α₇-nAChR-mediated suppression of hyperexcitability of colonic dorsal root ganglia neurons in experimental colitis

Abstract

Controlled clinical trials of nicotine transdermal patch for treatment of ulcerative colitis have been shown to improve histological and global clinical scores of colitis. Here we report that nicotine (1 μM) suppresses in vitro hyperexcitability of colonic dorsal root ganglia (DRG) (L₁–L₂) neurons in the dextran sodium sulfate (DSS)-induced mouse model of acute colonic inflammation. Nicotine gradually reduced regenerative multiple-spike action potentials in colitis mice to a single action potential. Nicotine's effect on hyperexcitability of inflamed neurons was blocked in the presence of an α₇-nicotinic acetylcholine receptor (nAChR) antagonist, methyllicaconitine, while choline, the α₇-nAChR agonist, induced a similar effect to that of nicotine. Consistent with these findings, nicotine failed to suppress hyperexcitability in colonic DRG neurons from DSS-treated α₇ knockout mice. Furthermore, colonic DRG neurons from DSS-treated α₇ knockout mice were characterized by lower rheobase (10 ± 5 vs. 77 ± 13 pA, respectively) and current threshold (28 ± 4 vs. 103 ± 8 pA, respectively) levels than DSS-treated C57BL/J6 mice. An interesting observation of this study is that 8 of 12 colonic DRG (L₁–L₂) neurons from control α₇ knockout mice exhibited multiple-spike action potential firing while no wild-type neurons did. Overall, our findings suggest that nicotine at low 1 μM concentration suppresses in vitro hyperexcitability of inflamed colonic DRG neurons in a mouse model of acute colonic inflammation via activation of α₇-nAChRs.

hyperexcitability of the peripheral sensory pathways defines an important mechanism for inflammatory pain (44). Dorsal root ganglia (DRG) neurons of colonic origin form peripheral afferent pathways that carry sensory information from the colon to the dorsal spinal cord (6, 33). Previous studies have demonstrated an increased excitability of the colonic DRG neurons in rat and guinea pig models of experimental colitis (4, 25, 28). These studies provide evidence for decreased voltage and current thresholds for action potential firing (rheobase) and an increased action potential discharge upon current injection in the inflamed colonic DRG neurons. This has been suggested to be due to an overexpression of TTX-resistant Na⁺ channels in the inflamed colonic DRG neurons; furthermore, suppression of K⁺ currents was also suggested to impact on the hyperexcitability in the inflamed colonic DRG neurons (4).

It is an interesting paradox that smoking has a favorable effect on the course and severity of ulcerative colitis (24, 39, 40). Patients with ulcerative colitis and who smoke intermittently often experience improvement of their symptoms during smoking. Controlled clinical trials using nicotine for treatment of ulcerative colitis have shown that the administration of a transdermal nicotine patch (21 mg/day) improved significantly the histological and global clinical score of colitis (5, 27, 30, 36). Despite clear beneficial effects of nicotine in patients with ulcerative colitis, its administration is often accompanied by adverse events such as nausea, lightheadedness, headache, associated with its nonselective action on various nicotinic acetylcholine receptor (nAChR) subtypes (19, 24).

Recent research also suggests that nicotine may induce anti-inflammatory effects (2). Eliakim et al. (13) reported that chronic oral nicotine administration (12.5 μg/ml) suppressed inflammation in the colon in IL-10 knockout mice via an enhanced expression of somatostatin and intestinal trefoil factor. A number of other studies suggested that nicotine can decrease proinflammatory and upregulate anti-inflammatory cytokines in cells of nonneuronal origin and that the effect develops by mechanisms mediated via α₇-nAChRs (10, 41, 42).

Although the involvement of neuronal nAChRs in anti-inflammatory mechanisms is a newer field of investigation, a number of studies have shown that nAChR agonists, such as the nonselective agonist, nicotine; an agonist with high affinity to α₄β₂-nAChRs, epibatidine; or an α₇-nAChR selective agonist, choline, induce an antinociceptive effect in animal models of acute and/or chronic pain (9, 17, 22). Although there is considerable evidence confirming the antinociceptive effects of nicotine and certain neuronal nAChR ligands, the central and peripheral mechanisms of the phenomenon are not yet fully understood.

Neuronal nAChRs expressed in colonic sensory neurons are potential candidates that can mediate the peripheral effects of nicotine in vivo. RT-PCR and immunofluorescence with nAChR subunit-specific antibodies revealed a spectrum of neuronal nAChR subunits, including α₂–α₇, α₉–α₁₀, and β₂–β₄, in rat lumbar DRG neurons (16, 38). The presence of α₇-nAChRs has been confirmed in mice DRG neurons using histochemistry (37), whereas the detailed knowledge on expression and function of other subtypes of neuronal nAChRs remains limited for mouse DRG neurons.

Here we report that nicotine, at the range of concentrations found in the serum of colitis patients treated effectively with transdermal nicotine patch (5, 30, 36), suppresses in vitro hyperexcitability of inflamed colonic DRG (L₁–L₂) neurons in a mouse model of acute colonic inflammation. Our findings suggest that nicotine returns the mode of multispike action potential firing to a single action potential per depolarization in the inflamed neurons but does not block the firing of the single action potential that is normally evoked by such a depolarization in control neurons. We also show using α₇-nAChR selective pharmacological agents and α₇-nAChR knockout mice that the suppressive effect of nicotine on hyperexcitability of inflamed colonic neurons is mediated via α₇-nAChRs. A preliminary account of this work has appeared as an abstract (1).

MATERIALS AND METHODS

Animals.

Mice were housed in a 21°C humidity-controlled Association for Assessment and Accreditation of Laboratory Animal Care approved animal care facility with food available ad libitum. The rooms were on a 12:12-h light-dark cycle (lights on at 7:00 A.M.). Mice were 8–10 wk of age and weighed ∼25–30 g at the start of the experiment. All experiments were performed during the light cycle and were approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University. Studies were conducted on male C57BL/J6 and α₇ knockout mice (C57BL/J6 background) (Jackson Laboratory, Bar Harbor, ME). All studies were performed under the guidelines for the Care and Use of Laboratory Animals as promulgated by the United States National Institutes of Health.

Retrograde labeling of colonic DRG neurons.

The extrinsic sensory colon innervation contains sympathetic and parasympathetic components that derive from small and medium (<40 pF) L₁–L₂ and L₆-S₁ DRG neurons, respectively (34). In this study, we investigated colonic neurons from L₁–L₂ DRGs. To identify DRG neurons of colonic origin, we prelabeled colonic neurons as described by Malykhina et al. (25). Briefly, the animals were fasted for 4 h before the surgical procedure. After mice were thoroughly anesthetized by pentobarbital sodium (40 mg/kg) injected intraperitoneally, a midline laparotomy was performed to gain access to the pelvic organs. The distal colon was exposed with cotton swabs, and 1,1-dioctadecyl-3–3-3′3′-tetramethylindocarbocyanine (DiI) (1.5% in DMSO; Invitrogen, Chicago, IL) was injected in the colonic wall using a Hamilton syringe at four to five sites (volume = 1.0 μl/injection). The colon was placed back in the abdominal cavity. To avoid nonspecific labeling of surrounding tissues and organs, several precautions were taken during the injection procedure: isolation of adjacent pelvic organs with gauze during dye injections; keeping the needle in place after the injection for 20 s and soaking up of dye reflux upon needle removal; and washing off any traces of dye from the organ surface with sterile saline before placing the organ in the pelvic cavity. Incisions were sutured in layers under sterile conditions, and mice were allowed to recover from anesthesia on a warm pad. After recovery from anesthesia, the mice were provided with free access to water, 12 h later to food, and monitored for signs of pain and discomfort.

Dextran sodium sulfate treatment and disease activity index evaluation.

To induce acute colon inflammation, treatment with dextran sodium sulfate (DSS, 40 kDa; MP Biomedicals, Solon, OH) was initiated 3–5 days after the surgery (7). DSS was added to drinking water and administered for the total duration of 7–8 days ad libitum. To confirm signs of colon inflammation, we evaluated the “disease activity index” on a daily basis as described previously by Jin et al. (21) on combined assessment of weight loss, blood in stool, and its consistency. Briefly, the scores were assigned ranging from zero to four, with zero being no weight loss, normal stool consistency, and no rectal bleeding, and four assigned when the weight loss was ≥20% of initial weight with gross bleeding. Blood in the stool was assessed using Hemoccult test (Beckman Coulter, Brea, CA). DSS was used at 5% concentration in C57BL/J6 mice (15, 32, 43). α7 Knockout mice required 2.5% DSS to show comparable disease activity index (∼4) to C57BL/J6 mice treated with 5% DSS.

Isolation of DRG neurons.

The mice were anesthetized by pentobarbital sodium (40 mg/kg ip) and killed by decapitation. The protocol for neuronal isolation was similar to that described previously (20). Briefly, DRGs (L₁–L₂) were removed under dissecting microscope magnification and collected in cold (4°C) PBS without Ca²⁺ or Mg²⁺ (Invitrogen). Ganglia were incubated with 15 U/ml papain (Worthington, Lakewood, NJ) in HBSS (Invitrogen) for 18 min at 37°C. After this initial enzyme treatment, the ganglia were rinsed three times in HBSS and then incubated for 18 min with 1.5 mg/ml collagenase (Sigma-Aldrich, Atlanta, GA) in HBSS at 37°C. After being washed three times with HBSS, ganglia were gently triturated with a flame-polished Pasteur pipette. The tissue fragments were centrifuged at 1,000 rpm for 5 min, and the pellet was resuspended in neurobasal culture media (Invitrogen) with 5% FBS (Invitrogen), 1% B27 (Invitrogen), 100 U/ml penicillin/streptomycin (Invitrogen), and 2 mM glutamax (Invitrogen). Cells were plated on poly-d-lysine-coated 12-mm glass cover slips (Thermanox; Nalge Nunc, Naperville, IL) and maintained at 37°C in a 95% air-5% CO₂ incubator overnight.

α₇ mRNA expression.

DRGs (L₁–L₂) were harvested from three control and three DSS-treated C57BL/J6 mice. Total RNA was extracted from ganglia with a PureLink Micro-to-Midi Total RNA Purification System (Invitrogen). The first-strand cDNA was amplified using the SensiMix One-Step Kit (Quantace, Watford, UK) at 42°C for 30 min. Primers were designed using “Vector NTI Suite” software according to published nucleotide sequences obtained from the Internet database “Pubmed Gene” (Gen Bank accession no. NM007390 for α₇) and from Ref. 31. The α₇ primer sequences used were as follows: 5′-TGCCACATTCCACACCAACG-3′ (forward) and 5′-CTACGGCGCATGGTTACTGT-3′ (reverse). r18S was used as a reference. The r18S primers sequences were as follows: 5′-ACGGACCAGAGCGAAAGCAT-3′ (forward) and 5′- GGACATCTAAGGGCATCACAGAC-3′ (reverse). Protocols for the PCR require an initial denaturation for 30 s at 95°C, 30 s annealing at 58°C, and 30 s elongation at 72°C. The PCR was terminated after 40 cycles. PCR products were confirmed by gel electrophoresis.

Electrophysiological recordings.

DRG neurons were studied within 1–2 days after plating on the cover slips. The cells were transferred to an experimental chamber and bathed (1–2 ml/min) in a solution containing (in mM) 137 NaCl, 5.4 KCl, 2 MgCl₂, 10 d-glucose, 10 HEPES, and 2 CaCl₂ (pH adjusted to 7.4 with NaOH). Labeled neurons were identified using a specific filter for DiI (Olympus, Tokyo, Japan) under an inverted fluorescent microscope (Olympus IX50; Olympus). The patch electrodes, pulled from borosilicate glass capillaries (Sutter Instrument, Novato, CA), had a resistance of 2.5–3.5 MΩ when filled with internal solution. In voltage-clamp experiments, in a majority of cells ∼85% of electrode resistance is compensated electronically, so that the effective series resistance in the whole cell configuration is accepted when <20 MΩ. Pipette solution contained (in mM): 30 KCl, 100 potassium aspartate, 6 EGTA, 1 MgCl₂, 10 HEPES, and 4 Mg-ATP (pH adjusted to 7.2 with KOH) or 20 CsCl, 110 CsMeSO₃, 10 EGTA, 1 MgCl₂, 10 HEPES, 0.1 CaCl₂, and 2 Mg-ATP (pH adjusted to 7.2 with CsOH). The potassium-containing internal solution was used to conduct the evoked action potential experiments, whereas cesium-containing solution was used to record nAChR-mediated currents.

Generation of voltage-clamp and current-clamp protocols and acquisition of the data are carried out using pCLAMP 9.0 software (Molecular Devices, Sunnyvale, CA). To study the evoked action potential characteristics, currents of 500-ms duration were applied in 20-pA steps in the current-clamp mode. Current steps to −20 pA (200 ms) were used to calculate the input resistance. Functional expression of nAChRs was evaluated in the voltage-clamp mode. If necessary, peak current amplitude was normalized to cell capacitance (pF) to determine current density. Experiments were performed at room temperature (22–25°C) and recorded using an Axopatch 200B amplifier (Molecular Devices).

Experimental solutions were applied on tested cells using a high-speed solution exchange system, HSSE-2 (ALA Scientific Instruments, Westbury, NY), a Y-ending application system as described previously (18), or via bath perfusion in some experiments. Atropine (1 μM) was added to ACh-containing solutions to omit the muscarinic component. pH of nicotine-containing solutions was adjusted to 7.4.

Chemicals.

All chemicals for patch-clamp experiments were obtained from Sigma Aldrich (Atlanta, GA) with the exception of DiI (Invitrogen) and DSS (MP Biomedicals).

Statistical analysis.

Results are presented as means ± SE for the number of cells (n) or average. Sigmaplot software was used for detection of differences in paired or unpaired t-tests, Chi square and the Fisher exact tests as appropriate, and values of P < 0.05 were regarded as significant.

RESULTS

Current injections of 500-ms duration were used to study the activation of action potentials in prelabeled colonic sensory neurons. Similar to previous studies in rat and guinea pig models of acute colonic inflammation (25, 28), we observed that the minimal current required to elicit a single action potential (rheobase) in colonic sensory neurons from DSS-treated C57BL/J6 mice (77 ± 13 pA, n = 12) was significantly less than the rheobase from control (184 ± 26 pA; n = 12) C57BL/J6 mice (Fig. 1, A–C). Increasing the amplitude of the injected current induced a greater number of action potentials in the DSS-treated but not in control C57BL/J6 mouse neurons (Fig. 1, A and B). The minimum current amplitude that elicited multiple action potentials in inflamed neurons from DSS-treated C57BL/J6 mice was 103 ± 8 pA (n = 6) and is plotted in Fig. 1C as a current threshold. In a representative neuron from a control C57BL/J6 mouse (Fig. 1A), the rheobase was 200 pA, and a twofold (2× rheobase) increase in the amplitude of the injected current induced only a single action potential in this neuron. In contrast, in the representative inflamed neuron shown in Fig. 1B, the single action potential was registered at 80 pA current injection (rheobase), and an increase of the amplitude of the injected current from 80 to 100 pA was sufficient to induce multiple action potentials in this neuron.

Fig. 1.

Effect of dextran sodium sulfate (DSS) treatment on excitability of colonic sensory neurons in C57BL/J6 mice. A and B: representative traces show the effect of colitis on rheobase and firing pattern in control (CTL) (A) and inflamed (B) neurons. C: inflamed neurons showed significantly lower rheobase than control neurons and fired multiple action potentials with a further increase of the amplitude of injected current (current threshold). *Significant difference from control neurons (P = 0.004). The duration of current injections was 500 ms in A–C. D: resting membrane potentials were similar in control and inflamed neurons.

Comparison of the resting membrane potentials in colonic sensory neurons from control and DSS-treated C57BL/J6 mice showed that the values were similar in these two groups of neurons, being equal to −42.2 ± 2.8 mV (n = 12) and −43.5 ± 2.7 mV (n = 12), respectively (Fig. 1D).

Next, we evaluated the effect of bath-applied nicotine (1 μM) on action potential discharge in inflamed colonic neurons isolated from C57BL/J6 mice. In these experiments, the action potentials were elicited by current injections (500 ms duration) at the level of the current threshold. The action potentials were initially elicited in the absence of nicotine, followed by recordings in the presence of nicotine at intervals of 1 min. As shown in Fig. 2A, nicotine gradually reduced the number of action potentials fired per depolarization in inflamed neurons evaluated at the current threshold level (n = 7). Increasing the amplitude of injected currents (2-fold) did not result in recovery of multiple action potential firing in neurons incubated with nicotine for 3 min (n = 4). The suppression of these multiple action potentials persisted even after 25 min washout with bath solution. In the absence of nicotine, multiple action potentials persisted in inflamed neurons from C57BL/J6 mice for >5–8 min, confirming that the suppression of action potentials was not due to rundown (data not shown, n = 3). Similar experiments conducted in colonic sensory neurons from control C57BL/J6 mice showed no change in action potential discharge in the presence of nicotine. A single potential was still induced in these cells in the presence of nicotine (n = 4; Fig. 2, B and C).

Fig. 2.

Nicotine in vitro suppresses action potential firing in inflamed colonic sensory neurons from C57BL/J6 mice. A and B: representative recordings of action potentials generated by a 500-ms current injection in the absence and presence of nicotine (1 μM) taken in 1-min intervals in colonic sensory neurons from DSS-treated (A) and control (B) C57BL/J6 mice. C: time course of the suppressive effect of nicotine on action potential discharge in neurons from DSS-treated and control C57BL/J6 mice. D: membrane potential of neurons from DSS-treated but not from control mice was shifted toward more positive values upon nicotine incubation. E: current activation in a representative inflamed neuron by prolonged application of 1 μM nicotine. Holding potential was −60 mV.

Exposure of inflamed neurons from C57BL/J6 mice to nicotine did not change significantly the input resistance (470 ± 102 MΩ vs. 478 ± 89 MΩ, P = 0.95, n = 5) or the resting membrane potential (from −43.6 ± 2.7 to −33.6 ± 5.8 mV, P = 0.09, n = 5). No change in the membrane potential was observed in neurons from control C57BL/J6 mice in the presence of nicotine (−45.3 ± 6.0 and −45.5 ± 6.4 mV, respectively; n = 4) (Fig. 2D).

Whole cell voltage-clamp experiments were carried out to record ionic currents activated by 1 μM nicotine in inflamed neurons. In five out of seven neurons, a small but slowly desensitizing current with an average amplitude of −22.2 ± 10.0 pA (n = 5) was activated at the holding potential of −60 mV (cell capacitance was ∼30 pF) (Fig. 2E).

To investigate whether specific subtype(s) of neuronal nAChRs mediate the suppressive effect of nicotine on hyperexcitability of inflamed colonic neurons, we compared pharmacological properties of neuronal nAChR-mediated currents in DSS-treated and control C57BL/J6 mice. Our data showed that, in control C57BL/J6 mice, 14 out of 53 (26.4%) labeled colonic DRG neurons responded to 1 mM ACh with various types of currents. In 5 out of those 15 neurons, the currents were abolished by the α₇-antagonist methyllicaconitine (MLA; Fig. 3A, top) and in four other neurons by the α₄β₂-antagonist dihydro-β-erythroidine (DHβE) (Fig. 3A, middle), suggesting that these ACh-induced currents were mediated by α₇- and α₄β₂-nAChRs, respectively. In 5 remaining cells out of 14 that responded to ACh, the currents were not affected either by DHβE or MLA, suggesting that the currents in those neurons were not due to activation of the α₄β₂- or α₇-nAChR subtypes (Fig. 3A, bottom). Similar whole cell experiments were performed on colonic DRG neurons isolated from DSS-treated C57BL/J6 mice. In contrast to control C57BL/J6 mice, the application of 1 mM ACh induced currents in 9 out of 12 tested neurons (75.0%) from DSS-treated C57BL/J6 mice, and in 6 neurons the currents were inhibited by MLA (Fig. 3B). Out of three remaining cells that responded to ACh in inflamed mice, one cell showed a current sensitive to DHβE, and in two other cells the currents were insensitive to both MLA and DHβE. These results suggest that acute colonic inflammation is accompanied by an increase in the number of colonic DRG neurons that respond to ACh (26.4 vs. 75.0%; Fisher exact test: P = 0.003). Although there was a trend in the increase in the neurons that were sensitive to the α₇-nAChR antagonist MLA (from 35.7 to 66.6%; Fig. 3C), this did not reach statistical significance (Fisher exact test: P = 0.15).

Fig. 3.

Effect of DSS-induced colonic inflammation on α₇-nicotinic acetylcholine receptor (α₇-nAChR) functional activity in colonic neurons from C57BL/J6 mice. A: a variety of nAChR-mediated currents in control colonic dorsal root ganglia (DRG) neurons isolated from a C57BL/J6 mouse. An example of ACh (1 mM)-evoked current that was suppressed by methyllicaconitine (MLA), suggesting the α₇-nAChR activation, is shown in top. An example of ACh-evoked current that was suppressed by dihydro-β-erythroidine (DHβE) but was insensitive to MLA, suggesting α₄β₂-nAChR activation, is shown in middle. The third type of current was insensitive to both DHβE and MLA, suggesting that the current was not due either to α₄β₂- or to α₇-nAChR activation, is shown in bottom. All three types of current were equally distributed among the neurons that responded to ACh. B: MLA-sensitive currents predominated in inflamed neurons, suggesting that the currents were due to α₇-nAChR activation. Examples are shown for two individual inflamed neurons from DSS-treated C57BL/J6 mouse. C: an increase in a relative number of neurons responding to ACh and sensitive to MLA in inflamed mice. Holding potential was −80 mV.

To examine the α₇-mRNA expression in colonic sensory neurons from DSS-treated and control C57BL/J6 mice, we collected the L₁–L₂ DRGs from three mice per each group and performed RT-PCR analysis. RNA agarose gel in Fig. 4 shows single bands for α₇ and internal control, 18S, in both DRG samples from DSS-treated and control C57BL/J6 mice.

Fig. 4.

Assessment of α₇-mRNA expression in control and inflamed (INF) colonic DRG (L₁–L₂) neurons. RNA agarose gel shows single bands for α₇ and internal control, 18S.

To further test the functional role of α₇-nAChRs during acute colon inflammation in colonic sensory neurons from C57BL/J6 mice, we used α₇ pharmacological agents and α₇ knockout mice in the next set of experiments on nicotine-induced inhibition of action potential firing in inflamed colonic sensory neurons. As demonstrated in Fig. 5A, nicotine (1 μM) failed to depress the action potential firing in inflamed colonic sensory neurons from C57BL/J6 mice in the presence of the α₇-antagonist MLA (n = 3). In agreement with this observation, the α₇ selective agonist choline, at a concentration of 10 μM, which has been shown to be comparable to 1 μM nicotine for current activation in recombinant α₇-nAChRs (29), mimicked the suppressive effect of nicotine on multispike action potential firing in inflamed neurons (n = 3; Fig. 5B).

Fig. 5.

Pharmacological evidence of essential role of α₇-nAChRs in the mechanism of nicotine-induced suppression of action potential firing in inflamed colonic neurons. A: nicotine fails to suppress multiple-spike action potential firing in inflamed C57BL/J6 neurons in the presence of the α₇ competitive antagonist MLA. B: α₇ agonist choline suppresses hyperexcitability of inflamed colonic sensory neurons. The duration of current injections was 500 ms.

To confirm that the nicotine-induced suppression of hyperexcitability in inflamed colonic sensory neurons occurs as a result of activation of α₇-nAChRs, we measured the effect of nicotine on action potentials in the α₇ knockout mice. Consistent with the pharmacological block of the suppressive effect of nicotine on action potential firing in inflamed colonic sensory neurons from C57BL/J6 mice by MLA (Fig. 5A), nicotine (1 μM) failed to suppress action potential firing in colonic sensory neurons from DSS-treated α₇ knockout mice (Fig. 6A). The number of action potential spikes at the current threshold and the membrane potential were not altered in colonic sensory neurons from DSS-treated α₇ knockout mice in the presence of nicotine (n = 6; Fig. 6B). In addition, the inflamed colonic sensory neurons from DSS-treated α₇ knockout mice were characterized by significantly lower rheobase (10 ± 5 pA, n = 12; P = 0.042), a lower current threshold (28 ± 4 pA, n = 12), and a higher number of action potential spikes at the threshold (8.6 ± 2.3; n = 5) than the inflamed neurons from C57BL/J6 mice.

Fig. 6.

Nicotine-induced suppression of action potential firing failed to occur in inflamed colonic neurons from DSS-treated α₇ knockout mice. A: nicotine fails to suppress action potential firing in the representative inflamed colonic neuron in the absence of α₇-nAChR. B: the average no. of spikes evaluated at the current threshold in individual inflamed neurons from α₇ knockout mice remains unchanged upon application of nicotine. C: membrane potential is similar in neurons from control and DSS-treated α₇ knockout mice and is not altered upon incubation with nicotine in inflamed neurons. D: rheobase and threshold levels in control and DSS-treated α₇ knockout mice. The duration of current injections was 500 ms in A, B, and D.

Colonic sensory neurons from control α₇ knockout mice had resting membrane potentials (−39.8 ± 0.7 mV; n = 12; Fig. 6C) that were not significantly different from those in C57BL/J6 mouse neurons (Fig. 6D). In contrast to control C57BL/J6 mouse neurons that fired a single action potential per depolarization (n = 12; Fig. 1A), 8 out of 12 neurons from control α₇ knockout demonstrated multiple action potentials at the current threshold of 66 ± 15 pA, whereas in the other four neurons only a single action potential fired upon an increase of the amplitude of the current injection (4× rheobase). This suggests that α₇-nAChR knockout per se induces hyperexcitability that is similar to colonic DRG neurons from DSS-treated C57BL/J6 mice but is significantly different from control C57BL/J6 mice (Fisher exact test: P = 0.001).

DISCUSSION

This study performed in a mouse model of acute colonic inflammation showed that: 1) inflamed colonic sensory neurons are characterized by hyperexcitability, 2) application of nicotine reverses the hyperexcitability of inflamed neurons to resemble those of control, i.e., shifts multiple action potential activation to activation of a single action potential, and 3) the suppressive effect of nicotine on hyperexcitability in inflamed colonic neurons is mediated via α₇-nAChRs.

Previous findings demonstrated that, in rat and guinea pig, colonic inflammation alters the intrinsic firing pattern of nociceptive neurons projecting to the colon (4, 25, 28). After 1 wk of 5% DSS treatment, multiple action potential discharge was typical for inflamed colonic sensory neurons from C57BL/J6 mice. In contrast, increasing the amplitude of current injections above the rheobase level induced only a single action potential in colonic neurons from control C57BL/J6 mice. The absence of repetitive action potential firing of colonic sensory neurons under control conditions is similar to that reported for rat and guinea pig colonic sensory neurons. Comparable rheobase level (165 pA) and somewhat more negative resting membrane potential values (−59.8 mV) were reported for rat DRG (L₁–L₃) neurons of colonic origin under control conditions (25). Also, in agreement with findings in rat and guinea pig models of colitis (25, 28), colonic inflammation resulted in a significant decrease of the rheobase level but did not affect the resting membrane potential of colonic sensory neurons in C57BL/J6 mice (Fig. 1, C and D). The decreased rheobase levels in inflamed neurons from DSS-treated C57BL/J6 mice are consistent with the ability of these neurons to fire multiple action potentials upon current injection, and validate their hyperexcitability.

Aδ- and C-fiber afferents are connected to small and medium DRG neurons that respond to nociceptive stimuli and carry nociceptive information to the central nervous system using neuropeptides and neurotransmitters (35). Sensitization of colonic DRG neurons leads to decreased threshold for activation and increased excitability in response to stimulation and has been suggested to underlie increased pain response to noxious stimuli in the animal models of colonic inflammation (4, 26).

The central finding of this study is that nicotine at low 1 μM concentration suppressed in vitro regenerative action potential firing in inflamed colonic neurons, but did not prevent the generation of the initial spike. In addition, the suppression was not observed in control neurons where only single action potential activation persisted. This suggests that the mechanism of nicotine's action in inflamed colonic DRG neurons is to suppress the ionic currents involved in generation of multiple-spike action potentials. The specific channels involved in regeneration of the action potentials in inflamed colonic DRG neurons remain to be determined. In this respect, King et al. (23) recently reported that the evoked hyperexcitability in inflamed colonic neurons is associated with an increase in voltage-gated Na⁺ 1.8 currents.

The fact that the suppressive effect of nicotine on hyperexcitability of inflamed colonic neurons developed gradually, but did not occur immediately (Fig. 2C), and the absence of the effect of nicotine in inflamed neurons from α₇ knockout mice suggest that the effect of nicotine is not due to direct interaction of nicotine with ion channels that mediate hyperexcitability in the inflamed neurons.

The suggestion that interaction of nicotine with α₇-nAChRs in inflamed colonic neurons plays a key role in the mechanism that underlies the suppression of hyperexcitability in those neurons is supported by the following findings. Our whole cell experiments revealed that colonic inflammation was accompanied by an increase in a number of colonic sensory neurons from C57BL/J6 mice that responded to 1 mM ACh and that the predominant fraction of those ACh-induced currents appeared to be sensitive to the α₇ antagonist MLA (Fig. 3, B and C). We found that only 26.4% of colonic DRG (L₁–L₂) neurons responded to 1 mM ACh in C57BL/J6 mice under control conditions. Similar to that, studies of lumbar DRG from adult rats found only 58% (64 neurons were sampled) or less (24%, 158 neurons were sampled) responded to 1 mM ACh (11, 14). There was an equal distribution between three types of currents in adult mice labeled colonic DRG (L₁–L₂) neurons with respect to their sensitivity to MLA and DHβE. In a similar manner, Dube et al. (11) classified ACh-induced whole cell currents into three groups in adult rat L₅–L₆ DRGs as induced by α₄β₂ (54%)- and α₇ (28%)-nAChRs, and of other origin (18%, slow kinetic). In contrast to control C57BL/J6 mice, the application of 1 mM ACh-induced currents in 75% of inflamed neurons from DSS-treated C57BL/J6 mice and 66% out of these ACh-induced currents were mediated by α₇-nAChRs. We confirmed the presence of α₇-mRNA in L₁–L₂ DRGs from both control and DSS-treated C57BL/J6 mice. Future studies utilizing laser capture will be carried out to examine changes in α₇ mRNA levels from labeled colonic L₁–L₂ DRG neurons. In agreement with the hypothesis that α₇ plays an important role in hyperexcitability of DRG neurons, we found that nicotine failed to induce its suppressive effect on repetitive action potential firing in the presence of the α₇-nAChR antagonist MLA or in mice that lack α₇-nAChRs. Furthermore, the α₇-agonist choline, at its low 10 μM concentration comparable in its potency to nicotine at 1 μM concentration for α₇-nAChRs (29), was able to mimic the effect of nicotine in inflamed neurons from C57BL/J6 mice. These data strongly suggest that the effect of nicotine on hyperexcitability of inflamed colonic neurons is mediated via α₇-nAChRs.

Interestingly, colonic sensory neurons from control α₇ knockouts possessed lower rheobase levels than the neurons from control C57BL/J6 mice. Furthermore, 66% of colonic sensory neurons from control α₇ knockouts fired multiple action potentials upon current stimulation similar to that in inflamed neurons from C57BL/J6 mice. This finding suggests that an endogenous α₇-nAChR tone may play a role in curtailing hyperexcitability.

Overall, our data show that nicotine suppresses in vitro hyperexcitability of colonic DRG (L₁–L₂) neurons via a α₇-nAChR-mediated mechanism. Furthermore, it is important that the lack of α₇-nAChRs predispose the development of hyperexcitability of colonic sensory neurons. This may underlie a novel functional role of α₇-nAChRs in addition to already identified properties such as presynaptic modulation of neurotransmitter release, high Ca²⁺ permeability, and activation of a number of intracellular signaling pathways (3, 8, 12). It appears that both the anti-inflammatory effect of nicotine in the colon (13) and nicotine-induced suppression of hyperexcitability in colonic DRG neurons are mediated via activation of neuronal α₇-nAChRs and can be potentially achieved using α₇-nAChR agonists that establishes these pharmacological agents as a promising tool for treatment of ulcerative colitis in humans.

GRANTS

This work was supported by National Institutes of Health Grants 5P50DA-05274 and DK-046367.

DISCLOSURES

No conflicts of interest are declared by the authors.