Relative contributions of TRPA1 and TRPV1 channels in the activation of vagal bronchopulmonary C-fibres by the endogenous autacoid 4-oxononenal
- T. E. Taylor-Clark1,
- M. A. McAlexander2,
- C. Nassenstein1,
- S. A. Sheardown3,
- S. Wilson3,
- J. Thornton3,
- M. J. Carr2 and
- B. J. Undem1
- 1Division of Allergy & Clinical Immunology, Johns Hopkins School of Medicine, Baltimore, MA, USA2Neuronal & Ion Channel Biology, Respiratory Drug Discovery, GlaxoSmithKline Pharmaceuticals, King of Prussia, PA, USA3Transgenics, Genotyping & Viral Vectors, GlaxoSmithKline Pharmaceuticals, Harlow, UK
- Corresponding author B. J. Undem. Division of Allergy & Clinical Immunology, Johns Hopkins School of Medicine, Baltimore, MA, USA. Email: bundem{at}jhmi.edu
Abstract
Transient receptor potential (TRP) A1 channels are cation channels found preferentially on nociceptive sensory neurones, including capsaicin-sensitive TRPV1-expressing vagal bronchopulmonary C-fibres, and are activated by electrophilic compounds such as mustard oil and cinnamaldehyde. Oxidative stress, a pathological feature of many respiratory diseases, causes the endogenous formation of a number of reactive electrophilic alkenals via lipid peroxidation. One such alkenal, 4-hydroxynonenal (4HNE), activates TRPA1 in cultured sensory neurones. However, our data demonstrate that 100 μm 4HNE was unable to evoke significant action potential discharge or tachykinin release from bronchopulmonary C-fibre terminals. Instead, another endogenously produced alkenal, 4-oxononenal (4ONE, 10 μm), which is far more electrophilic than 4HNE, caused substantial action potential discharge and tachykinin release from bronchopulmonary C-fibre terminals. The activation of mouse bronchopulmonary C-fibre terminals by 4ONE (10–100 μm) was mediated entirely by TRPA1 channels, based on the absence of responses in C-fibre terminals from TRPA1 knockout mice. Interestingly, although the robust increases in calcium caused by 4ONE (0.1–10 μm) in dissociated vagal neurones were essentially abolished in TRPA1 knockout mice, at 100 μm 4ONE caused a large TRPV1-dependent response. Furthermore, 4ONE (100 μm) was shown to activate TRPV1 channel-expressing HEK cells. In conclusion, the data support the hypothesis that 4-ONE is a relevant endogenous activator of vagal C-fibres via an interaction with TRPA1, and at less relevant concentrations, it may activate nerves via TRPV1.
Transient receptor potential (TRP) A1 is a non-selective cation channel found preferentially on TRPV1-expressing sensory neurones that is critical for sensory nerve activation and pain behaviours caused by a wide variety of reactive irritants and mediators of inflammation and tissue damage (Bandell et al. 2004; Jordt et al. 2004; Bautista et al. 2006; Kwan et al. 2006; Macpherson et al. 2007b; McNamara et al. 2007; Trevisani et al. 2007; Taylor-Clark et al. 2008). TRPA1 can be activated through covalent modification of cysteine residues within its cytosolic N-terminus by reactive electrophilic molecules (Hinman et al. 2006; Macpherson et al. 2007a; Maher et al. 2008). This mechanism could play a significant role in pathological neural hypersensitivity, as oxidative stress, which can be caused by inflammation, drug metabolism and exogenous irritants including cigarette smoke, can generate reactive electrophilic molecules including acrolein, 8-iso prostaglandin A2 and 4HNE (Esterbauer et al. 1991; Anderson et al. 1997; Roberts & Morrow, 2002) that can activate TRPA1 (Bautista et al. 2006; Macpherson et al. 2007b; Trevisani et al. 2007; Taylor-Clark et al. 2008).
Oxidative stress occurs in a variety of pathological conditions and can lead to the production of bioactive fragments of macromolecule catabolism (Negre-Salvayre et al. 2008). In particular, the respiratory tract, whose antioxidant defense mechanisms are already challenged by its oxygen-rich environment, can be subjected to substantial oxidative stress, both in the form of inhaled irritants and pollutants such as particulate matter, ozone and cigarette smoke, and during chronic inflammatory diseases such as chronic obstructive pulmonary disease (COPD), asthma and allergic rhinitis (Bowler & Crapo, 2002; Rahman et al. 2006a,b). Reactive electrophiles have been implicated in the pathogenesis of respiratory diseases (Rahman et al. 2002; Boldogh et al. 2005) and polymorphisms compromising antioxidant defenses have been linked to the susceptibility and/or progression of respiratory disorders including asthma and COPD (Ishii et al. 1999; Gilliland et al. 2004; Imboden et al. 2007).
The respiratory tract possesses a dense plexus of sensory fibres, most of which are nociceptive, i.e. their nerve terminals are activated by potentially noxious stimuli, which initiates nocifensive reflexes such as cough, mucus secretion and bronchoconstriction (Coleridge & Coleridge, 1984; Carr & Undem, 2003; Taylor-Clark & Undem, 2006). Gene expression and electrophysiological evidence suggest that functional TRPA1 and TRPV1 channels exist in respiratory nociceptive sensory neurones and their terminals (Kollarik & Undem, 2004; Nassenstein et al. 2008). In addition, the TRPA1-selective agonist cinnamaldehyde evoked action potential discharge in bronchopulmonary C-fibres and caused nocifensive reflex-induced decreases in respiratory rate analogous to those caused by capsaicin inhalation (Braun et al. 2004; Nassenstein et al. 2008).
4-Oxononenal (4ONE) is a highly reactive electrophilic oxoalkenal (Lee & Blair, 2000; Lee et al. 2001) that has been shown to be produced during oxidative stress-induced lipid peroxidation in vitro in quantities similar to those of the hydroxyalkenal 4HNE (Jian et al. 2007). Both 4ONE and 4HNE are formed downstream of arachidonic acid and linoleic acid and share the same immediate precursor (4-hydroperoxy-2-nonenal) (Blair, 2006). Critically, however, 4ONE and 4HNE differ at the C4 position where 4ONE possesses a ketone group in place of 4HNE's hydroxyl group (Lee & Blair, 2000). This dramatically increases the electrophilic reactivity of 4ONE (Doorn & Petersen, 2002; Lin et al. 2005), thus increasing its potential to form Michael adducts with susceptible amino acid residues such as cysteine residues and thus, hypothetically, activate TRPA1 channels.
While it has been elegantly demonstrated that the endogenous reactive electrophile 4HNE can activate TRPA1 and increase cytosolic calcium in dissociated sensory neurones (Macpherson et al. 2007b; Trevisani et al. 2007), the ability of 4HNE and other related products of oxidative stress to cause TRPA1-mediated activation of C-fibre terminals in situ is unknown. Here, we address the hypothesis that the reactive oxoalkenal 4ONE is a potent TRPA1 channel agonist that is capable of activating vagal nociceptive sensory nerves innervating the respiratory tract. We confirm the TRPA1 agonist activity of the hydroxyalkenal 4HNE and demonstrate that another hydroxyalkenal 4-hydroxyhexenal (4HHE), formed downstream of docoshexaenoic acid, eicosapentaenoic acid and linolenic acid (Blair, 2006), also activates heterologously expressed TRPA1, albeit with lower potency. 4ONE proved to be 10-fold more potent than 4HNE at activating TRPA1. Moreover, 4ONE evokes axon reflex-dependent contraction of isolated guinea pig bronchial smooth muscle and initiates action potential discharge from bronchopulmonary C-fibres with considerably more potency than either of the hydroxyalkenals tested. Furthermore, we demonstrate that at high concentrations (∼100 μm), 4ONE also activates heterologously expressed TRPV1 and causes TRPV1-like responses in cultured sensory neurones from TRPA1-deficient mice. However, TRPV1 channels appear not to be critical for 4ONE-induced action potential discharge from bronchopulmonary C-fibres at these concentrations, as 4ONE-induced responses (even at 100 μm) were absent in TRPA1-deficient mice.
Methods
All experiments were approved by the Johns Hopkins Animal Care and Use Committee or conducted according to the requirements of the United Kingdom Animals (Scientific Procedures) Act 1986 and strictly conformed to the ethical standards of GlaxoSmithKline Pharmaceuticals as appropriate.
HEK293 cell culture
In addition to wild-type HEK293 cells, cells stably expressing human TRPA1 (hTRPA1-HEK (Hill & Schaefer, 2007)) or human TRPV1 (hTRPV1-HEK (Hayes et al. 2000)) were used in the current study. Cells were maintained in an incubator (37°C, 5% CO2) in DMEM (containing 110 mg l−1 pyruvate) supplemented with 10% FBS and 500 mg ml−1 Geneticin as a selection agent. Cells were removed from their culture flasks by treatment with Accutase (Sigma), then plated onto poly-D-lysine-coated cover slips and incubated at 37°C for > 1 h before experimentation.
TRPA1 knockout mice
Genomic fragments homologous to the mouse TRPA1 locus were cloned by proofreading PCR (Phusion HF, Finnzymes) from C57BL/6J genomic DNA and positioned in the targeting vector such that exon 21 was flanked by LoxP sites and the LoxP flanked Neo selectable cassette was inserted into intron 2/3 (see Fig. 3A). The linearized targeting vector was electroporated into C57BL/6J ES cells (CR-UK B6 2.1) and correctly targeted ES cell clones were identified by Southern blot analysis (data not shown). Targeted clones were microinjected into BALB/c-derived blastocysts and gave rise to germline transmitting chimaeric mice. Male chimaeras were crossed with C57BL/6J females to give heterozygous offspring; these were crossed with a transgenic line expressing Cre recombinase which resulted in deletion of the Neo cassette along with exon 21 of the target gene. Mice were subsequently interbred to generate homozygous mutants.
Genotype analysis of mice was performed by PCR of tail DNA using primers ANK:10088U19 (5′-AAC ATC CCA GCG TAT ATG C) and ANK:10322L15 (5′-CAG AGC CCT CTA CCC) which produce a 300 bp band from the wild-type allele, and primers AnknullF (5′-TTT ATC TAG GGA CTT GGC) and AnknullR (5′-CCC AAC CTA TTT GTC AAC T) which produce a 718 bp band from the null allele. The offspring produced the expected Mendelian ratio of homozygotes to heterozygotes to wild-types.
Trigeminal ganglia RT-PCR
Reverse transcription and PCR amplification were performed according to the manufacturer's recommendations (Qiagen GmbH, Valencia, CA, USA). In short, total RNA was extracted from trigeminal ganglia homogenates of wild-type and TRPA1 KO mice lung homogenates (RNeasy Plus Mini Kit, Qiagen) and eluted in a final volume of 35 μl. Reverse transcription was performed with 12.75 μl of total RNA from each sample and random hexamer primers (Roche Applied Biosciences, Indianapolis, IN, USA) in a final volume of 20 μl (Omniscript RT-Kit, Qiagen). As control, an equal volume of RNA of each sample was treated according to the same protocol with addition of water instead of the reverse transcriptase. Reaction products were then used for PCR amplification of mouse β-actin, TRPV1 and TRPA1 receptors by the HotStar Taq Polymerase Kit (Qiagen) according to the manufacturer's recommendations in a final volume of 20 μl. After an initial activation step at 95°C for 15 min, cDNAs were amplified with custom-synthesized intron-spanning primers (Invitrogen, Carlsbad, CA, USA) (β-actin (Accession number NM_007393): forward primer GTGGGAATGGGTCAGAAGG, reverse primer GAGGCATACAGGGACAGCA (product size: 302); TRPV1 (Accession number NM_001001445): forward primer GCTGCTAACGGGGACTTCTT, reverse primer CTTCAGTGTGGGGTGGAGTT (product size: 285); TRPA1(1) (Accession number NM_177781): forward primer ATATGCAGTGGCAATGTGGA, reverse primer CTGAGGCCAAAAGCCAGTAG (product size: 175); and TRPA1(2): forward primer CAATGCTCTGGAATGGGTTA, reverse primer CCAAAGGTCAGGACTGGGTA (product size: 405)) by 40 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s and extension at 72°C for 1 min followed by a final extension at 72°C for 10 min. Products were then visualized in ethidum bromide-stained 1.5% agarose gels.
Dissociation of mouse vagal neurones
The methods were modified from those previously described (Taylor-Clark et al. 2008). Briefly, male C57BL/6J mice (20–40 g) were killed by CO2 overdose and the vagal ganglia rapidly dissected and cleared of adhering connective tissue. The ganglia were carefully cut open with a fine scalpel and incubated in 2 mg ml−1 collagenase type 1A and 2 mg ml−1 dispase II in 2 ml Ca2+-free, Mg2+-free Hank's buffered salt solution (HBSS) (18 h, 4°C; then 10 min, 37°C). Neurones were dissociated by trituration, washed by centrifugation, resuspended in L-15 medium containing 10% fetal bovine serum (FBS) and then transferred onto circular 25 mm glass coverslips (Bellco Glass Inc., Vineland, NJ, USA) coated with poly d-lysine (0.1 mg ml−1) and laminin (5 μg ml−1, 25 μl per coverslip). Coverslips were used within 24 h.
Calcium imaging
HEK293-covered coverslips were loaded with fura-2 acetyoxymethyl ester (fura-2 AM; 8 μm) (Molecular Probes, Carlsbad, CA, USA) in DMEM (containing 110 mg l−1 pyruvate) supplemented with 10% FBS and incubated (40 min, 37°C, 5% CO2). Neuron-covered coverslips were loaded with fura-2 AM (8 μm) in L-15 media containing 20% FBS and incubated (40 min, 37°C). For imaging, the coverslip was placed in a custom-built chamber (bath volume of 600 μl) and superfused at 4 ml min−1 with Locke solution (34°C) for 15 min before and throughout each experiment by an infusion pump. Changes in intracellular free calcium concentration (intracellular [Ca2+]free) were measured by digital microscopy (Universal; Carl Zeiss, Inc., Thornwood, NY, USA) equipped with in-house equipment for ratiometric recording of single cells. The field of cells was monitored by sequential dual excitation, 352 and 380 nm, and the analysis of the image ratios used methods previously described to calculate changes in intracellular [Ca2+]free (MacGlashan, 1989). The ratio images were acquired every 6 s. Superfused buffer was stopped 30 s before each drug application, when 300 μl buffer was removed from the bath and replaced by 300 μl of 2 × test agent solution added between image acquisitions. Following treatments, neurones were exposed to KCl (30 s, 75 mm) to confirm voltage sensitivity. At the end of experiments, both neurones and HEK cells were exposed to ionomycin (30 s, 1 μm) to obtain a maximal response.
For the analysis of fura-2 AM-loaded cells, the measurement software converted ratiometric information to intracellular [Ca2+]free using Tsien parameters ([Ca] = Kd ((R – Rmin)/(Rmax – R)) (β)) (Grynkiewicz et al. 1985) particular to this instrumentation and the HEK cells and dissociated mouse vagal neurones. Preliminary calibration studies yielded an Rmin (352/380 ratio under calcium-free conditions) of 0.3 for both HEK cells and mouse vagal neurones and an Rmax (352/380 ratio under calcium-saturating conditions) of 18 and 14 for HEK cells and neurones, respectively. β (380 in calcium-free conditions/380 in calcium-saturating conditions) was estimated as being 10 and the Kd was estimated as being 224 nm. In the following experimental studies we did not specifically calibrate the relationship between ratiometric data and absolute calcium concentration for each specific cell, choosing instead to use the parameters provided from the calibration studies and relate all measurements to the peak ionomycin response in each viable cell. This effectively provided the needed cell-to-cell calibration for enumerating individual cellular responses. Only cells that had a robust response to ionomycin were included in analyses. At each time point for each cell, data were presented as the percentage change in intracellular [Ca2+]free, normalized to ionomycin: responsex = 100 × ([Ca2+]x – [Ca2+]bl)/([Ca2+]max – [Ca2+]bl), where [Ca2+]x was the apparent [Ca2+]free of the cell at a given time point, [Ca2+]bl was the cell's mean baseline apparent [Ca2+]free measured over 120 s, and [Ca2+]max was the cell's peak apparent [Ca2+]free during ionomycin treatment. For the neuronal experiments, neurones were defined as ‘responders’ to a given compound if the mean response was greater than the mean baseline plus 2 × the standard deviation. Only neurones that responded to KCl were included in analyses. Given that vagal and trigeminal ganglia are probably comprised of heterogeneous neuronal populations it is important to emphasize the point that results are presented in two distinct ways. Firstly, the number of neurones responding (based on the criteria described above) to a given stimulus compared to the total number of neurones is reported. Secondly, the mean percentage change in intracellular [Ca2+]free normalized to ionomycin of those neurones that (based on the above criteria) were defined as ‘responders’ is reported.
Tachykinergic contraction of bronchi
Guinea pigs were killed by asphyxiation with CO2 followed by exsanguination. The main stem bronchi and distal trachea were removed, trimmed of excess tissue, placed in tissue baths and tied with silk surgical suture to force-displacement transducers (FT03C, Grass Instrument Co., Quincy, MA, USA) for recording of isometric tension on a Grass polygraph, as previously described (Undem et al. 1990; Ellis & Undem, 1994). Resting tension was set at 1 g. Tissue baths contained 10 ml of Kreb's bicarbonate solution with indomethacin (3 mm) maintained at 37°C and gassed with 95% O2–5% CO2 and replaced every 15 min during a 60 min equilibration period.
After the equilibration period, cumulative concentration–response curves were obtained with 4ONE, 4HNE, 4HHE and capsaicin, followed by carbachol (1 mm). In some studies, concentration–response curves were obtained in the presence of a combination of SR 140333 (NK1 antagonist, 1 μm) and SR 48968 (NK2 antagonist, 1 μm) – concentrations 100–1000 times greater than the affinity constants for their respected receptors that have previously been shown to block sensory nerve-dependent tachykinin-mediated bronchial contraction (Carr et al. 2000). In the desensitization studies, a high dose of capsaicin (10 μm) was first applied for 30 min and then washed every 15 min until the tissue contraction returned to baseline levels prior to the cumulative concentration–response curves. In the glutathione conjugation studies, cumulative concentration–response curves were obtained with 4ONE and capsaicin following their incubation with equimolar glutathione ethyl ester. For the bronchial contraction studies, contractions were expressed as a percentage of the maximum tissue response to 1 mm carbachol.
C-fibre extracellular recordings
Mice were killed by CO2 asphyxiation followed by exsanguination. The innervated isolated trachea–bronchus preparation was prepared as previously described (Nassenstein et al. 2008). Briefly, the airways and lungs with their intact extrinsic innervation (vagus nerve including vagal ganglia) were taken and placed in a dissecting dish containing Krebs bicarbonate buffer solution composed of (mm): 118 NaCl, 5.4 KCl, 1.0 NaH2PO4, 1.2 MgSO4, 1.9 CaCl2, 25.0 NaHCO3 and 11.1 dextrose, and equilibrated with 95% O2 and 5% CO2 (pH 7.2–7.4) (also containing indomethacin (3 μm)). Connective tissue was trimmed away leaving the trachea and lungs with their intact nerves. The airways were then pinned to the larger compartment of a custom-built two-compartment recording chamber which was lined with silicone elastomer (Sylgard). A vagal ganglion was gently pulled into the adjacent compartment of the chamber through a small hole and pinned. Both compartments were separately superfused with buffer (37°C). A sharp glass electrode was pulled by a Flaming Brown micropipette puller (P-87; Sutter Instruments, Novato, CA, USA) and filled with 3 m NaCl solution. The electrode was gently inserted into the vagal ganglion so as to be placed near the cell bodies. The recorded action potentials were amplified (Microelectrode AC amplifier 1800; A-M Systems, Everett, WA, USA), filtered (0.3 kHz of low cut-off and 1 kHz of high cut-off), and monitored on an oscilloscope (TDS340; Tektronix, Beaverton, OR, USA) and a chart recorder (TA240; Gould, Valley View, OH, USA). The scaled output from the amplifier was captured and analysed by a Macintosh computer using NerveOfIt software (Phocis, Baltimore, MD, USA). To measure conduction velocity, an electrical stimulation (S44; Grass Instruments) was applied to the centre of the receptive field. The conduction velocity was calculated by dividing the distance along the nerve pathway by the time delay between the shock artifact and the action potential evoked by electrical stimulation. Only fibres with conduction velocities < 0.7 m s−1 (C-fibres) were studied. Drugs were intratracheally applied as a 1 ml bolus over 10 s.
In the extracellular recording studies, the action potential discharge was quantified off-line and recorded in 1 s bins. A response was considered positive if the number of action potentials in any 1 s bin was > 2 the average background response. The background activity was usually either absent or less than 2 Hz. The peak frequency evoked by a stimulus was quantified as the maximum number of action potentials that occurred within any 1 s bin.
Chemicals
Stock solutions (200× +) of all agonists were dissolved in 100% ethanol (final concentration of 0.5% ethanol or less). Alkenals were purchased from Cayman Chemicals (Ann Arbor, MI, USA). Fura-2 AM was purchased from Molecular Probes. All other chemicals were purchased from Sigma-Aldrich.
Statistics
All data are mean ± standard error of the mean (s.e.m.). Unpaired t tests were used for statistical analysis when appropriate. A P value of less than 0.05 was taken as significant.
Results
We have previously shown that HEK293 cells stably transfected with human TRPA1 (hTRPA1-HEK), responded in calcium imaging assays to the selective TRPA1 channel agonist allyl isothiocyanate (AITC), and other reactive electrophilic molecules including the product of oxidative tissue damage 8-isoprostane A2 (Taylor-Clark et al. 2008). Using the same hTRPA1-HEK cells, we investigated the activation of TRPA1 channels by the oxoalkenal 4ONE and the hydroxyalkenals 4HNE and 4HHE (chemical structures shown in Fig. 1A). 4ONE (0.1–30 μm) caused an increase in cytosolic calcium (55 ± 1.9% of ionomycin, n = 274) that was absent in non-transfected HEK cells (nt-HEK cells, 2.5 ± 0.4% of ionomycin, n = 341) (Fig. 1B). 4HNE (1–300 μm) and 4HHE (1–300 μm) also activated hTRPA1-HEK cells in a dose-dependent manner, but as predicted, their responses were less robust than those of 4ONE: the −log EC50 for 4ONE, 4HNE and 4HHE was approximately 5.8, 5 and ≥ 4.3, respectively (EC50 for 4HHE is calculated from the maximum response at 300 μm) (Fig. 1C). Neither 4HNE nor 4HHE activated nt-HEK cells (data not shown). In addition, the vehicle used for all three alkenals, 0.5% ethanol, had no effect on hTRPA1-HEK cells or nt-HEK cells (2.2 ± 0.4% of ionomycin (n = 148) and 3.5 ± 0.7% of ionomycin (n = 168), respectively).
To determine if this potency ratio of TRPA1 activation (4ONE > 4HNE > 4HHE) was retained in a more complex mammalian system, we investigated the effect of these alkenal products of lipid peroxidation on sensory nerve terminal-evoked, neuropeptide-dependent axon reflex contractions of the guinea pig bronchus (Undem et al. 1990; Ellis & Undem, 1994). 4HNE (0.1–100 μm, n = 8) was relatively ineffective in this assay, causing a contraction at 100 μm of only 13 ± 5.8%. By contrast, 4ONE (0.1–100 μm, n = 8) effectively contracted the isolated bronchus, with a response at 100 μm that was nearly 50% of the tissue maximum (47 ± 3.2%). 4HHE failed to contract the bronchus even at 100 μm (n = 3) (Fig. 2A).
The mechanism of 4ONE-induced contraction was further investigated. Previous work has shown that sensory neuroactive compounds (e.g. capsaicin) induce bronchial contractility in a manner that is dependent on the activation of NK1 and NK2 receptors on the airway smooth muscle following tachykinin release from the nociceptor nerve terminals (Ellis & Undem, 1994). Pretreatment of the bronchi with a combination of SR 140333 (NK1 antagonist, 1 μm) and SR 48968 (NK2 antagonist, 1 μm) also abolished the 4ONE response (n = 3, Fig. 2B). Previous work has also shown that pretreatment of the bronchus with a high dose of capsaicin depletes the sensory terminals of tachykinins and renders the sensory terminals unresponsive to further chemical or electrical stimulation (Undem et al. 1990). Here we found that capsaicin pretreatment (10 μm) abolished the response to subsequent exposure of 4ONE (n = 4, Fig. 2B). Based on the hypothesis that TRPA1 channel activation by 4ONE depends on the reactivity of the latter's two electrophilic carbonyl groups, we investigated the effect of conjugation of these reactive groups with the glutathione membrane-permeable analogue glutathione ethyl ester. Incubation of equimolar 4ONE with glutathione ethyl ester produced a solution that was unable to induce bronchial contraction in 3 out of 3 experiments. Non-specific effects of the glutathione analogue were assessed by repeating the experiment with capsaicin instead of 4ONE (Fig. 2C). Capsaicin (10 nm to 10 μm) produced a dose-dependent contraction of the bronchi that was unaffected by the presence of equimolar glutathione ethyl ester (Fig. 2C).
The studies of 4ONE-induced tachykinergic-dependent contractions of the isolated guinea pig bronchus do not prove a TRPA1-dependent activation of nociceptive sensory nerves. Therefore we generated mice with targeted deletion of the TRPA1 channel (TRPA1−/−, see Methods and Fig. 3A) and investigated the direct effect of 4ONE on dissociated mouse vagal neurones. As part of the validation of these TRPA1−/− mice we compared their expression of the genes for TRPA1, TRPV1 and β-actin with wild-type mice of the same background strain (C57BL/6J). RT-PCR of trigeminal ganglia revealed TRPA1 channel mRNA expression (determined by two independent pairs of primers – TRPA1(1) and TRPA1(2)) was lacking in the TRPA1−/− mice, but TRPV1 and β-actin mRNA was still present (Fig. 3B). As expected TRPA1, TRPV1 and β-actin mRNA was present in the trigeminal ganglia of wild-type mice.
We also tested the validity of our TRPA1−/− mice using a functional assay of neuronal activation by known TRPA1 agonists. In calcium imaging assays of dissociated neurones from vagal sensory ganglia, 37 out of 107 wild-type neurones responded to AITC (100 μm). By contrast, AITC was ineffective in increasing cytosolic calcium in neurones from TRPA1−/− mice (Fig. 3C). In addition, we have previously shown that the PGD2 metabolite 15-deoxy-Δ12,14-prostaglandin J2 (100 μm) stimulates hTRPA1-HEK cells and wild-type mouse trigeminal neurones (81 out of 162 neurones responded, 27 ± 2% of ionomycin) via TRPA1 channel activation (Taylor-Clark et al. 2008). Trigeminal neurones dissociated from our TRPA1−/− mice failed to respond to 15-deoxy-Δ12,14-prostaglandin J2 (100 μm) (15 out of 77 neurones responded, 4 ± 0.6% of ionomycin, data not shown).
4ONE (0.1–100 μm) caused a dose-dependent increase in cytosolic calcium in 182 out of 276 wild-type vagal sensory neurones (27 ± 1% of ionomycin, Fig. 4). A maximal effect of 4ONE was produced by 10 μm. Vagal neurones isolated from TRPA1−/– mice failed to respond to concentrations up to 10 μm 4ONE (Fig. 4). Interestingly, at 100 μm 4ONE stimulated a calcium response in 63 out of 144 neurones lacking TRPA1 (29 ± 3% of ionomycin). A 20 min pretreatment of wild-type neurones with the non-selective TRP channel blocker ruthenium red (30 μm) led to a dramatic reduction in 4ONE-induced responses at all doses (maximal response to 4ONE (100 μm) of 27 ± 1% of ionomycin versus 3 ± 1% of ionomycin; wild-type versus ruthenium red-treated, respectively) (Fig. 4). Therefore, in vagal dissociated neurones it appeared that 4ONE caused an increase in cytosolic calcium via at least two ruthenium red-sensitive mechanisms, a relatively potent mechanism that was dependent on the expression of TRPA1 channels, and another non-TRPA1-dependent mechanism seen only at concentrations ≥ 100 μm.
Given that ruthenium red is an inhibitor of TRPV channels as well as TRPA1 channels, we hypothesized that a TRPV channel may be mediating the high dose effects of 4ONE. According to a recent report, the reactive irritant allicin activates TRPV1 channels (Salazar et al. 2008); we therefore hypothesized that a likely candidate was the TRPV1 channel. We investigated the effect of 4ONE on HEK293 cells stably transfected with human TRPV1 (hTRPV1-HEK). 4ONE (1 and 10 μm) had no effect on hTRPV1-HEK, whereas 100 μm 4ONE caused an increase in cytosolic calcium in these cells (29 ± 1% of ionomycin, n = 395, Fig. 5). As expected capsaicin (300 nm) also activated the hTRPV1-HEK cells (64 ± 1% of ionomycin, n = 395) and had no effect on nt-HEK cells.
We next investigated the responses of dissociated vagal sensory neurones harvested from TRPA1−/− mice to 4ONE (0.1–100 μm) in the presence of the selective TRPV1 antagonist iodo-resiniferatoxin (I-RTX, 1 μm, 10 min pretreatment). Reminiscent of the neuronal responses to 4ONE following pretreatment with ruthenium red, genetic deletion of TRPA1 and pharmacological blockade of TRPV1 dramatically reduced the neuronal response to 4ONE at all doses (Fig. 6): 66 out of 151 neurones responded with a maximal response to 4ONE (100 μm) of 8 ± 2% of ionomycin (compared with 182 out of 276 responding wild-type neurones, maximal response to 4ONE (100 μm) of 27 ± 1% of ionomycin). I-RTX (1 μm) also abolished neuronal responses to capsaicin (1 μm), verifying its effective inhibition of TRPV1. In order to prove that TRPV1 did not have modulatory effects on the responses to the lower concentrations of 4ONE, we compared the responses of wild-type neurones to neurones harvested from TRPV1−/− mice (Davis et al. 2000). As expected, TRPV1−/− neurones failed to respond to capsaicin. Responses to 4ONE (0.1–10 μm) were not reduced in the TRPV1−/− neurones as compared to wild-type, but the response to 100 μm 4ONE was inhibited (Fig. 6). This lack of a robust response to 4ONE at 100 μm may indicate a tachyphylaxis of TRPA1-mediated responses, as has been previously shown for other TRPA1 agonists in similar neuronal assays (Taylor-Clark et al. 2008).
Based on these findings, 4ONE increases cytosolic calcium in dissociated vagal sensory neurones at low micromolar concentrations via TRPA1 channels and at higher concentrations via TRPV1 channels. Single cell analysis of the wild-type neurones that responded (as defined by the criteria described in Methods) to 0.1–1 μm 4ONE (TRPA1-specific) and capsaicin (TRPV1-specific) indicates that out of 276 neurones 72 responded to both 4ONE and capsaicin, 26 responded only to 4ONE, 87 responded only to capsaicin and 91 responded to neither compounds (Fig. 7). The small percentage of TRPA1-sensitive TRPV1-insensitive mouse vagal neurones is similar to an identical nerve phenotype previously described in the mouse trigeminal ganglia (Taylor-Clark et al. 2008). Consistent with the pharmacological and knockout evidence that 4ONE activates both TRPA1 and TRPV1 (presented above); 4ONE (at any dose) only activated TRPA1−/− neurones that were also capsaicin sensitive.
To investigate whether 4ONE evoked action potential discharge from C-fibre nerve terminals through a TRPA1-dependent mechanism, we employed an ex vivo vagally innervated mouse lung preparation, where evoked action potential discharge from the nerve terminals were recorded using an extracellular electrode micropositioned in the vagal sensory ganglia. Using this preparation, we have previously shown that 300 μm of cinnamaldehyde, the selective TRPA1 agonist, activated capsaicin-sensitive bronchopulmonary C-fibres (conduction velocity < 0.7 ms−1) in a ruthenium red-sensitive manner (Nassenstein et al. 2008). We addressed the hypothesis that 4ONE would also activate bronchopulmonary C-fibres that conducted action potentials at less than 0.7 ms−1. Six out of six C-fibres responded robustly to 4ONE (10 μm), with a peak frequency of 14.7 ± 4.5 Hz (Fig. 8). All fibres conducting at less than 0.7 ms−1 responded to capsaicin (300 nm) at the end of each experiment (peak frequency of 13.3 ± 4.1 Hz). In two fibres tested 1 μm 4ONE failed to induce action potential discharge. We also assessed the ability of 4HNE to activate bronchopulmonary C-fibre terminals. Even at 100 μm, 4HNE failed to evoke action potential discharge appreciably (peak frequency of 0.5 ± 0.3 Hz, Fig. 8).
Given that 10 μm 4ONE activates bronchopulmonary C-fibre terminals, we hypothesized that TRPA1 channels are critical to this activation. After pretreating the preparation with ruthenium red (30 μm) for 15 min, 4ONE (10 μm) only activated 2 out of 5 fibres with a significantly reduced action potential discharge (2.8 ± 2.1 Hz, P < 0.05) (Fig. 8). While this observation suggested the involvement of TRPA1, we confirmed this by testing the responses of C-fibre terminals in lungs from TRPA1−/− mice. None out of 3 fibres responded to 10 μm 4ONE (peak frequency of 0.7 ± 0.4 Hz, P < 0.05, Fig. 8). Interestingly, 0 out of 4 fibres also responded to 100 μm 4ONE (peak frequency of 0.8 ± 0.3 Hz, P < 0.05, Fig. 8), the concentration shown to activate TRPV1 in the cell soma. All four TRPA1−/− C-fibres responded to capsaicin applied at the end of the experiment (peak frequency of 12.5 ± 4.4 Hz) (data not shown).
Discussion
TRPA1 channels are non-selective cation channels found on nociceptive sensory neurones (Bandell et al. 2004; Jordt et al. 2004; Bautista et al. 2006; Kwan et al. 2006), which can be activated following covalent modification by reactive electrophiles (Hinman et al. 2006; Macpherson et al. 2007a; Maher et al. 2008). Recent studies have described an electrophilic product of oxidative stress, 4-hydroxynonenal (4HNE), as a potential endogenous TRPA1 channel agonist (Macpherson et al. 2007b; Trevisani et al. 2007). Data presented here support the hypothesis that 4-oxononenal (4ONE), another electrophilic product of oxidative stress, is a more potent endogenously produced TRPA1 channel agonist, and at equivalent concentrations is more likely than 4HNE to be involved in the activation of afferent C-fibres at sites of oxidative stress.
Structurally, 4ONE differs from 4HNE in that it has a ketone group at the C4 position instead of 4HNE's hydroxy group. This single change increases the electrophilic reactivity of 4ONE dramatically, perhaps even by one hundredfold, as has been shown by its adduction to glutathione (Doorn & Petersen, 2002). As TRPA1 channels have been shown to be activated by a range of electrophilic compounds (Bandell et al. 2004; Jordt et al. 2004; Bautista et al. 2006; Kwan et al. 2006; Macpherson et al. 2007b; McNamara et al. 2007; Trevisani et al. 2007; Taylor-Clark et al. 2008), we predicted that 4ONE potentially represents the most potent endogenous activator of TRPA1 channels. Calcium imaging studies of HEK293 cells stably transfected with human TRPA1 (hTRPA1-HEK) showed that 4ONE was ∼10 times more potent in activating TRPA1 channels compared with 4HNE and ∼40 times more potent than 4-hydroxyhexenal (4HHE), another endogenously produced electrophilic alkenal. These data are similar to those recently published for these ligands in Chinese hamster ovary cells expressing mouse TRPA1 channels (Andersson et al. 2008). This order of potency was also seen in the tachykinergic-dependent contraction of the isolated guinea pig bronchus, which is a model of airway C-fibre nerve terminal activation (Undem et al. 1990). Again, 4ONE evoked far greater nerve activation than 4HNE, which only caused minor bronchial contraction even at 100 μm (4HHE had no effect at 100 μm). 4ONE's contraction of the bronchus was inhibited by capsaicin pretreatment and by a combination of tachykinin receptor antagonists, confirming that nociceptive nerve activity was required for 4ONE's actions, as have been shown previously for capsaicin-induced bronchial contraction (Undem et al. 1990; Ellis & Undem, 1994). The 4ONE-induced bronchial contraction, unlike that of capsaicin, was abolished by its prior conjugation with the nucleophilic scavenger glutathione, suggesting that 4ONE's activation of bronchial nerve terminals is dependent on its electrophilic nature.
We have previously shown that cinnamaldehyde, a selective TRPA1 channel agonist, is capable of directly activating a large proportion of capsaicin-sensitive dissociated mouse vagal sensory neurones, as assayed by calcium imaging, and that TRPA1 channel mRNA expression is co-localized in almost all TRPV1 channel mRNA expressing bronchopulmonary neurones (Nassenstein et al. 2008). 4ONE was able to directly activate wild-type capsaicin-sensitive neurones even at low micromolar neurones. Interestingly, although TRPA1−/− dissociated vagal neurones failed to significantly respond to 4ONE up to 10 μm (which is similar to a recent report of the actions of 3 μm 4ONE on neurones dissociated from TRPA1−/− mouse dorsal root ganglia (Andersson et al. 2008)), 4ONE did cause activation at a concentration of 100 μm. This response was abolished in wild-type dissociated neurones by the non-selective TRP inhibitor ruthenium red, and was abolished in the TRPA1−/− neurones by the TRPV1 antagonist I-RTX. Moreover, 4ONE (100 μm) was shown to activate HEK293 cells stably transfected with human TRPV1. Interestingly, another reactive compound, allicin (an extract of garlic), has also been shown to activate both TRPA1 and TRPV1 channels, albeit with greater potency at TRPA1 channels (Macpherson et al. 2005). Like TRPA1, TRPV1 channels are cation channels found preferentially on nociceptive sensory nerves and they have been implicated in inflammatory pain (Caterina et al. 1997; Woolf & Costigan, 1999; Montell, 2005; Wang & Woolf, 2005). A recent report demonstrated that TRPV1 channel activation by allicin was dependent on direct modification of a key cysteine group on the TRPV1 channel N-terminus (Salazar et al. 2008). It is possible that many electrophilic compounds including those that are endogenously produced are capable of direct TRPV1 activation but only the most reactive compounds like 4ONE can do so at reasonable concentrations.
Predictions of nerve terminal action potential discharge based on calcium imaging data of dissociated neuronal cell bodies should only be made cautiously, especially with respect to relative potencies. We have typically found that a rise in intracellular calcium to be a more sensitive assay of TRP channel activation than action potential discharge in tissue C-fibres (an effect that can occur independently of calcium). Consistent with this, although 4HNE consistently evoked a large calcium response at 100 μm in hTRPA1-HEK, we found that even at this relatively large concentration, 4HNE was a poor activator of action potential discharge in the vagal C-fibre terminals (we predict that at higher concentrations 4HNE will stimulate action potential discharge, but this was not studied due to confounding ethanol vehicle effects). By contrast, 4ONE at 10 μm caused action potential discharge in the C-fibres that was similar in magnitude to that seen with capsaicin. That this effect depended on TRPA1 is supported by the observation that 4ONE failed to evoke action potential discharge in capsaicin-sensitive C-fibres in TRPA1−/− mice. Given that 100 μm 4ONE was capable of activating TRPV1, as indicated in the calcium assay, it was perhaps surprising that the capsaicin-sensitive C-fibres from TRPA1−/− mice failed to respond to 4ONE even at 100 μm. One might argue that although 100 μm is suprathreshold for the detection of calcium influx via TRPV1, it is below the threshold dose required for action potential discharge via TRPV1 at the nerve terminals. It may be relevant here to note that selective permeability of TRPV1 favours calcium ions over sodium ions (Montell, 2005), thus activation of TRPV1 by a low-potency TRPV1 agonist (e.g. 4ONE) may induce calcium influx at concentrations too low to evoke TRPV1-dependent action potential discharge from the nerve terminal. In any event, data presented here suggest that TRPA1 channels are the major and possibly only mechanism that contributes to 4ONE's activation of nerve terminals at physiologically relevant concentrations.
Oxidative stress can occur through multiple mechanisms in a variety of tissues, in particular during the inflammatory response (Bowler & Crapo, 2002; Rahman et al. 2006a,b; Negre-Salvayre et al. 2008). There is a link between inflammation and nociceptive activity, giving rise to painful and unpleasant sensations and/or inducing local and central reflexes (Woolf & Costigan, 1999; Carr & Undem, 2001; Wang & Woolf, 2005), depending on the tissue innervated by the nociceptive nerve. 4ONE and 4HNE are both electrophilic products of lipid peroxidation that can be induced by enzymatic (e.g. cyclooxygenase and lipoxygenase) or non-enzymatic (e.g. reactive oxygen species (ROS)) mechanisms (Blair, 2006). Both are formed from the same precursor, 4-hydroperoxy-2-nonenal, an intermediate that can be produced downstream of arachidonic acid and linoleic acid metabolism. Recent studies demonstrate using chemical analyses of both lipid hydroperoxide products and their glutathione adducts that oxidative stress-induced 4ONE production in vitro rivals that of 4HNE and may even predominate (Lee & Blair, 2000; Jian et al. 2007); with the predicted concentration range of these products in pathophysiological conditions being approximately 1–100 μm (Esterbauer et al. 1991; Blair, 2006; Farooqui & Horrocks, 2006). The results presented here suggest that 4ONE may be a relevant mediator in the link between oxidative stress, inflammation and visceral nociceptive nerve activation in vivo.
Footnotes
-
(Received 5 March 2008; accepted after revision 16 May 2008; first published online 22 May 2008)
-
T. E. Taylor-Clark and M. A. McAlexander contributed equally to this work.
References
Figure 1. Activation of hTRPA1-HEK cells by 4ONE and other electrophilic alkenals A, structural formulae of the oxoalkenal 4ONE and the hydroxyalkenals 4HNE and 4HHE. B, mean ± s.e.m. Ca2+ responses of hTRPA1-HEK cells to putative endogenous TRPA1 agonist 4ONE (0.1–30 μm). All drugs were applied for 60 s (bars beneath traces). Continuous line denotes responses of hTRPA1-HEK cells (n = 247), dashed line denotes responses of nt-HEK cells (n = 341). C, dose–response relationships of Ca2+ responses of hTRPA1-HEK cells for 4ONE, 4HNE and 4HHE (0.1–300 μm) (data comprise > 274 cells). Data represent the maximum response during the 60 s agonist treatment taken from mean cell response versus time curves (note that the s.e.m. is often contained within symbols).
Figure 2. Contraction of isolated guinea pig bronchus by 4ONE and other electrophilic alkenals A, dose–response relationships of contraction of isolated guinea pig bronchus in vitro for 4ONE, 4HNE and 4HHE (0.1–100 μm) (n = 8 for 4ONE and 4HNE, n = 3 for 4HHE). B, dose–contraction response relationship for 4ONE alone (n = 8), following pretreatment with NK inhibitors (NK1 antagonist SR 140333 (1 μm) and NK2 antagonist SR 48968 (1 μm)) (n = 3), and following capsaicin (Caps, 10 μm) pretreatment (n = 4). C, dose-contraction response relationship for 4ONE alone (n = 8), capsaicin alone (n = 4), a solution containing 4ONE and equimolar glutathione ethyl ester (GSHee, n = 3), and a solution containing capsaicin and equimolar glutathione ethyl ester (n = 3). All data are first normalized to the maximum contraction elicited by 1 mm carbachol, then expressed as mean ± s.e.m.
Figure 3. Targeted deletion of exon 21 of the TRPA1 ion channel A, gene targeting strategy (see Methods for details). B, RT-PCR of trigeminal ganglia confirming the absence of TRPA1 channels in a TRPA1−/− mouse (−/−), but not in a wild-type mouse (WT) using 2 pairs of independent primers (TRPA1(1) and TRPA1(2)). Samples in which the reverse transcriptase was omitted (–RT) served as negative control. Identical results were observed in a repeat experiment (data not shown). C, mean ± s.e.m. Ca2+ responses of vagal neurones dissociated from wild-type (WT, black squares) and TRPA1−/− mice (grey squares) responding to AITC (100 μm). Response to capsaicin (Caps, 1 μm) also shown. Bars denote the 60 s application of agonist. All neurones responded to KCl (75 mm) applied immediately prior to ionomycin.
Figure 4. Contribution of TRPA1 channels to 4ONE-induced calcium responses in dissociated sensory neurones Mean ± s.e.m. Ca2+ responses of vagal neurones responding to 4ONE (0.1–100 μm). Response to capsaicin (Caps, 1 μm) also shown. Data comprised of neurones from wild-type mice (WT, black squares, 182/276 neurones responding), neurones from TRPA1−/− mice (grey squares, 67/144), and neurones from wild-type mice following pretreatment with 30 μm ruthenium red (WT + RR, white squares, 17/39). Bars denote the 30 s application of agonist. All neurones responded to KCl (75 mm) applied immediately prior to ionomycin.
Figure 5. Activation of hTRPV1-HEK cells by 4ONE Mean ± s.e.m. Ca2+ responses of hTRPV1-HEK cells to TRPA1 agonist 4ONE (1, 10 and 100 μm) and to TRPV1 agonist capsaicin (Caps, 300 nm). All drugs were applied for 60 s (bars). Continuous line denotes responses of hTRPV1-HEK cells (n = 395), dashed line denotes responses of nt-HEK cells (n = 208).
Figure 6. Contribution of TRPV1 channels to 4ONE-induced calcium responses in dissociated sensory neurones Mean ± s.e.m. Ca2+ responses of vagal neurones responding to 4ONE (0.1–100 μm). Response to capsaicin (Caps, 1 μm) also shown. Data comprised of neurones from wild-type mice (WT, black squares, 182/276 neurones responding), neurones from TRPA1−/− mice following pretreatment with 1 μm iodo-resiniferatoxin (TRPA1−/− + IRTX, grey squares, 66/151), and neurones from TRPV1−/− mice (white squares, 77/166). Bars denote the 30 s application of agonist. All neurones responded to KCl (75 mm) applied immediately prior to ionomycin.
Figure 7. Sensitivity of dissociated vagal neurones to 4ONE and capsaicin Percentage of wild-type neurones that respond (based on criteria described in Methods) in fura-2 AM calcium assay to low doses of 4ONE (0.1–1 μm) and capsaicin (Caps, 1 μm).
Figure 8. TRPA1 channels are critical for the activation of bronchopulmonary C-fibre terminals by 4ONE Mean ± s.e.m. peak frequency action potential discharge of wild-type bronchopulmonary C-fibres to 4ONE (10 μm) in the presence (n = 5) or absence (n = 6) of ruthenium red (30 μm, RR) and 4HNE (100 μm, n = 4); and mean ± s.e.m. peak frequency action potential discharge of TRPA1−/– mice bronchopulmonary C-fibres to 4ONE (10 μm, n = 3 and 100 μm, n = 4). Inset, representative trace of action potential discharge from a wild-type bronchopulmonary C-fibre evoked by 4ONE (10 μm) (top) and representative traces of action potential discharge from a TRPA1−/− bronchopulmonary C-fibre by 4ONE (100 μm) and capsaicin (Caps, 300 nm) (bottom). All C-fibres responded to capsaicin (300 nm) at the end of the experiment. Drugs were perfused into the lungs via the trachea. Action potentials elicited from the nerve terminals of individual bronchopulmonary C-fibres were recorded by an extracellular electrode micropositioned in the vagal sensory ganglion.
