Antispasmodics are used clinically to treat a variety of gastrointestinal disorders by inhibition of smooth muscle contraction. The main pathway for smooth muscle Ca2+ entry is through L-type channels; however, there is increasing evidence that T-type Ca2+ channels also play a role in regulating contractility. Otilonium bromide, an antispasmodic, has previously been shown to inhibit L-type Ca2+ channels and colonic contractile activity. The objective of this study was to determine whether otilonium bromide also inhibits T-type Ca2+ channels. Whole cell currents were recorded by patch-clamp technique from HEK293 cells transfected with cDNAs encoding the T-type Ca2+ channels, CaV3.1 (α1G), CaV3.2 (α1H), or CaV3.3 (α1I) alpha subunits. Extracellular solution was exchanged with otilonium bromide (10−8 to 10−5 M). Otilonium bromide reversibly blocked all T-type Ca2+ channels with a significantly greater affinity for CaV3.3 than CaV3.1 or CaV3.2. Additionally, the drug slowed inactivation in CaV3.1 and CaV3.3. Inhibition of T-type Ca2+ channels may contribute to inhibition of contractility by otilonium bromide. This may represent a new mechanism of action for antispasmodics and may contribute to the observed increased clinical effectiveness of antispasmodics compared with selective L-type Ca2+ channel blockers.
- voltage dependence
- smooth muscle
antispasmodics have been used clinically for many years to treat gastrointestinal disorders thought to be associated with increased contractility (4, 8, 11, 24). The mechanisms of action of this class of drugs are diverse and not fully understood. A key mechanism of action is thought to be inhibition of smooth muscle contractility, hence their class name. Together with their antimuscarinic effects, several have been shown to inhibit L-type Ca2+ channels (1, 15, 29). Entry of Ca2+ into gastrointestinal smooth muscle through L-type Ca2+ channels is required for initiation of gastrointestinal contractions. However, selective L-type Ca2+ channel blockers such as nifedipine do not have the same clinical effectiveness as the antispasmodic class of drugs, suggesting other mechanisms of action.
One antispasmodic is otilonium bromide (OB), a quaternary ammonium salt used clinically for the treatment of irritable bowel syndrome (3), a functional disorder of the gastrointestinal tract with high prevalence in Western countries (19). The etiology of irritable bowel syndrome is incompletely understood. Major complementary hypotheses include afferent neuronal hypersensitivity, postinfective hypersensitivity, abnormal bacteria/host interactions, small intestinal bacterial overgrowth, psychosocial factors, and smooth muscle hypercontractility (9). The latter, documented in motility studies, together with the presence of crampy abdominal pain in patients, form the basis for the treatment of irritable bowel syndrome by smooth muscle relaxants. OB can inhibit spontaneous intestinal contractile activity through several mechanisms (3, 4, 14, 26) including block of L-type Ca2+ channel currents in smooth muscle cells from human jejunum and rat colon, in the latter leading to an abolishment of muscle contractions (16, 29). L-type Ca2+ channels are voltage dependent, and opening requires a depolarization of the membrane potential. The phasic depolarization required to allow L-type Ca2+ channels to open in gastrointestinal smooth muscle cells is generated by interstitial cells of Cajal (ICC) and then propagated through the ICC network to smooth muscle cells.
Increasing evidence points toward a role for Ca2+ channels in propagation of the intestinal slow wave and specifically for a role for T-type-like Ca2+ channels (6, 13, 25, 33). T-type-like Ca2+ channels are expressed in gastrointestinal smooth muscle cells and ICC (6, 12). Inhibition of T-type Ca2+ channels in the small intestine is postulated to disrupt propagation of the slow wave to smooth muscle, causing a decrease in smooth muscle contractility. The aim of this study was to determine the effect of OB on T-type Ca2+ channels CaV3.1, CaV3.2, and CaV3.3 exogenously expressed in HEK 293 cells.
Transfection of HEK cells.
Construction of the expression vectors containing the genes encoding T-type Ca2+ channels (older nomenclature used to identify alpha subunit in parentheses) CaV3.1 (α1G), CaV3.2 (α1H), and CaV3.3 (α1I) used in this study has been described previously (2, 7, 23). Each of the expression vectors along with green fluorescent protein, pEGFP-C1 (Clontech, Palo Alto, CA), were transiently cotransfected into HEK293 cells (American Type Culture Collection, Manassas, VA) using Lipofectamine 2000 Reagent (Invitrogen, Carlsbad, CA). Transfected HEK cells were cultured for 2 days prior to study.
Whole cell currents were recorded by standard patch-clamp technique from transfected cells identified by fluorescent microscopy (28, 29). Patch-clamp electrode glass Kimble KG-12 was pulled on a P-97 puller (Sutter Instrument, Novato, CA). Electrodes were coated with R6101 (Dow Corning, Midland, MI) and fire polished to a final resistance of 3–5 MΩ. Currents were amplified, digitized, and processed by use of an Axopatch 200B amplifier, a Digidata 1322A, and pCLAMP 9 software (Molecular Devices, Union City, CA). For CaV3.1 and CaV3.2, cells were stepped from a holding voltage of −100 mV to −90 through +35 mV in 5-mV intervals for 400 ms each. Start-to-start time was 1 s. Whole cell records from CaV3.1 and CaV3.2 were sampled at 10 kHz and filtered at 4 kHz with an eight-pole Bessel filter. Sampling rates of data for CaV3.1 and CaV3.2 were reduced (decimated) fivefold to 2 kHz during analysis. Ca2+ currents completely inactivated before 400 ms in HEK cells transfected with CaV3.1 or CaV3.2 but not CaV3.3. Therefore, the protocol was modified slightly for CaV3.3; cells were stepped from a holding voltage of −100 mV to −90 through +30 mV in 10-mV intervals for 2 s each. Start-to-start time was 4 s. Whole cell records from CaV3.3 were sampled at 2 kHz and filtered at 1 kHz with an eight-pole Bessel filter. The sampling rate of data for CaV3.3 was not reduced during analysis, and 70–80% series resistance compensation (lag of 60 μs) was applied during all recordings.
For use- and time-dependent block experiments, HEK cells transfected with T-type Ca2+ channels were held at −120 mV and repeatedly stepped to −30 mV (a voltage positive of peak inward current, chosen to ensure full open probability) for 15 ms (CaV3.1, CaV3.2) or 30 ms (CaV3.3) at 2 or 40 s start-to-start intervals. Total elapsed time per record was 200 s. Eight subsweeps of P/N leak subtraction were applied to each trace during recording.
Measurement of contractions from human colonic smooth muscle.
The Institutional Review Board approved the use of specimens of human colonic tissue resected for nonobstructive colon cancer. Normal sigmoid colon at least 10 cm away from the tumor was harvested directly into chilled F12 buffer solution (F-12: GIBCO, Invitrogen, 14 mmol/l NaHCO3; pH 7.35) and transported to the laboratory within 20 min. After the mucosa was removed, a strip of muscle (2 × 10 mm) was cut with the long axis of the muscle strip parallel to the circular muscle layer. One end of the muscle strip was pinned down to the Sylgard-coated floor of a recording chamber and the other end of the strip was tied with a string and attached to a force transducer. A stretch was applied to the strip, which extended its length to 1.5 to 1.6 times its original length. The recording chamber was continuously perfused with oxygenated normal Krebs solution (NKS) at 37°C and at a rate of 3 ml/min. Recording started ∼2 h after the muscle strip was placed into the recording chamber. The composition of the NKS was (in mM) 137.4 Na+, 5.9 K+, 2.5 Ca2+, 1.2 Mg2+, 124 Cl−, 15.5 HCO3−, 1.2 H2PO4−, and 11.5 glucose. It was continuously bubbled with 97% O2-3% CO2, and maintained at pH 7.4. Values are given as means ± SE and a P value of <0.05 is considered significant.
Drugs and solutions.
The pipette solution contained (in mM) 145 Cs+, 35 Cl−, 5 Na+, 5 Mg2+, 2 EGTA, 5 HEPES, 125 methanesulfonate. pH was adjusted to 7.0 with CsOH. The bath solution contained normal Ringer solution (in mM) 149.2 Na+, 4.74 K+, 159 Cl−, 2.54 Ca2+ buffered with 5 mM HEPES. pH was adjusted to 7.4 with NaOH. Osmolality of solutions was 300 mmol/kg. Chemicals and mibefradil (3 μM) were purchased from Sigma Chemical (St. Louis, MO). OB was obtained from Menarini Ricerche, Florence, Italy. For washout experiments, 6–15 ml of normal Ringer solution was applied over the course of 3–8 min. The median time for all washout experiments was 5.1 min (n = 20). After a seal and access were established, control recordings were taken in NaCl Ringer solution for 2–5 min until T-type Ca2+ current reached a steady value. Next, each cell was exposed for 30–120 s with Ringer solution containing increasing concentrations of OB (0 M OB control rinse followed by 10−8 to 10−5 M OB).
Data were analyzed via Clampfit and Excel (Microsoft, Redmond, WA). Current-voltage (I-V) curves plot peak inward current vs. voltage steps. Data were normalized by using the formula Inorm = 100(Iv)/(Imax), where Inorm is the normalized peak inward current at a particular voltage sweep, Iv is the current at each voltage sweep, and Imax is the maximum peak inward current from the set of traces (usually the peak inward current at −50 mV). Dose-responses (Fig. 1⇓–3) were fitted with a logistic three-parameter curve (Hill equation): y = a/[1+(x/x0)b], where a is the y-intercept (∼100%), b is the slope, and x0 is the IC50 value. Time to peak (Fig. 6) was measured from the start of each pulse to the time of peak inward current. Inactivation kinetics (Fig. 6) were fitted by a two-term weighted exponential function: , where τ1 and τ2 are time constants of inactivation and activation, respectively. Currents recorded in the presence of OB were vertically scaled to visually compare the kinetics of partially blocked currents to those from control (Fig. 6). Time to peak and time constant of inactivation are measurements of time, and kinetic measurements from scaled currents were confirmed to be the same as unscaled within 0.01 ms. Significance was determined via GraphPad software (La Jolla, CA) by nonparametric repeated-measures ANOVA with Friedman posttest (Figs. 1⇓–3, 5) or Student's t-test (Figs. 6–7). A P value of < 0.05 was considered significant.
Cell capacitances were similar between the cells that expressed CaV3.1 (13.0 ± 2.2 pF, n = 13), CaV3.2 (11.4 ± 1.6 pF, n = 12), and CaV3.3 (9.5 ± 1.7 pF, n = 13) (P > 0.05). The average peak current density of CaV3.2 (−22.4 ± 5.3 pA/pF, n = 12) was significantly less than either CaV3.1 (−60.3 ± 13.3 pA/pF, n = 13, P < 0.05) or CaV3.3 (−60.9 ± 14.5 pA/pF, n = 13, P < 0.05).
OB blocks T-type Ca2+ channel alpha subunits expressed in HEK293 cells with a lower concentration for CaV3.3 than CaV3.1 or CaV3.2.
CaV3.1 current inactivated by 200 ms, peaked at approximately −50 mV step voltage, and was significantly blocked by 82.6 ± 2.2% at 3 μM and 96.7 ± 0.7% at 10 μM OB (Fig. 1, n = 6, P < 0.05). Ca2+ current through CaV3.2 completely inactivated before 400 ms, peaked at about −45 mV step voltage, and was significantly blocked by 76.0 ± 3.6% at 3 μM and 87.4 ± 3.2% at 10 μM OB (Fig. 2, n = 6, P < 0.05). CaV3.3 current completely inactivated by 1 s, peaked at −50 to −40 mV step voltage, and was significantly blocked by 73.2 ± 4.2% at 1 μM, 91.6 ± 2.2% at 3 μM, and 96.3 ± 1.4% at 10 μM OB (Fig. 3, n = 7, P < 0.05).
Dose-response curves generated from the above data (Figs. 1C, 2C, and 3C) demonstrated that OB inhibited T-type Ca2+ channel subunit CaV3.3 at lower concentrations than CaV3.1 or CaV3.2 (Table 1). The corresponding IC50 value for CaV3.3 (451 ± 90 nM, slope 1.23 ± 0.07, n = 7) was significantly less than CaV3.1 (774 ± 109 nM, n = 6, P < 0.05) or CaV3.2 (1,070 ± 269 nM, n = 6, P < 0.05). Although the IC50 values for CaV3.1 and CaV3.2 were similar, their respective slope constants were significantly different (1.54 ± 0.14 vs. 1.11 ± 0.08, n = 6 each, P < 0.05).
Time course for OB block.
Cells were incubated with increasing concentrations of OB. IC50 values may be skewed if block did not reach a plateau at each concentration or if the block were use dependent. Therefore, we tested block of T-type Ca2+ currents opening in response to steps to −30 mV at two different frequencies, 0.5 and 0.025 Hz (Fig. 4). OB (3 μM) blocked T-type Ca2+ currents down to a plateau by 30–60 s at 0.5 Hz and 80 s at 0.025 Hz (Fig. 4, right), suggesting that the block was not dependent on channel openings because the plateau at 0.5 Hz was reached after more pulses than at 0.025 Hz. Time to plateau (Fig. 4) also was shorter than sequential rinse intervals (Fig. 1⇑–3), suggesting that the time period between rinses was adequate.
Block of T-type Ca2+ channels by OB is reversible.
To determine whether inhibition of T-type Ca2+ channels by OB was reversible, we recorded current after washing off the drug (Fig. 5). T-type Ca2+ currents could not be recorded from HEK 293 cells for longer than 1–2 min in the presence of 10 μM OB. Therefore, we applied 3 μM before washout. Following significant block of CaV3.1 (79.2 ± 3.0% block, n = 6, P < 0.05), CaV3.2 (67.1 ± 3.9% block, n = 6, P < 0.05), and CaV3.3 (79.1 ± 5.4% block, n = 8, P < 0.05) by OB (3 μM), peak Ca2+ currents for all three T-type channels returned to levels at or below baseline (CaV3.1: 83.2 ± 9.6% of baseline current, n = 6, P > 0.05; CaV3.2: 80.0 ± 7.2%, n = 6, P > 0.05; CaV3.3: 74.2 ± 6.1%, n = 8, P > 0.05) (Fig. 5) within 5 min.
OB slows the activation and inactivation kinetics of CaV3.1.
T-type Ca2+ currents at the six increasing concentrations of OB shown in Fig. 1 were vertically scaled to match the baseline level of peak inward current of 0 M controls and then analyzed for change in activation times (Fig. 6, middle) or inactivation time constants (Fig. 6, right). To control for any changes in kinetics induced by time or rinses, we carried out control experiments with sham rinses of control solution. In response to rinsing the bath up to six times with drug-free Ringer solution, CaV3.1 activated the same or faster and inactivated faster at all voltage steps tested (Fig. 6A). At the 3 μM dose of OB, CaV3.1 was slower to both activate and inactivate at voltage steps from −100 to −60 through 0 mV (Fig. 6B; n = 6, P < 0.05). In contrast, 3 μM OB changed activation and inactivation channel kinetics of CaV3.2 or CaV3.3 at only select voltage steps (Fig. 6, C and D; n = 6 each, P < 0.05 by Student's t-test where indicated by asterisk).
Mibefradil reduces frequency of contractions in human colon.
We determined the potential contribution of T-type Ca2+ channels to smooth muscle contractions in human colon. Both high-frequency, low-amplitude contractions and low-frequency, high-amplitude contractions were observed in muscle strips (Fig. 7). In NKS the frequency was 6.03 ± 1.33 min−1 for the high-frequency, low-amplitude contractions and 0.73 ± 0.18 min−1 for the low-frequency, high-amplitude contractions (n = 4). Twenty minutes after application of mibefradil (3 μM), the frequency of the high-frequency, low-amplitude contractions was 4.73 ± 0.78 min−1 (P = 0.068, Fig. 7) whereas the frequency of the low-frequency, high-amplitude contractions decreased to 0.46 ± 0.06 min−1 (P = 0.05, Fig. 7).
The main finding of this study was that the antispasmodic drug OB inhibited T-type Ca2+ channels at concentrations that are clinically significant, therefore representing an additional potential mechanism of action of this drug. OB therefore appears to have diverse mechanisms of action, targeting a variety of colonic channels and receptors.
T-type Ca2+ channels have been identified in a wide array of tissues, including neurons, cardiac pacemaking cells, and smooth muscle cells from bronchial, vascular, uterine, and gastrointestinal tissues (34). T-type Ca2+ channels may play a role in electrophysiological pacing in tissues when present, including the gastrointestinal tract.
In the gastrointestinal tract of various species, T-type or T-type-like Ca2+ currents have been recorded from stomach (30), small intestine (6, 27), and colon (10, 31, 32), the latter representing the predominant tissue affected by OB. The T-type Ca2+ channel family is made up of three genes, CACNA1G, CACNA1H, and CACNA1I, which respectively encode the T-type Ca2+ channels CaV3.1, CaV3.2, and CaV3.3 (also known as α1G, α1H, and α1I) (22). Cloned “in silico” by human brain EST tags, CaV3.1 in the gastrointestinal tract has been confirmed by Northern blot (20) and RT-PCR (21). In mouse, nifedipine- and cadmium-resistant, mibefradil-sensitive currents have been identified in jejunum myocytes, and loss of CaV3.2 expression decreased the upstroke and slow wave frequency of slow waves by ICC (6). The presence of CaV3.3 in gastrointestinal smooth muscle is not clear, with evidence both for (21) and against its expression (6).
The main findings of this study were that OB reversibly blocked the gastrointestinal T-type Ca2+ channels CaV3.1 (α1G) and CaV3.2 (α1H). OB was also effective at blocking CaV3.3 (α1I), with a lower IC50 suggesting that OB had a greater affinity to this T-type channel subunit. The IC50s for all three channels were within the range (∼1–10 μM) of the known concentration of OB in gastrointestinal smooth muscle (5). These values are significantly less than the IC50 values for two other gastrointestinal smooth muscle channels when expressed heterologously, the L-type Ca2+ channel CaV1.2 [alpha subunit α1C with auxiliary subunit β2, 2.3 ± 0.5 μM (29)] and Na+ channel NaV1.5 (α pore encoded by SCN5A, 8.8 ± 2.4 μM, Table 1). Overall, OB blocks channels at IC50 values in the following order: CaV3.3 < CaV3.1 = CaV3.2 < CaV1.2 < NaV1.5. Therefore, the block observed in this study may be clinically significant. This statement is supported by the effect of mibefradil on human sigmoid colon contractile activity, although the effect of mibefradil on the complex motor patterns of the human colon will require further study.
OB appears to block T-type Ca2+ channels at doses in a similar range as the established T-type Ca2+ blocker, mibefradil. IC50 values for the block of transfected Ca2+ channels by mibefradil for CaV3.1 (0.27 μM), CaV3.2 (0.14 μM), CaV3.3 (0.29 μM) (17, 28) were close to the IC50 values of OB reported here (Table 1). However, mibefradil does not appear to be as specific for T-type Ca2+ channels as OB with a threefold selectivity for mibefradil on T-type channels over NaV1.5 channels compared with 9-fold for otilonium (18, 28).
In conclusion, OB inhibits T-type Ca2+ channels at clinically relevant concentrations. This effect on T-type Ca2+ channels is likely limited to gastrointestinal T-type Ca2+ channels as OB is poorly absorbed, with plasma levels at least 1,000 times lower than those reached in the colonic wall.(5) Therefore, inhibition of gastrointestinal T-type Ca2+ channels may contribute to the inhibitory effects of OB on smooth muscle contractility. Whether this effect is specific for OB or is a general property of antispasmodics will require future investigation.
This research was supported by National Institutes of Health grants DK52766 and DK57061 (G. Farrugia) and NS38691 (E. Perez-Reyes) and an unrestricted grant from Menarini Ricerche.
We thank Kristy Zodrow for secretarial assistance.
- Copyright © 2010 the American Physiological Society