation of reactive oxygen species and the electrophilic dicarbonyl compound methylglyoxal (MG). Here we show that MG stimulates heterologously expressed TRPA1 in CHO cells and natively expressed TRPA1 in MDCK cells and DRG neurons. MG evokes [Ca2+]i-responses in TRPA1 expressing DRG neurons but is without effect in neurons cultured from Trpa1−/− mice. Consistent with a direct, intracellular action, we show that methylglyoxal is significantly more potent as a TRPA1 agonist when applied to the intracellular face of excised membrane patches than to intact cells. Local intraplantar administration of MG evokes a pain response in Trpa1+/+ but not in Trpa1−/− mice. Furthermore, persistently increased MG levels achieved by two weeks pharmacological inhibition of glyoxalase-1 (GLO-1), the rate-limiting enzyme responsible for detoxification of MG, evokes a progressive and marked thermal (cold and heat) and mechanical hypersensitivity in wildtype but not in Trpa1−/− mice. Our results thus demonstrate that TRPA1 is required both for the acute pain response evoked by topical MG and for the long-lasting pronociceptive effects associated with elevated MG in vivo. In contrast to our observations in DRG neurons, MG evokes indistinguishable [Ca2+]i-responses in pancreatic β-cells cultured from Trpa1+/+ and Trpa1−/− mice. In vivo, the TRPA1 antagonist HC030031 impairs glucose clearance in the glucose tolerance test both in Trpa1+/+ and Trpa1−/− mice, indicating a non-TRPA1 mediated effect and suggesting that results obtained with this compound should be interpreted with caution. Our results show that TRPA1 is the principal target for MG in sensory neurons but not in pancreatic β-cells and that activation of TRPA1 by MG produces a painful neuropathy with the behavioral hallmarks of diabetic neuropathy.
Citation: Andersson DA, Gentry C, Light E, Vastani N, Vallortigara J, Bierhaus A, et al. (2013) Methylglyoxal Evokes Pain by Stimulating TRPA1. PLoS ONE 8(10): e77986. doi:10.1371/journal.pone.0077986
Editor: Alexander G. Obukhov, Indiana University School of Medicine, United States of America
Received: July 31, 2013; Accepted: September 16, 2013; Published: October 22, 2013
Copyright: © 2013 Andersson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The work was supported by grants from the Medical Research Council (G0500847), Deutsche Forschungsgemeinschaft (BI-1281/3-1; NA 138/7-1) and the Dietmar-Hopp-Stiftung. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Sensory neuropathy is a common complication of type 1 and 2 diabetes and one of the major causes of neuropathic pain. The percentage of patients with neuropathy is generally considered to be about 50% for patients with long-standing diabetes and the prevalence is rising as a consequence of the increased global burden of type 2 diabetes . The major symptom of diabetic polyneuropathy is sensory loss, but many patients suffer from a variety of abnormal sensations including paraesthesias, thermal and mechanical hypersensitivities, and spontaneous pain. Several processes have been suggested to contribute to the development of diabetic neuropathy. Prominent among these are oxidative stress and an increased formation of reactive carbonyl and dicarbonyl metabolites from glucose during hyperglycemia , . We and others have shown that a number of lipid peroxidation products and reactive oxygen species (ROS) exert a direct pronociceptive action by stimulating an ion channel, TRPA1, expressed in nociceptive sensory neurons –. Although these reactive molecules do occur at increased levels during diabetes –, increased production of these chemical factors is also seen in other pathologies that are not primarily characterized by sensory neuropathies (e.g. Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis and atherosclerosis) . However, hyperglycemic episodes in diabetic patients are associated with an increased formation of the electrophilic reactive dicarbonyl compound methylglyoxal (MG). MG is primarily formed from triose intermediates in situations associated with increased glycolysis, such as hyperglycemia and fasting , . During hyperglycemia, cells with an insulin independent glucose uptake (of particular importance in neurons) can experience dramatic increases in the rate of glycolysis . Under these conditions, MG and other dicarbonyl containing glucose metabolites covalently modify proteins, giving rise to advanced glycation endproducts (AGE). The formation of AGE can directly interfere with the normal function of proteins, but also stimulates intracellular signaling pathways through activation of Receptor for AGE . Here we confirm earlier reports that methylglyoxal (MG), one of the major dicarbonyl metabolites produced from glucose during hyperglycemia, stimulates the nociceptive ion channel TRPA1 ,  directly through a reversible, intracellular interaction. In vivo, we show that pharmacologically elevated MG levels produce painful hypersensitivities to thermal and mechanical stimulation in wildtype, but not in Trpa1−/− mice. Our findings mirror clinical observations well, since altered sensitivity to thermal stimuli (hot and cold) produced by damage to small diameter nerve fibers is a reliable indicator that can distinguish between painful and painless diabetic neuropathy . Together, these results demonstrate that increased concentrations of MG, in the absence of hyperglycemia, produce a sensory neuropathy similar to that seen in diabetic patients and in experimental animal models of diabetes.
DRG neurons were prepared from adult male or female mice using methods described previously . Isolated neurons were maintained in MEM AQ (5.6 mM glucose) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 µg/ml streptomycin and 50 ng/ml NGF (Promega, Southampton, UK) for less than 24 hours before experimentation.
Mouse islets of Langerhans were isolated from the exocrine pancreas by collagenase digestion (1 mg/ml, type XI, Sigma, Poole, UK). Islets were dissociated by 15 min incubation in trypsin (0.5%, Sigma, Poole, UK) followed by trituration through a flame polished glass Pasteur pipette and the islet cells were seeded on glass cover slips.
MDCK cells were grown in DMEM (Invitrogen) supplemented with 5% fetal bovine serum, 100 U/ml penicillin and 100 µg/ml streptomycin.
Untransfected CHO cells and CHO cells expressing mouse or human TRPA1 were grown in MEM Glutamax or AQ medium supplemented with penicillin (100 U/ml), streptomycin (100 µg/ml), and FCS (10%). Plasmids encoding mutated TRPA1 channels were kindly provided by Dr. Ardem Patapoutian . All media, serum and antibiotics were from Invitrogen (Paisley, UK) or Sigma (Poole, UK).
Cells were loaded with 2 µM Fura-2 AM (Molecular Probes, Paisley, UK) in the presence of 1 mM probenecid for ∼1 hr. The experiments were performed in a physiological saline solution containing (in mM) 140 NaCl, 5 KCl, 10 glucose, 10 HEPES, 2 CaCl2, and 1 MgCl2, buffered to pH 7.4 with NaOH. Compounds were applied to cells by local continuous microperfusion of solution through a fine tube placed in close proximity of the cells studied. Images of a group of cells were captured every 2 sec using 340 and 380 nm excitation wavelengths with emission measured at 520 nm with a microscope based imaging system (PTI, New Jersey). Analyses of emission intensity ratios at 340 nm/380 nm excitation (R, in individual cells) were performed using the ImageMaster suite of software.
96 well plate assays
Changes in intracellular calcium ([Ca2+]i) in response to agonists were determined in TRPA1 expressing CHO cells and MDCK cells using a Flexstation 3 (Molecular Devices). Cells grown in 96 well black walled plates (Costar, Tewksbury, MA) were loaded with Fura 2-AM in the presence of 1 mM probenecid at 37°C for 1–1.5 hours and assays were carried out at 25°C. Basal emission ratios (340 nm/380 nm) were measured and changes in ratio determined at various times after compound addition.
DRG neurons and TRPA1 CHO cells were studied under voltage-clamp conditions using an Axopatch 200B amplifier and pClamp 10.0 software (Molecular Devices, Sunnyvale, CA). Except where stated otherwise, whole cell recordings were performed at a holding potential of –60 mV using an extracellular solution with the composition described above for [Ca2+]i measurements. Borosilicate glass pipettes (2–5 MΩ, 75–80% series resistance compensation) were filled with (in mM) 140 KCl, 1 CaCl2, 2 MgATP, 10 EGTA, and 10 HEPES buffered to pH 7.4 (KOH). Excised inside-out patches were recorded using a Ca2+-free solution containing (in mM) 140 NaCl, 5 KCl, 1 MgCl2, 10 HEPES and 1 EGTA, pH 7.4 (NaOH) in the pipettes and perfused with 140 KCl, 2 MgATP, 1 EGTA, and 10 HEPES (pH 7.4). DRG neurons were studied using an intracellular solution containing (in mM) 140 CsCl, 1 CaCl2, 2 MgATP, 10 EGTA and 10 HEPES, pH 7.4 (CsOH). Drugs were applied by local microperfusion with a rapid solution changer (RSC-200, Biologic, Claix, France).
All animal studies were carried out according to the U.K. Home Office Animal Procedures (1986) Act (project license PPL 70/0740) and were approved by the King’s College London Ethical Review Panel. Mice were killed by cervical dislocation.
Data shown are from male and female homozygote Trpa1−/− and Trpa1+/+. The Trpa1-null mice and wild-type littermates were bred from heterozygotic mice provided by Drs. Kelvin Kwan and David Corey (Harvard Medical School, Boston, MA) .
Mechanical thresholds were measured using an Analgesymeter (Ugo-Basile, Milan). Mice were kept in their holding cages to acclimatize (10–15 min) to the experimental room. The experimenter then lightly restrained the mouse and applied an increasing pressure stimulus to the dorsal surface of the hind paw using a blunt conical probe. The nociceptive threshold was defined as the force in grams at which the mouse withdrew its paw. In order to avoid tissue injury a 150 g force cut-off value was used.
Tactile allodynia was assessed by measuring withdrawal thresholds to calibrated von Frey filaments (0.008–2 g). Animals were placed in a Perspex chamber with a metal grid floor giving access to the underside of their paws and allowed to acclimatize prior to the start of the experiment. Von Frey filaments were applied perpendicular to the mid plantar surface of the hind paw with sufficient force to cause slight bending against the paw and held for a few seconds. This was repeated 5 times at intervals of 1–2 s. A positive response was noted if the paw was sharply withdrawn or there was flinching upon removal of the hair. If no response was noted a higher force hair was tested. Using a modified ‘up-down’ method the lowest force filament producing a positive response was recorded as the threshold.
Heat and cold sensitivities were assessed by measuring the time for paw withdrawal from a 50°C hotplate or 10°C coldplate (Ugo Basile, Milan) of lightly restrained mice .
Methylglyoxal (250 nmoles in 25 µl saline) was injected subcutaneously into the plantar surface of one of the hind paws using a 50 µl luer-syringe (Hamilton Co.) fitted with a 26-gauge x 3/8 inch intradermal needle. Immediately after injection, mice were placed inside a Perspex chamber and the duration of the pain-related behaviors (licking and biting or flinching and shaking of the injected paw) recorded using a digital stop-watch. Total pain response times over the first 2 minutes were used for analysis as the pain behaviors were largely confined to this period. The glyoxalase 1 inhibitor Sr-p-Bromobenzylglutathione cyclopentyl diester (50 mg/kg) was dissolved in 2% DMSO/20% Tween 80 in saline and administered by intraperitoneal (i.p.) injection (5 ml/kg).
The glucose tolerance test was performed in mice (not fasted) treated i.p. with 2 g/kg glucose or vehicle (0.2 ml, 0.9% saline). HC030031 (100 mg/kg) or vehicle (DMSO 10%, Tween 80 5%, 0.9% saline 85%) was administered i.p. 30 min before glucose or vehicle. Glucose concentrations were determined using a Contour blood glucose meter (Bayer Diabetes Care) in blood samples taken from the tail-tip.
Formalin fixed plantar skin was embedded in paraffin wax and cut in 8 µm sections and stained with rabbit polyclonal anti-PGP9.5 (1 in 5000, Ultraclone Ltd, UK) after citric acid heat induced epitope retrieval. Staining was visualized using anti-rabbit IgG labeled with Alexa-594. The number of PGP9.5 positive fibers crossing the basal membrane into the epidermis was determined in n = 6 mice for each group on a Zeiss Axioplan 2 microscope. The observer was blinded to the treatment.
Drugs and chemicals
The glyoxalase 1 inhibitor Sr-p-Bromobenzylglutathione cyclopentyl diester was provided by Dr. Thomas Fleming and was prepared and characterized as previously described , . AP18, 4-(4-chlorophenyl)-3-methylbut-3-en-2-one oxime, was from Maybridge (Tintagel, UK). Methylglyoxal and all other reagents were from Sigma (Poole, UK).
Methylglyoxal stimulates heterologously expressed TRPA1
We examined the effect of MG on [Ca2+]i in untransfected CHO cells and CHO cells expressing TRPA1. MG evoked concentration dependent increases in [Ca2+]i in cells expressing mouse TRPA1 (EC50 = 0.9±0.1 mM) or human TRPA1 (EC50 = 0.7±0.2 mM; Fig. 1A), but had no effect on [Ca2+]i in untransfected CHO cells or in the absence of extracellular Ca2+ (data not shown), demonstrating that MG evokes [Ca2+]i-responses by stimulating Ca2+-influx through TRPA1. In whole cell voltage-clamp experiments, MG produced inward currents that displayed a current waveform with Ca2+-mediated potentiation and inactivation similar to that previously observed with other TRPA1 agonists (Fig. 1B; , ). These results agree well with those of earlier reports characterizing MG as a TRPA1 agonist , , . Since methylglyoxal is a reactive, short-lived molecule formed intracellularly ,  and the TRPA1 residues targeted by electrophilic agonists are located intracellularly , , we examined the effect of MG applied to the intracellular face of excised inside-out patches (Fig. 1CD). In this configuration, MG rapidly stimulated TRPA1 channel activity and was markedly more potent than when applied extracellularly (see Fig. 1A). MG stimulated TRPA1 channel activity with an EC50 of 74 µM in the inside-out configuration (Fig. 1D), strongly suggesting that methylglyoxal acts at an intracellular site. TRPA1 channels activated by methylglyoxal in excised inside-out patches displayed a voltage dependent inactivation or block at positive membrane potentials (Fig. 1E), similar to that observed with other TRPA1 agonists . The single channel conductance was larger at positive membrane potentials (120 pS at +60 mV and 150 pS at +100 mV) and somewhat smaller (70-80 pS) in the physiological voltage range (–20 to –60 mV; Fig. 1F), consistent with the properties of TRPA1 .
(A) Application of MG evokes concentration-dependent [Ca2+]i-responses in CHO cells expressing mouse or human TRPA1. (B) MG (0.5 mM) elicits inward currents with the characteristic rapid onset and inactivation in the presence of 2 mM Ca2+ (top panel). In Ca2+-free solutions (1 mM EGTA) the inward current grew more slowly and addition of Ca2+ produced the typical current surge followed by a rapid inactivation of the current (bottom panel). (C) MG rapidly activates TRPA1 in excised inside-out patches, with a markedly higher potency than when applied extracellularly (D, compare with A). (E) The current voltage relationship in membrane patches containing several channels demonstrated a reversal potential close to 0 mV and channel block or inactivation at positive membrane potential. (F) Current-voltage relationship for a single channel in an inside-out patch stimulated with MG.
Methylglyoxal stimulates TRPA1 in MDCK cells
MG has previously been reported to stimulate [Ca2+]i-influx through an unknown mechanism in Madin-Darby Canine Kidney (MDCK) cells . We used [Ca2+]i-measurements to determine the involvement of TRPA1 in MG evoked [Ca2+]i-responses in MDCK cells. As illustrated in Fig. 2A, MG (5 mM) and the TRPA1 agonists AITC (100 µM), cinnamaldehyde (1 mM) and hydrogen peroxide (20 mM) all evoked [Ca2+]i-influx responses in MDCK cells (Fig. 2A). In concentration-response experiments, MG evoked [Ca2+]i-responses with an EC50 of 2.2 ± 0.4 mM and AITC produced an EC50 value of 9.1 ± 2.0 µM (Fig. 2BC). AP18 is a selective and potent inhibitor of TRPA1 , which likely acts through a saturable allosteric rather than a competitive mechanism . AP18 (10 µM) produced similar rightward shifts of the concentration-response curves for AITC and MG in MDCK cells, indicating a common target in TRPA1 (Fig. 2BC).
(A) The TRPA1 agonists AITC, H2O2, MG and cinnamaldehyde stimulate [Ca2+]i-responses in MDCK cells. (B, C) The selective TRPA1 antagonist AP18 (10 µM) produces a rightward shift of AITC and MG evoked [Ca2+]i-responses in MDCK cells.
TRPA1 mediates methylglyoxal responses in DRG neurons
TRPA1 is a sensory neuron transduction channel for pronociceptive irritant chemicals and inflammatory mediators ,  and we next determined whether MG stimulates nociceptive neurons directly. Similar to the results described by Eberhardt et al., (2012), we found that MG induced [Ca2+]i-responses in 72% (240 of 333) of capsaicin-sensitive (TRPV1 positive) DRG neurons cultured from Trpa1+/+ mice, but failed to increase [Ca2+]i in DRG neurons (0%, 1 of 322) cultured from Trpa1−/− mice (Fig. 3AB). We also examined the effect of MG on DRG neurons using voltage clamp recordings (Fig. 3C). In a proportion of DRG neurons MG evoked robust inward currents (in solutions containing 15 µM Ca2+ which prevents the Ca2+ mediated current surge , ) that were inactivated by readmission of 2 mM Ca2+ in the same way as currents evoked by other electrophilic TRPA1 agonists in DRG neurons , .
(A) Pseudocoloured images illustrating [Ca2+]i-concentrations in Fura-2 loaded cultured DRG neurons. MG evokes [Ca2+]i-responses in TRPA1 containing neurons. (B) Typical examples of [Ca2+]i-responses in capsaicin sensitive DRG neurons from Trpa1+/+ and Trpa1−/− mice. (C) MG activates characteristic TRPA1 currents in DRG neurons (holding potential -60 mV). Before addition of 2 mM Ca2+ this neuron was superfused with a solution containing 15 µM Ca2+, which prevents the Ca2+-induced current surge.
Site of action
Electrophilic TRPA1 agonists stimulate the channel through covalent modification of N-terminal cysteine residues , . To determine whether the TRPA1 cysteine residues that have previously been identified as targets for covalent modification by electrophilic agonists influence the sensitivity to MG, we examined the effect of MG on human TRPA1 channels containing the single point mutations C621S, C641S or C665S (Fig. 4A). None of these amino acid substitutions affected the potency of MG significantly, but each substitution very substantially reduced the TRPA1 sensitivity to AITC (Fig. 4B). Although many of the 31 cysteine residues present in TRPA1 can be modified by electrophilic agents , a limited number of these residues are required for channel activation , . Furthermore, it is likely that the relative importance of different residues varies between agonists .
The cysteine substitutions C665S, C641S and C621S (in human TRPA1) do not significantly affect responses evoked by MG (A), but dramatically reduce the sensitivity to AITC (B). (C) The lysine substitution K712Q (mouse TRPA1) reduced [Ca2+]i-responses evoked by AITC (50 µM). (D) The K712Q mutation reduced the amplitude of [Ca2+]i-responses evoked by MG (2 mM) and Δ9-tetrahydrocannabiorcol (20 µM; C16) in individual CHO cells (n = 22–64). (E) Outward TRPA1 current responses (+60 mV) evoked by MG (1 mM) in CHO cells decay relatively slowly when MG is removed (recorded under Ca2+-free conditions). Application of DTT (2 mM) produces a rapid, partial current inactivation. (F) Currents evoked by AITC (50 µM) under the same conditions inactivate relatively slowly, but are resistant to DTT and leave TRPA1 refractory to stimulation. (G) Currents elicited by the oxidant H2O2 (5 mM) remain at a stable level after removal of H2O2, but are rapidly and reversibly inactivated by DTT. Data were analyzed by ANOVA followed by Dunnett’s post-hoc test (panel B) or by t-test (*P<0.05, **P<0.01, ***P<0.001 compared to the wildtype channel).
In addition to the N-terminal cysteine residues, lysine K712 (in mouse) has been implicated as a target for electrophilic TRPA1 agonists . We therefore examined the effect of the point mutation K712Q on agonist responses evoked by MG, AITC and the non-covalent agonist Δ9-tetrahydrocannabiorcol . The amplitude of [Ca2+]i-responses evoked by AITC (50 µM), MG (2 mM) and Δ9-tetrahydrocannabiorcol in cells expressing mTRPA1 K712Q were all significantly reduced compared to those in WT mTRPA1 (Fig. 4CD). The fact that the sensitivity to all three agonists was reduced in the K712Q mutant indicates that this residue is important for the functional integrity of TRPA1 and does not exclusively act as a nucleophilic acceptor for electrophilic compounds.
It is possible that MG stimulates TRPA1 through formation of reversible hemithioacetals with cysteine residues other than those examined here , . To investigate this possibility we examined the effect of the reducing agent dithiothreitol (DTT, 2 mM) on outward TRPA1 currents recorded in the absence of Ca2+ (5 mM BAPTA intracellularly, 1 mM EGTA extracellularly) to avoid Ca2+-mediated inactivation or desensitization. We used DTT to determine whether disulfide bridge formation or reversible noncovalent interactions contribute to the agonist activity of MG. DTT is expected to rapidly reduce cystine bridges to free thiols, but also to form hemithioacetals with MG, thereby accelerating current decay. Covalent modification of arginine and lysine residues by MG and modification of cysteine residues by AITC are on the other hand resistant to DTT . Under these recording conditions, MG rapidly evoked an outward TRPA1 current that partially decayed with time after removal of MG (Fig. 4E), suggesting a reversible interaction between MG and TRPA1. Application of DTT immediately after MG removal dramatically accelerated the TRPA1 current decay consistent with a reversible interaction (Fig. 4E). When current responses evoked by a pharmacologically active concentration of H2O2 were examined in the same way, it was notable that they remained at a stable amplitude after removal of H2O2, but that application of DTT rapidly reduced the current amplitude (Fig. 4G). This is consistent with the expected oxidative formation of cystine bridges and their reduction by DTT . The contrast between the stable currents produced by H2O2 and the spontaneously reversible currents evoked by MG highlights the reversible nature of the interaction between MG and TRPA1. Currents evoked by AITC declined with time after removal of AITC, but were resistant to treatment with DTT (Fig. 4F). We also observed that repeated applications of MG and H2O2 evoked reproducible current responses of similar amplitudes, whereas an application of AITC prevented subsequent channel activation by AITC challenges (Fig. 4C-E). We conclude that MG acts through a reversible interaction with TRPA1 and that H2O2 relies on DTT sensitive oxidative modification of TRPA1 protein for its activity. These results strongly suggest that MG primarily acts through hemithioacetal formation as concluded by Eberhardt et al. . In contrast, the covalent modification produced by AITC appears irreversible and renders TRPA1 refractory to subsequent activation.
MG evokes pain and neuropathy by stimulating TRPA1
Local administration of TRPA1 agonists evokes acute pain – and we therefore examined whether this is also the case with MG. Trpa1+/+ and Trpa1−/− mice were given intraplantar injections of MG (250 nmoles/paw) and the time spent dedicated to pain-related behaviors during the 2 min following injection was quantified. MG induced a significant nociceptive response in Trpa1+/+ mice, but was without effect in Trpa1−/− mice, confirming that TRPA1 is required for the pronociceptive effect of MG in vivo, even after administration of a relatively high bolus dose (Fig. 5A).
(A) Intraplantar injections of MG (250 nmoles in 25 µl) evokes nociceptive behaviors in wildtype mice but is without effect in Trpa1−/− mice (n = 6). Injections of the GLO-1 inhibitor Sr-p-Bromobenzylglutathione cyclopentyl diester (GLOI, 50 mg/kg, 1h before test) did not acutely affect the withdrawal threshold in the paw pressure (B) or von Frey (C) tests compared to mice treated with vehicle (n = 6). (D-G) Effect of GLOI or vehicle injections administered every 2 days during a 2 week treatment period on the paw withdrawal latency from a 50°C hotplate (D), 10°C coldplate (E), the paw withdrawal threshold in the paw pressure test (F) and in response to stimulation with von Frey filaments (G). The data points are mean ± SEM from n = 6 mice, except for the last time point (15 days after the final injection) where n = 3. Data were analyzed by t-test (panel A) or ANOVA followed by Tukey’s HSD test (**P<0.01, ***P<0.001 compared to vehicle treated group).
Although acute administration of MG identified a potential role for TRPA1 as a sensor of elevated MG levels, it is not an experimental situation that accurately reflects the long-lasting, widely increased MG levels associated with hyperglycemia in diabetes. To model this situation more closely, we made use of the selective glyoxalase-1 (GLO-1) inhibitor Sr-p-Bromobenzylglutathione cyclopentyl diester . GLO-1 is the rate-limiting enzyme responsible for the hydrolysis of the majority of MG formed in vivo and inhibition of this enzyme significantly increases MG concentration in cultured neuronal and non-neuronal cells, as well as in plasma and brain in vivo , . Acute inhibition of GLO-1 did not influence nociception, since the sensitivities in the paw pressure and von Frey filaments tests were unaffected 1 h after administration of the GLO-1 inhibitor (Fig. 5B,C).
To determine the effect of prolonged elevation of MG levels in vivo on nociception, we administered the GLO-1 inhibitor on 7 occasions over 2 weeks to Trpa1+/+ and Trpa1−/− mice . During the course of the experiment we examined the nociceptive responses to cold, heat and mechanical stimulation (using the paw-pressure test and punctate stimulation with von Frey filaments) on days following administration of the GLO-1 inhibitor. Consistent with the expected time-course of the pharmacologically induced increase in MG concentrations , we observed no increased sensitivity in any of the tests during the first 4 days of the study (Fig. 5D-G). This indicates that the GLO-1 inhibitor itself is well tolerated and is without any direct effect on nociception. After day 4, the sensitivity to all four test modalities was progressively increased in the Trpa1+/+, but not the Trpa1−/− mice, reaching a plateau after 1–2 weeks of dosing. The observed hypersensitivities were long-lasting, with no sign of reversal even two weeks after cessation of drug administration.
Inhibition of GLO-1 does not reduce the number of intraepidermal nerve fibers
The early stages of diabetic neuropathy are characterized by damage to small nerve fibers and a reduction of intraepidermal nerve fibers (IENFs) is a useful clinical diagnostic tool for sensory neuropathies of different etiologies , . We examined skin sections from Trpa1+/+ and Trpa1−/− mice treated with the GLO-1 inhibitor or vehicle for 2 weeks to assess whether the hypersensitivities produced by inhibition of GLO-1 were accompanied by a loss of IENFs, similar to that seen in mouse models of diabetes. Inhibition of GLO-1 failed to produce a significant reduction of the IENF density compared to that seen in vehicle treated mice in either genotype (Fig. 6AB), suggesting that 2 weeks of increased MG levels is not sufficient to damage IENFs. The density of IENFs was significantly lower in vehicle treated Trpa1−/− mice compared to wildtype littermates (Fig. 6B). It is possible that this anatomical difference between the two genotypes contributes to the reduced sensitivity to cold and mechanical stimuli in naïve TRPA1-null mice , .
(A) PGP9.5 positive nerve fibers crossing the basal membrane into the epidermis (the white arrow highlights an example) in 8 µm sections of plantar skin (scale bar = 20 µm). (B) Following two weeks treatment with the GLO-1 inhibitor (GLO-I, 50 mg/kg every other day), the number of intra-epidermal nerve fibers crossing into the epidermis/mm skin was unchanged in both Trpa1+/+ mice and in Trpa1−/− littermates. Data were analyzed by ANOVA followed by Tukey’s HSD test (n = 6 mice, *P<0.05 compared to vehicle treated Trpa1+/+ mice).
MG stimulates mouse pancreatic beta cells independently of TRPA1
‘MG formed from glucose has been suggested to stimulate basal insulin release in response to elevated glucose concentrations through activation of TRPA1 expressed in pancreatic β-cells . To identify a possible role for TRPA1 as a MG receptor in beta cells directly, we compared the effect of MG and AITC on beta cells cultured from Trpa1+/+ and Trpa1−/− mice.. AITC failed to elicit [Ca2+]i-responses (data not shown) in beta cells cultured from either wildtype or Trpa1−/− mice, suggesting that mouse beta cells do not express TRPA1. Intriguingly, MG evoked [Ca2+]i-responses in a subpopulation of tolbutamide sensitive beta cells, but the number of MG sensitive cells and the response amplitude evoked by MG were similar in cultures from Trpa1+/+ and Trpa1−/− mice, demonstrating that MG stimulates beta cells independently of TRPA1, at least in the mouse (Fig. 7AB).
(A) MG (0.5 mM) evokes indistinguishable [Ca2+]i-responses in β-cells from Trpa1+/+ (black) and Trpa1−/− (red) mice (n = 15-28). (B) Illustration of the time course of MG and tolbutamide (200 µM) evoked [Ca2+]i-responses in β-cells. (C, D) Effect of the TRPA1 antagonist HC030031 (HC, 100 mg/kg) or vehicle (Veh) administered 30 min before glucose or vehicle in the glucose tolerance test (2 g/kg, i.p.) performed in Trpa1+/+ (C) and Trpa1−/− mice (D). Data points are mean ± s.e.m of n = 6 mice and were analyzed by ANOVA followed by Tukey’s HSD test (*P<0.05, **P<0.01, ***P<0.001 compared to the group treated with vehicle and glucose, †P<0.05 compared to the group treated with vehicle and vehicle).
We also examined whether TRPA1 influences blood glucose levels in vivo. In this experiment we examined the effect of the TRPA1 antagonist HC030031 (100 mg/kg, i.p.) on the elevation in blood glucose evoked by administration of glucose (2 g/kg in 0.2 ml, i.p.). Administration of glucose produced similar increases in blood glucose concentration in mice of both genotypes (Fig. 7CD). A notable finding was that administration of HC030031 dramatically enhanced the increase in blood glucose seen following a glucose challenge, both in Trpa1+/+ and in Trpa1−/−, suggesting that this antagonist interferes with glucose metabolism independently of TRPA1. Moreover, HC030031 also increased the blood glucose in mice that were not challenged with glucose (Fig. 7CD).
We have demonstrated that MG stimulates natively expressed TRPA1 in DRG neurons and MDCK cells and that acute administration of MG evokes pain mediated by TRPA1. We exploited a pharmacological inhibitor of GLO-1 to raise the levels of MG for a period of 2 weeks  and observed a progressive development of hypersensitivities to cold, heat and mechanical stimuli. Critically, we found that TRPA1 is required for the development of painful hypersensitivities and our results thus demonstrate that persistent stimulation of TRPA1 is sufficient to produce some of the well-established sensory characteristics of painful diabetic sensory neuropathy.
Glucose uptake is insulin-independent in sensory neurons and can therefore vary dramatically with fluctuations in plasma glucose . Increased glucose flux during hyperglycemia in turn leads to a dramatic increase in the intracellular formation of MG , . Increased formation of MG during hyperglycemia in diabetic patients is considered a major causal factor for the development of diabetic sensory neuropathy and other long-term diabetic complications , , , . It has recently been proposed that MG elicits sensory neuron hypersensitivity by modifying the properties of a sub-type of voltage gated sodium channel, NaV1.8, . Our results demonstrate that a direct effect of MG on NaV1.8 is not sufficient to produce pain or hypersensitivity in the absence of TRPA1, but suggest that actions at both TRPA1 and NaV1.8 are required for the behavioral sensitization observed following treatment with a GLO-1 inhibitor.
The chemotherapeutic agents paclitaxel and oxaliplatin produce persistent, TRPA1-dependent pain and hypersensitivities , , similar to our results with GLO-1 inhibition. Together, these findings are consistent with the notion that TRPA1 is central for the development of several common and clinically important classes of painful metabolic or pharmacologically induced neuropathies.
MG as a TRPA1 agonist
MG evokes [Ca2+]i-responses in cultured DRG neurons, MDCK cells and TRPA1 expressing CHO cells by stimulating TRPA1. Studies of MG modification of proteins have shown that MG forms reversible hemithioacetals with cysteine residues but stable covalent glycation products with lysine and in particular, arginine residues . Our results are consistent with a reversible interaction between MG and cysteines, since application of DTT largely reversed TRPA1 currents evoked by MG . This conclusion is further supported by the fact that application of MG to the intracellular face of excised patches stimulates TRPA1 with an apparent affinity that is close to estimates for the interaction between MG and N-acetylcysteine . Our observation that the amino acid substitutions C621S, C641S, and C665S did not affect the potency of MG whereas each mutation substantially reduced the TRPA1 sensitivity to AITC, is consistent with MG stimulating TRPA1 through formation of hemithioacetals , a mechanism distinct from that underlying the agonist effect of AITC . It is possible that cysteine residues other than those examined here are targeted by MG and the possibility that unrelated cysteines can compensate for the loss of the mutated cysteines involved in MG induced disulfide formation cannot be excluded . In agreement with the results presented here, C641 has been identified as a critical residue for the agonist activity of AITC , but this residue has also been shown not to influence the sensitivity of TRPA1 to cinnamaldehyde . Taken together, these observations therefore suggest that different electrophilic cysteine-reactive TRPA1 agonists do not interact with an identical subset of TRPA1 cysteine residues . The small DTT resistant component of the MG evoked current may involve covalent modification of arginine or lysine residues . Lysine 712 (K710 in human TRPA1) has previously been identified as a target for AITC, but in our hands appeared to compromise normal channel function, since it reduced TRPA1 sensitivity to MG, AITC and the non-covalently acting agonist Δ9-tetrahydrocannabiorcol to a similar extent.
TRPA1 is expressed in the gastrointestinal tract , but to our knowledge this is the first demonstration of functional TRPA1 channels in a cell line derived from the kidney. Expression of TRPA1 in MDCK cells suggests that some agents that have previously been shown to increase [Ca2+]i in MDCK cells, may do so by stimulating TRPA1 , .
MG has been proposed to regulate insulin release in vivo and in rat pancreatic β-cells by stimulating TRPA1 . Our results demonstrate that this is not the case in the mouse. Although MG evoked [Ca2+]i-responses in mouse β-cells, these were identical in cultures from wildtype and TRPA1-null mice and thus independent of TRPA1 channels. We further found that the TRPA1 antagonist HC030031 reduced glucose clearance in the glucose tolerance test very significantly in both Trpa1+/+ and Trpa1−/− mice. HC030031 must therefore interfere with glucose tolerance independently of TRPA1 and this off-target effect suggests that results obtained with this compound should be interpreted with caution.
TRPA1 as a mediator of diabetic neuropathy
Our results show that sustained elevations of endogenously formed MG by pharmacological inhibition of GLO-1 leads to the development of a painful diabetic neuropathy-like condition in mice. Activation of TRPA1 is essential for the observed sensory changes, since mice lacking TRPA1 appear unaffected by MG. TRPA1 and GLO1 are preferentially expressed in small-diameter, peptidergic DRG neurons ,  and the GLO1 activity is reduced in DRG neurons from diabetic mice . Although our results imply that increased formation of MG is sufficient to produce painful diabetic sensory neuropathy, it is likely that increased levels of reactive oxygen species and lipid peroxidation products also contribute to the development of sensory neuropathy in diabetes , , , .
Although diabetic neuropathy is a common cause of neuropathic pain, loss of sensation is the major clinical symptom. It is possible that a loss of peripheral sensory nerve fibers as a consequence of persistent activation of TRPA1 also contributes to the development of insensate features of diabetic sensory neuropathy . We did not detect a loss of IENFs or a loss of sensation and it is possible that degeneration of epidermal fibers takes longer than the two weeks that we examined in this study. It is also possible that long-lasting inhibition of GLO-1 produced higher levels of MG than would be the case in diabetes mellitus. Hyperglycemia reduces the cellular concentration of reduced glutathione in DRG neurons, leaving the cells more vulnerable to oxidative insults . Since glutathione is a necessary co-factor for GLO1, episodes of hyperglycemia or oxidant stress would also reduce the cellular capacity to convert MG to lactate . Reactive oxygen species and electrophilic metabolites are potent endogenous TRPA1 agonists  and are important for the development of diabetic neuropathy , . Oxidative stress and increased formation of MG thus converge on the nociceptive ion channel TRPA1 during hyperglycemia, thereby selectively stimulating nociceptive fibers.
Conceived and designed the experiments: SB DAA EL TF AB. Performed the experiments: DAA CG EL NV SB JV. Analyzed the data: DAA CG EL NV SB JV. Contributed reagents/materials/analysis tools: TF AB. Wrote the paper: DAA EL SB.
- 1. Edwards JL, Vincent AM, Cheng HT, Feldman EL (2008) Diabetic neuropathy: Mechanisms to management. Pharmacol Ther.
- 2. Frischmann M, Bidmon C, Angerer J, Pischetsrieder M (2005) Identification of DNA adducts of methylglyoxal. Chem Res Toxicol 18: 1586–1592.
- 3. Tomlinson DR, Gardiner NJ (2008) Glucose neurotoxicity. Nat Rev Neurosci 9: 36–45.
- 4. Andersson DA, Gentry C, Moss S, Bevan S (2008) Transient receptor potential A1 is a sensory receptor for multiple products of oxidative stress. J Neurosci 28: 2485–2494.
- 5. Cruz-Orengo L, Dhaka A, Heuermann RJ, Young TJ, Montana MC, et al. (2008) Cutaneous nociception evoked by 15-delta PGJ2 via activation of ion channel TRPA1. Mol Pain 4: 30.
- 6. Macpherson LJ, Xiao B, Kwan KY, Petrus MJ, Dubin AE, et al. (2007) An ion channel essential for sensing chemical damage. J Neurosci 27: 11412–11415.
- 7. Taylor-Clark TE, Undem BJ, Macglashan Jr DW, Ghatta S, Carr MJ, et al.. (2007) Prostaglandin-induced activation of nociceptive neurons via direct interaction with TRPA1. Mol Pharmacol.
- 8. Trevisani M, Siemens J, Materazzi S, Bautista DM, Nassini R, et al. (2007) 4-Hydroxynonenal, an endogenous aldehyde, causes pain and neurogenic inflammation through activation of the irritant receptor TRPA1. Proc Natl Acad Sci U S A 104: 13519–13524.
- 9. Blanco M, Moro MA, Davalos A, Leira R, Castellanos M, et al. (2005) Increased plasma levels of 15-deoxyDelta prostaglandin J2 are associated with good outcome in acute atherothrombotic ischemic stroke. Stroke 36: 1189–1194.
- 10. Calabrese V, Mancuso C, Sapienza M, Puleo E, Calafato S, et al. (2007) Oxidative stress and cellular stress response in diabetic nephropathy. Cell Stress Chaperones 12: 299–306.
- 11. Coughlan MT, Mibus AL, Forbes JM (2008) Oxidative stress and advanced glycation in diabetic nephropathy. Ann N Y Acad Sci 1126: 190–193.
- 12. Forbes JM, Coughlan MT, Cooper ME (2008) Oxidative stress as a major culprit in kidney disease in diabetes. Diabetes 57: 1446–1454.
- 13. Aldini G, Dalle-Donne I, Facino RM, Milzani A, Carini M (2007) Intervention strategies to inhibit protein carbonylation by lipoxidation-derived reactive carbonyls. Med Res Rev 27: 817–868.
- 14. Ahmed N, Babaei-Jadidi R, Howell SK, Thornalley PJ, Beisswenger PJ (2005) Glycated and oxidized protein degradation products are indicators of fasting and postprandial hyperglycemia in diabetes. Diabetes Care 28: 2465–2471.
- 15. Turk Z, Nemet I, Varga-Defteardarovic L, Car N (2006) Elevated level of methylglyoxal during diabetic ketoacidosis and its recovery phase. Diabetes Metab 32: 176–180.
- 16. Bierhaus A, Nawroth PP (2009) Multiple levels of regulation determine the role of the receptor for AGE (RAGE) as common soil in inflammation, immune responses and diabetes mellitus and its complications. Diabetologia 52: 2251–2263.
- 17. Cao DS, Zhong L, Hsieh TH, Abooj M, Bishnoi M, et al. (2012) Expression of transient receptor potential ankyrin 1 (TRPA1) and its role in insulin release from rat pancreatic beta cells. PLoS One 7: e38005.
- 18. Eberhardt MJ, Filipovic MR, Leffler A, de la Roche J, Kistner K, et al. (2012) Methylglyoxal activates nociceptors through transient receptor potential channel A1 (TRPA1): a possible mechanism of metabolic neuropathies. J Biol Chem 287: 28291–28306.
- 19. Kramer HH, Rolke R, Bickel A, Birklein F (2004) Thermal thresholds predict painfulness of diabetic neuropathies. Diabetes Care 27: 2386–2391.
- 20. Bevan S, Winter J (1995) Nerve growth factor (NGF) differentially regulates the chemosensitivity of adult rat cultured sensory neurons. J Neurosci 15: 4918–4926.
- 21. Hu H, Bandell M, Petrus MJ, Zhu MX, Patapoutian A (2009) Zinc activates damage-sensing TRPA1 ion channels. Nat Chem Biol 5: 183–190.
- 22. Kwan KY, Allchorne AJ, Vollrath MA, Christensen AP, Zhang DS, et al. (2006) TRPA1 contributes to cold, mechanical, and chemical nociception but is not essential for hair-cell transduction. Neuron 50: 277–289.
- 23. Gentry C, Stoakley N, Andersson DA, Bevan S (2010) The roles of iPLA2, TRPM8 and TRPA1 in chemically induced cold hypersensitivity. Mol Pain 6: 4.
- 24. Vince R, Daluge S, Wadd WB (1971) Studies on the inhibition of glyoxalase I by S-substituted glutathiones. J Med Chem 14: 402–404.
- 25. Lo TW, Thornalley PJ (1992) Inhibition of proliferation of human leukaemia 60 cells by diethyl esters of glyoxalase inhibitors in vitro. Biochem Pharmacol 44: 2357–2363.
- 26. Nagata K, Duggan A, Kumar G, Garcia-Anoveros J (2005) Nociceptor and hair cell transducer properties of TRPA1, a channel for pain and hearing. J Neurosci 25: 4052–4061.
- 27. Koivisto A, Hukkanen M, Saarnilehto M, Chapman H, Kuokkanen K, et al. (2012) Inhibiting TRPA1 ion channel reduces loss of cutaneous nerve fiber function in diabetic animals: sustained activation of the TRPA1 channel contributes to the pathogenesis of peripheral diabetic neuropathy. Pharmacol Res 65: 149–158.
- 28. Phillips SA, Thornalley PJ (1993) The formation of methylglyoxal from triose phosphates. Investigation using a specific assay for methylglyoxal. Eur J Biochem 212: 101–105.
- 29. Thornalley PJ (1993) The glyoxalase system in health and disease. Mol Aspects Med 14: 287–371.
- 30. Hinman A, Chuang HH, Bautista DM, Julius D (2006) TRP channel activation by reversible covalent modification. Proc Natl Acad Sci U S A 103: 19564–19568.
- 31. Macpherson LJ, Dubin AE, Evans MJ, Marr F, Schultz PG, et al. (2007) Noxious compounds activate TRPA1 ion channels through covalent modification of cysteines. Nature 445: 541–545.
- 32. Nilius B, Prenen J, Owsianik G (2011) Irritating channels: the case of TRPA1. J Physiol 589: 1543–1549.
- 33. Jan CR, Chen CH, Wang SC, Kuo SY (2005) Effect of methylglyoxal on intracellular calcium levels and viability in renal tubular cells. Cell Signal 17: 847–855.
- 34. Petrus M, Peier AM, Bandell M, Hwang SW, Huynh T, et al. (2007) A role of TRPA1 in mechanical hyperalgesia is revealed by pharmacological inhibition. Mol Pain 3: 40.
- 35. Defalco J, Steiger D, Gustafson A, Emerling DE, Kelly MG, et al. (2010) Oxime derivatives related to AP18: Agonists and antagonists of the TRPA1 receptor. Bioorg Med Chem Lett 20: 276–279.
- 36. Bautista DM, Jordt SE, Nikai T, Tsuruda PR, Read AJ, et al. (2006) TRPA1 mediates the inflammatory actions of environmental irritants and proalgesic agents. Cell 124: 1269–1282.
- 37. Wang YY, Chang RB, Waters HN, McKemy DD, Liman ER (2008) The nociceptor ion channel TRPA1 is potentiated and inactivated by permeating calcium ions. J Biol Chem 283: 32691–32703.
- 38. Takahashi N, Mizuno Y, Kozai D, Yamamoto S, Kiyonaka S, et al. (2008) Molecular characterization of TRPA1 channel activation by cysteine-reactive inflammatory mediators. Channels (Austin) 2: 287–298.
- 39. Andersson DA, Gentry C, Alenmyr L, Killander D, Lewis SE, et al. (2011) TRPA1 mediates spinal antinociception induced by acetaminophen and the cannabinoid Delta(9)-tetrahydrocannabiorcol. Nat Commun 2: 551.
- 40. Lo TW, Westwood ME, McLellan AC, Selwood T, Thornalley PJ (1994) Binding and modification of proteins by methylglyoxal under physiological conditions. A kinetic and mechanistic study with N alpha-acetylarginine, N alpha-acetylcysteine, and N alpha-acetyllysine, and bovine serum albumin. J Biol Chem 269: 32299–32305.
- 41. Andersson DA, Gentry C, Moss S, Bevan S (2009) Clioquinol and pyrithione activate TRPA1 by increasing intracellular Zn2+. Proc Natl Acad Sci U S A 106: 8374–8379.
- 42. Bandell M, Story GM, Hwang SW, Viswanath V, Eid SR, et al. (2004) Noxious cold ion channel TRPA1 is activated by pungent compounds and bradykinin. Neuron 41: 849–857.
- 43. Jordt SE, Bautista DM, Chuang HH, McKemy DD, Zygmunt PM, et al. (2004) Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTM1. Nature 427: 260–265.
- 44. McNamara CR, Mandel-Brehm J, Bautista DM, Siemens J, Deranian KL, et al. (2007) TRPA1 mediates formalin-induced pain. Proc Natl Acad Sci U S A 104: 13525–13530.
- 45. Thornalley PJ, Edwards LG, Kang Y, Wyatt C, Davies N, et al. (1996) Antitumour activity of S-p-bromobenzylglutathione cyclopentyl diester in vitro and in vivo. Inhibition of glyoxalase I and induction of apoptosis. Biochem Pharmacol 51: 1365–1372.
- 46. Thornalley PJ (2008) Protein and nucleotide damage by glyoxal and methylglyoxal in physiological systems—role in ageing and disease. Drug Metabol Drug Interact 23: 125–150.
- 47. Bierhaus A, Fleming T, Stoyanov S, Leffler A, Babes A, et al. (2012) Methylglyoxal modification of Nav1.8 facilitates nociceptive neuron firing and causes hyperalgesia in diabetic neuropathy. Nat Med 18: 926–933.
- 48. Devigili G, Tugnoli V, Penza P, Camozzi F, Lombardi R, et al. (2008) The diagnostic criteria for small fibre neuropathy: from symptoms to neuropathology. Brain 131: 1912–1925.
- 49. Sommer C, Lauria G (2007) Skin biopsy in the management of peripheral neuropathy. Lancet Neurol 6: 632–642.
- 50. Fleming T, Cuny J, Nawroth G, Djuric Z, Humpert PM, et al. (2012) Is diabetes an acquired disorder of reactive glucose metabolites and their intermediates? Diabetologia 55: 1151–1155.
- 51. Xiao WH, Zheng H, Bennett GJ (2012) Characterization of oxaliplatin-induced chronic painful peripheral neuropathy in the rat and comparison with the neuropathy induced by paclitaxel. Neuroscience 203: 194–206.
- 52. Nassini R, Gees M, Harrison S, De Siena G, Materazzi S, et al. (2011) Oxaliplatin elicits mechanical and cold allodynia in rodents via TRPA1 receptor stimulation. Pain 152: 1621–1631.
- 53. Eberhardt MJ, Filipovic MR, Leffler A, de la Roche J, Kistner K, et al.. (2012) Methylglyoxal activates nociceptors through TRPA1 - a possible mechanism of metabolic neuropathies. J Biol Chem.
- 54. Wang L, Cvetkov TL, Chance MR, Moiseenkova-Bell VY (2012) Identification of in vivo disulfide conformation of TRPA1 ion channel. J Biol Chem 287: 6169–6176.
- 55. Ahmed N, Dobler D, Dean M, Thornalley PJ (2005) Peptide mapping identifies hotspot site of modification in human serum albumin by methylglyoxal involved in ligand binding and esterase activity. J Biol Chem 280: 5724–5732.
- 56. Nozawa K, Kawabata-Shoda E, Doihara H, Kojima R, Okada H, et al. (2009) TRPA1 regulates gastrointestinal motility through serotonin release from enterochromaffin cells. Proc Natl Acad Sci U S A 106: 3408–3413.
- 57. Jan CR, Jiann BP, Lu YC, Chang HT, Yu CC, et al. (2003) Oxidation by thimerosal increases calcium levelsin renal tubular cells. Pharmacol Toxicol 93: 123–127.
- 58. Jan CR, Wu SN, Tseng CJ (1999) Zn2+ increases resting cytosolic Ca2+ levels and abolishes capacitative Ca2+ entry induced by ATP in MDCK cells. Naunyn Schmiedebergs Arch Pharmacol 360: 249–255.
- 59. Jack MM, Ryals JM, Wright DE (2011) Characterisation of glyoxalase I in a streptozocin-induced mouse model of diabetes with painful and insensate neuropathy. Diabetologia 54: 2174–2182.
- 60. Story GM, Peier AM, Reeve AJ, Eid SR, Mosbacher J, et al. (2003) ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell 112: 819–829.
- 61. Vincent AM, McLean LL, Backus C, Feldman EL (2005) Short-term hyperglycemia produces oxidative damage and apoptosis in neurons. FASEB J 19: 638–640.
- 62. Vincent AM, Callaghan BC, Smith AL, Feldman EL (2011) Diabetic neuropathy: cellular mechanisms as therapeutic targets. Nat Rev Neurol 7: 573–583.