Coordinated gastrointestinal motility is required for. A Mechanosensitive Calcium Channel in Human Intestinal Smooth Muscle Cells

Transcription

1 GASTROENTEROLOGY 1999;117: A Mechanosensitive Calcium Channel in Human Intestinal Smooth Muscle Cells GIANRICO FARRUGIA,*, ADRIAN N. HOLM,*, ADAM RICH,*, MICHAEL G. SARR, JOSEPH H. SZURSZEWSKI,*, and JAMES L. RAE* *Department of Physiology and Biophysics, Division of Gastroenterology and Hepatology, and Department of Surgery, Mayo Clinic, Rochester, Minnesota Background & Aims: Gastrointestinal smooth muscle strips devoid of enteric nerve cells can contract in response to stretch, suggesting that mechanosensitivity and mechanotransduction can occur at the level of the smooth muscle cell. The aim of this study was to determine whether stretch-activated calcium channels are present in gastrointestinal smooth muscle cells. Methods: Whole-cell and single-channel calcium currents were measured from human jejunal circular smooth muscle cells in response to increased intracellular pressure, bath perfusion, and membrane stretch. Results: At 10 mm Hg positive pressure, peak calcium current increased from pa to pa. Bath perfusion at 10 ml/min increased calcium current from pa to pa. Single-channel open probability increased in response to negative pipette pressure. All increases were blocked by nifedipine. Conclusions: A stretch-activated, nifedipine-sensitive calcium channel is present in human jejunal circular smooth muscle cells. The channel is activated by both an increase in intracellular pressure and by external shear forces. The presence of a stretch-activated calcium channel in gastrointestinal smooth muscle cells may allow the smooth muscle cells to act directly as mechanotransducers and to participate in the regulation of smooth muscle tone and intestinal motility. Coordinated gastrointestinal motility is required for efficient digestion and absorption of food and the elimination of waste products. Abnormalities in gastrointestinal motility have been implicated in numerous gastrointestinal disorders such as constipation, 1 pseudoobstruction, 2 and irritable bowel syndrome. 3,4 Gastrointestinal motility is coordinated by an interplay of extrinsic nerves, the enteric nervous system, interstitial cells of Cajal, and smooth muscle cells. Gastrointestinal smooth muscle strips devoid of enteric nerve cells can contract in response to stretch. 5 This observation suggests that mechanosensitivity and mechanotransduction can occur at the level of the smooth muscle cell, although the mechanism behind stretch-induced contraction is unknown. Mechanosensitivity is central to enabling living organisms to sense and respond to changes in external mechanical stimuli such as stress, strain, osmotic and mechanical pressure, shear forces, acceleration, vibration, and sound. 6 At a cellular level, the elementary unit for mechanotransduction is the ion channel. Mechanosensitive ion channels have been described in plants, bacteria, fungi, and a multitude of vertebrate and invertebrate animal cells. 7 Mechanosensitive ion channels are currently classified into stretch-sensitive channels, displacement-sensitive channels, and shear stress sensitive ion channels. 6,7 Stretch-sensitive ion channels are divided into stretch-activated (SA) ion channels and the less common stretch-inactivated ion channels. 6,7 SA ion channels are further divided according to their ion selectivity. Calcium influx is essential to maintaining intestinal smooth muscle contractility. The main pathway for entry of Ca 2 into human jejunal circular smooth muscle cells is through L-type Ca 2 channels. 8 Although SA Ca 2 channels have been inferred to exist, 9 none have been recorded at the single-channel or whole-cell current level. To determine whether SA Ca 2 channels are present in gastrointestinal smooth muscle cells, isolated human jejunal circular smooth muscle cells were exposed to changes in intracellular pressure and to external shear forces. A nifedipine-sensitive SA Ca 2 channel was found to be present in human and canine jejunal circular smooth muscle cells. Activation of the channel occurred at pressures normally generated in the gastrointestinal tract, suggesting that single gastrointestinal smooth muscle cells may act as mechanotransducers and participate in the regulation of smooth muscle tone and contractile activity. Abbreviation used in this paper: SA, stretch-activated by the American Gastroenterological Association /99/$10.00

2 October 1999 MECHANOSENSITIVE INTESTINAL SMOOTH MUSCLE 901 Materials and Methods Cells Single, isolated, circular smooth muscle cells were obtained from human jejunum. The use of human jejunum was approved by the Institutional Review Board of the Mayo Clinic. Human tissue was obtained as surgical waste tissue during gastric bypass operations performed for morbid obesity. Procedures used to obtain single, relaxed, circular smooth muscle cells were as previously described. 10,11 Patch-Clamp Recordings Whole-cell patch-clamp recordings were made using traditional whole-cell techniques. Whole-cell and singlechannel recordings were obtained using Kimble KG-12 glass pulled on a P-97 puller (Sutter Instruments, Novato, CA). Electrodes were coated with R6101 (Dow Corning, Midland, MI) and fire polished to a final resistance of 3 5 M. Currents were amplified, digitized, and processed using an Axopatch 200-A amplifier, a Digidata 1200, and pclamp 7 software (Axon Instruments, Foster City, CA). The pulse protocols used are shown in the figures. Drugs were applied by complete bath changes with the solution containing the drug. The time for a complete bath solution change (4 bath volume) was 30 seconds. Whole-cell records were sampled at 2 khz and filtered at 1 khz, and single-channel records were sampled at 5 khz and filtered at 2 khz. Positive pressure was applied to the recording pipette to increase intracellular pressure and examine the effects of an increase in intracellular pressure on inward Ca 2 currents. After obtaining a gigaseal, whole-cell access was obtained by gentle suction. Inward currents were measured for at least a 5-minute equilibrium period. Intracellular pressure was increased using a battery-powered Bio-Tek DPM-1 pneumatic transducer. A range of pressures varying from 5 to 20 mm Hg ( kpa) were used. The on-cell patch-clamp technique was used to obtain single-channel recordings. After obtaining a gigaseal control, records were obtained using a series of 300-millisecond pulses from a holding voltage of 80 to 30 mv. A 2-second interval was used between pulses to avoid channel inactivation. Negative pressure was applied to the pipette to activate SA ion channels in the on-cell patch. All records were obtained at room temperature (22 C). Solutions For whole-cell current and single-channel measurements, the pipette solution contained (in mmol/l) Cs, 145; Cl, 20; EGTA, 2; HEPES, 5; methanesulfonate, 130; and Ba 2, 80; Cl, 160; HEPES, 5, respectively. The bath solution contained Na, 146; K, 4.7; Cl, 154.7; Ca 2, 2; HEPES, 5, for whole-cell recordings and K, 150; Cl, 154; Ca 2,2; HEPES, 5, for single-channel recordings. Data Analysis Data were analyzed using Clampfit or custom macros in Excel (Microsoft, Redmond, WA) and comparisons between currents made at peak measured values. Single-channel open probability was measured from all-points amplitude histograms. Single-channel current amplitude was measured from gaussian fits to the histograms as well as from single, well-resolved openings. The paired Student t test was used to evaluate statistical significance. Results Effect of Positive Pressure Single human jejunal circular smooth muscle cells were patch-clamped under recording conditions that favored inward Ca 2 current. After traditional whole-cell access was obtained, whole-cell currents were recorded for at least a 5-minute period. The pipette pressure was then increased to 10 mm Hg above atmospheric pressure, and whole-cell currents were recorded immediately. At 10 mm Hg positive pressure, peak inward Ca 2 current increased from to pa (n 9, P 0.01; Figure 1). Positive pressure did not change cell capacitance. Cell capacitance was 56 4 pa at atmospheric pressure and pa at 10 mm Hg positive pressure (n 9, P 0.05). When normalized for cell capacitance, peak inward Ca 2 current was 0.64 pa/pf at atmospheric pressure and 0.97 pa/pf at 10 mm Hg positive pressure (n 9, P 0.05). The increase in Ca 2 current peaked within 30 seconds of initiation of positive pressure. Voltage at peak inward current was unchanged with positive pressure. Removal of positive pressure did not immediately return peak inward current to baseline. Figure 1. Activation of inward Ca 2 current by increased intracellular pressure. (A) Peak inward current at atmospheric pressure measured 48 pa. (B) Increase in intracellular pressure to 10 mm Hg above atmospheric increased peak inward Ca 2 current to 110 pa. (C) No change occurred in the voltage at which peak inward current was measured.

3 902 FARRUGIA ET AL. GASTROENTEROLOGY Vol. 117, No. 4 Figure 2. Increase in inward current evoked by sequential increases in intracellular pressure. (A) Whole-cell inward current. Peak inward Ca 2 current was 26 pa at atmospheric pressure, 37 pa at 10 mm Hg above atmospheric pressure, 51 pa at 20 mm Hg, and 16 pa at 20 mm Hg in the presence of nifedipine (1 µmol/l). (B) Respective current-voltage relationships:, 0mmHg;, 10mm Hg;,20mmHg;,20mmHg nifedipine (10 µmol/l). Rather, a slow decrease in current was observed over 5 minutes. Initial effects of positive pressure were first found at 5 mm Hg, the lowest pressure applied. A pressure-current response curve was obtained in 4 cells. Mean inward Ca 2 current measured pa in control records, pa at 10 mm Hg, at 15 mm Hg, and pa at 20 mm Hg (all P 0.05 compared with control, Figure 2). Increases in pressure above 20 mm Hg were usually accompanied by a sudden change in cell shape from fusiform to round and a decrease in peak Ca 2 current. In 3 cells, an increase in positive pressure from 10 to 20 mm Hg did not further increase Ca 2 current. The increase in inward Ca 2 current was blocked by nifedipine (10 µmol/l, pa before and pa after nifedipine, n 4, P 0.05; Figure 2). Bath Perfusion To determine whether external mechanical stimulation increased Ca 2 current, the bath was perfused with normal Ringer s solution at 5, 10, and 20 ml/min in separate experiments. At 5 ml/min, inward Ca 2 current increased from pa to pa (n 7, P 0.05); at 10 ml/min, inward Ca 2 current increased from pa to pa (n 21, P 0.05; Figure 3). At 20 ml/min, inward Ca 2 current peaked at pa (n 4, P 0.05). When normalized for cell capacitance, mean peak inward current was 1.45 pa/pf before and 1.81 pa/pf after perfusion at 10 ml/min. The increase in inward current was not secondary to a change in other conductances that contribute to the whole-cell current. Input resistance, measured at 70 mv, was G before and 12 2G after perfusion at 10 ml/min (P 0.05, n 18). Also, the whole-cell current evoked by the voltage pulse that gave rise to peak inward Ca 2 current, when measured at 100 milliseconds from the start of the voltage step, where the contribution of L-type Ca 2 current is negligible, was pa before and pA(P 0.05, n 18) after perfusion at 10 ml/min. Perfusion-current response curves were also obtained (Figure 4). Increase in inward Ca 2 current was completely blocked by nifedipine (1 µmol/l). Increase in inward Ca 2 current was first noted within 8 seconds of Figure 3. Percentage increase in peak inward current after perfusion (10 ml/min) for each of the cells studied. Wide bar represents the mean increase in peak inward current (31% 2%; n 21, P ).

4 October 1999 MECHANOSENSITIVE INTESTINAL SMOOTH MUSCLE 903 Figure 4. (A) Activation of inward Ca 2 current by external shear force. Peak control inward Ca 2 current at a bath flow rate of 0 ml/min measured 66 pa. Inward current increased to 70 pa at a bath flow rate of 5 ml/min, 78 pa at 10 ml/min, and 97 pa at 20 ml/min. (B) Peak inward current returned to baseline 6 minutes after bath perfusion was stopped. Numbers within graph indicate flow rate (ml/min) at that point. initiation of bath perfusion and returned to baseline levels 4 8 minutes after perfusion was stopped (Figure 4). Single-Channel Recordings Single-channel recordings were obtained from on-cell patches to determine the single-channel conductance that underlied the mechanosensitive whole-cell current. Barium (80 mmol/l) was used as the charge carrier in the pipette to increase the amplitude of single-channel openings, and 150 mmol/l KCl was used in the bath to depolarize the resting membrane potential to 0 mv. Single-channel openings consistent with calcium channels were observed in 5 of 14 patches at atmospheric pressure. Negative pressure ( 40 mm Hg) increased the open probability of the calcium channel in all 5 patches (Figure 5). Channel openings were inhibited in the presence of nifedipine (10 µmol/l; Figure 5). The single-channel conductance was 16 3 picosiemens (ps) in 80 mmol/l barium at atmospheric pressure and 17 3 ps at negative pipette pressure ( 40 mm Hg). In 5 on-cell patches with no measurable single-channel activity at atmospheric pressure, negative pressure ( 40 mm Hg) activated a channel with similar characteristics to the baseline channel observed in the 5 patches. Rundown was observed in all patches, with loss of channel activity within 5 minutes despite continued application of negative pressure. In 3 patches infrequent openings of second channel type ( 12 ps) were also seen. This channel was not studied further. Discussion This study provides whole-cell current and singlechannel evidence for the existence of SA, nifedipinesensitive L-type Ca 2 channels and of their presence in human gastrointestinal smooth muscle cells. The gastrointestinal tract must continually adapt to changes in intraluminal content as chyme boluses traverse its length. An SA ion channel located on smooth muscle cells may allow smooth muscle to respond directly to distention or to contraction of smooth muscle cells without involvement of neuronal structures. SA ion channels may also provide a mechanism through which gastrointestinal smooth muscle can regulate smooth muscle tone. A similar mechanism has been described for arterial smooth muscle. 12 The presence of SA ion channels in gastrointestinal smooth muscle may require a change in the current view of gastrointestinal smooth muscle as a pure motor

5 904 FARRUGIA ET AL. GASTROENTEROLOGY Vol. 117, No. 4 Figure 5. Single-channel recordings obtained from an on-cell patch with 80 mmol/l barium as the charge carrier in the pipette. Records were obtained using the voltage protocol shown in the inset and capacitance changes subtracted off-line. Little channel activity was noted at atmospheric pipette pressure (A, open probability [NPo] 0.005). Negative pressure ( 40 mm Hg) activated a nifedipine-sensitive channel (B, NPo 0.06; C, NPo ). organ to an organ that may exhibit both sensory and motor function. Changes in gastrointestinal smooth muscle shape, either as a result of passive stretching or secondary to a contraction, may activate SA Ca 2 channels, increasing Ca 2 influx and triggering intracellular messenger cascades, including the contractile cascade. The smooth muscle layers of the gastrointestinal tract are continually exposed to large changes in intraluminal pressure and wall tension, during both contraction and distention. Because the SA Ca 2 channel discussed in the present study was activated by changes in intracellular pressure and by perfusion-induced shear stress, the SA Ca 2 channel may participate in the regulation of normal intestinal function. SA Ca 2 channels in the smooth muscle membrane may enable a smooth muscle cell to respond directly to acute changes in the local environment of the cell, such as changes in membrane tension during passage of a food bolus, or to chronic changes such as changes in cell size (hypertrophy) and changes in the luminal or intramural resistance caused by strictures and inflammation. The pathophysiology of gastrointestinal motility disorders is to a large extent unknown. In cardiac myocytes, SA ion channels have been implicated in the pathophysiology of compensatory hypertrophy and cardiac arrhythmias. 16 Although it is attractive to hypothesize that abnormalities in SA ion channel expression and regulation may underlie gastrointestinal motility disorders characterized by abnormalities in the response of the gastrointestinal wall to stretching, direct evidence for this hypothesis is presently lacking. Recently there has been renewed interest in the mechanisms by which gastrointestinal smooth muscle cells communicate with the pacemaker cells of the gut the interstitial cells of Cajal. At the electronmicroscopic level, numerous peg and socket junctions have been shown to exist between interstitial cells of Cajal (sockets) and smooth muscle cells (pegs). 17 It has been proposed that these peg and socket junctions

6 October 1999 MECHANOSENSITIVE INTESTINAL SMOOTH MUSCLE 905 transduce mechanical events in interstitial cell of Cajal to smooth muscle cells. A mechanosensitive Ca 2 channel seems an ideal interface to couple these 2 cell types. The mechanism by which tension alters ion channel gating is not well understood. We found that a change in pipette pressure increased open probability of the SA Ca 2 channel without altering single-channel conductance. This observation is in agreement with previous findings in studies of other SA channel types, suggesting that a direct stretching of the conductance pore is not involved. An elastic transduction model for SA ion channels has been proposed. In this model the selectivity pore is not affected by membrane tension, and therefore channel ionic selectivity is unaltered by stretching. Rather, the SA ion channel is modeled as having 2 states: a more distensible, smaller, closed state and a larger, stiffer, open state. An increase in tension favors the larger, stiffer, open state resulting in a shift in the thermal equilibrium of the SA ion channel toward the open state. 7 The mechanism by which tension is transduced to the mechanosensitive ion channel is also poorly understood but appears to involve the cytoskeleton. 6 In the present study the effects of perfusion on inward Ca 2 current appeared to be more quickly reversible than the effects of positive pressure. A possible reason for this discrepancy is that positive pressure is more likely to disrupt the cytoskeleton than perfusion. Disruption of the actin cytoskeleton has been shown to enhance ion channel mechanosensitivity. 6,18 The data reported in this study provide evidence for a novel SA nifedipine-sensitive Ca 2 channel and suggest that isolated human jejunal circular smooth muscle cells express the SA nifedipine-sensitive Ca 2 channel. SA of the Ca 2 channel may provide a novel mechanism by which gastrointestinal smooth muscle cells can act as both sensory and motor organs. Whether this mechanism is active under normal physiological conditions remains to be determined. References 1. Bassotti G, Chiarioni G, Vantini I, Betti C, Fusaro C, Pelli MA, Morelli A. Anorectal manometric abnormalities and colonic propulsive impairment in patients with severe chronic idiopathic constipation. Dig Dis Sci 1994;39: Camilleri M, Phillips SF. Acute and chronic intestinal pseudoobstruction. Adv Intern Med 1991;36: Small PK, Loudon MA, Hau CM, Noor N, Campbell FC. Large-scale ambulatory study of postprandial jejunal motility in irritable bowel syndrome. Scand J Gastroenterol 1997;32: Gorard DA, Libby GW, Farthing, MJ. Ambulatory small intestinal motility in diarrhoea predominant irritable bowel syndrome. Gut 1994;35: Burnstock G, Prosser CL. Responses of smooth muscles to quick stretch; relation of stretch to conduction. Am J Physiol 1960;198: Sachs F. Mechanical transduction in biological systems. Crit Rev Biomed Imaging 1988;16: Morris CE. Mechanosensitive ion channels. J Membr Biol 1990; 113: Farrugia G. Modulation of ionic currents in isolated canine and human jejunal circular smooth muscle cells by fluoxetine. Gastroenterology 1996;110: Gannier F, White E, Garnier D, Le Guennec JY. A possible mechanism for large stretch-induced increase in intracellular calcium in isolated guinea-pig ventricular myocytes. Cardiovasc Res 1996;32: Farrugia G, Rae JL, Sarr MG, Szurszewski JH. Potassium current in circular smooth muscle of human jejunum activated by fenamates. Am J Physiol 1993;265:G873 G Farrugia G, Rae JL, Szurszewski JH. Characterization of an outward potassium current in canine jejunal circular smooth muscle and its activation by fenamates. J Physiol 1993;468: Davis MJ, Donovitz JA, Hood JD. Stretch-activated single-channel and whole cell currents in vascular smooth muscle cells. Am J Physiol 1992;262:C1083 C Bustamante JO, Ruknudin A, Sachs F. Stretch-activated channels in heart cells: relevance to cardiac hypertrophy. J Cardiovasc Pharmacol 1991;17:S110 S Kent RL, Hoober JK, Cooper G, IV. Load responsiveness of protein synthesis in adult mammalian myocardium: role of cardiac deformation linked to sodium influx. Circ Res 1989;64: Simpson DG, Sharp WW, Borg TK, Price RL, Terracio L, Samarel AM. Mechanical regulation of cardiac myocyte protein turnover and myofibrillar structure. Am J Physiol 1996;270:C1075 C Hansen DE, Craig CS, Hondeghem LM. Stretch-induced arrhythmias in the isolated canine ventricle. Evidence for the importance of mechanoelectrical feedback. Circulation 1990;81: Thuneberg L, Peters S. Peg and socket junctions: accumulating evidence of a role in intestinal mechanotransduction. Neurogastroenterol Motil 1998;10: Guhary F, Sachs F. Stretch-activated single ion channel currents in tissue-cultured embryonic chick skeletal muscle. J Physiol 1984;352: Received January 25, Accepted June 29, Address requests for reprints to: Gianrico Farrugia, M.D., Mayo Clinic, Guggenheim 8, 200 First Street Southwest, Rochester, Minnesota Farrugia.gianrico@mayo.edu; fax: (507) Supported by National Institutes of Health grants DK17238, DK52766, DK39337, EY03282, and EY06005.