INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE COLONIC MUCOSA exhibits the ability to absorb but also to secrete electrolytes and water, the former mainly by the surface cells, whereas the latter is thought to occur in the colonic crypts (23), although there are reports stating that surface cells also are capable of secretion (12). There is normally a net absorption of fluid from the intestinal contents, the result being formed fecal material in the rectum. In some pathological conditions, however, there is a net secretion of fluid resulting in more or less severe diarrhea. The expulsion of chloride ions from the apical parts of the crypt cells is thought to be the driving force for this secretion. The movement of chloride ions cannot occur without a number of different ion channels and transport systems. A previously suggested (1) simplified secretory model consists of an apical Cl channel, a basolateral 2Cl-K⁺-Na⁺ cotransport system, a basolateral K⁺ channel, the Na⁺-K⁺-ATPase, and intercellular transport of sodium ions. The Cl secretion can be activated through different intracellular pathways, e.g., via activation of an adenylate cyclase, which increases the levels of adenosine 3',5'-cyclic monophosphate (cAMP), and mobilization of intracellular calcium via the inositol 1,4,5-trisphosphate system. In this study we have used the colon carcinoma cell line HT-29, which is a commonly used model for studies of Cl secretion, to investigate the effects of carbachol (CCh), a parasympatheticomimetic agent known to increase intracellular calcium. With the perforated whole cell mode of the patch-clamp technique, we were able to show that CCh caused a pattern of depolarizing oscillations of the membrane potential, which was generated by Cl and K⁺ permeabilities. With the fluorescent dye fluo 3 we could further show that the intracellular calcium concentration ([Ca²⁺]_i) also exhibited a pattern of oscillations, linking these events to the changes in membrane potential. In inside-out patches, calcium-dependent Cl channels could be demonstrated.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Cells. The colon carcinoma cell line HT-29 was grown in Dulbecco's modification of Eagle's medium supplemented with bovine calf serum (5% vol/vol), L-glutamine (2 mM), penicillin (10⁵ U/l), and streptomycin (100 mg/l) and was kept in an atmosphere of 5% CO₂ at 37°C. The cells were seeded in petri dishes (35 mm, Corning Glass, Corning, NY). Only isolated cells, grown for up to 36 h, were used for the patch-clamp experiments (with or without simultaneous calcium measurements). For calcium measurements on cell monolayers, experiments were performed up to 7 days after subculture. The experiments were performed at room temperature (20-22°C).

Solutions. The normal extracellular bathing solution had the following composition (in mM): 140 NaCl, 4 KCl, 1.5 CaCl₂, 1 MgCl₂, 5 glucose, and 10 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), adjusted to pH 7.4 with NaOH. In the substitution experiments, Cl was replaced with glucuronate on a mole per mole basis. All other solutions are described in RESULTS. Junction potentials with different combinations of solutions were calculated using the Axoscope program from Axon Instruments (Foster City, CA). The membrane potentials were corrected for after the experiments were performed. The nystatin solution was made as a stock solution with 5 mg of nystatin dissolved in 100 µl of dimethyl sulfoxide (DMSO). From this stock solution 10 µl were dissolved in 3 ml of pipette solution, giving a final nystatin concentration of 180 µmol/l.

Whole cell measurements. The pipettes were pulled from aluminosilicate glass (Hilgenberg, Malsfeld, Germany; OD 1.6 mm, ID 1.28 mm) on a two-stage vertical puller. The tips were covered with Sylgard (Dow Corning, Seneffe, Belgium) and fire polished before use. The pipette resistance was 1.5-5 M when filled with the different intracellular solutions. To obtain a stable whole cell recording and to minimize the disturbance of the intracellular solution, the perforated patch technique was used with nystatin as the perforating agent (10). The tips were filled with a nystatin-free solution, whereas the rest of the pipette was backfilled with the nystatin solution described above. After formation of a giga-seal (>2 G), the capacitative current induced by a test pulse slowly increased and became more rapid. From this current, the series resistance could be calculated. The experiments started when the series resistance was <30 M, which was usually achieved within 10 min. The final series resistance was 7.5-30 M, depending on the size of the pipette. Before the experiments started, the series resistance and whole cell capacitance were compensated for. Conventional whole cell experiments, i.e., with light suction applied to break the cell membrane, was not successful because of resealing of the membrane with accompanying increase in access resistance to values >50 M. During the experiments, the chamber was continually perfused hydrostatically with normal extracellular solution or one of the test solutions. The perfusion system allowed an exchange of solution within 5-15 s. Whole cell voltages and currents were recorded using an Axopatch 200A patch-clamp amplifier together with the pCLAMP (5.5 or 6.03) and Axotape software (Axon Instruments). The data were also displayed on a pen recorder.

Cell-attached and inside-out experiments. The experimental procedure was the same as described above concerning pipettes, perfusion, data handling, and so forth. The pipettes were of similar size as in the whole cell experiments.

Patch-clamp conventions. In all experiments, a positive current is a membrane outward current, i.e., positive ions moving from the inside of the membrane to the outside or negative ions moving the opposite way.

Calcium measurements. The [Ca²⁺]_i was measured using the fluorescent calcium indicator fluo 3 (Molecular Probes, Eugene, OR). The cells were incubated at room temperature in 8-10 µM fluo 3 containing 0.04% Pluronic and 0.4% DMSO for 30-60 min. The concentration and time used for incubation did not seem to be critical for the loading of fluo 3. The fluorescence was recorded using an inverted microscope (IM-35, Zeiss, Oberkochen, Germany) equipped with an epifluorescent system. A 75-W halogen lamp was used for fluorescence excitation, and a filter combination of 480, 40×/505/510-565 nm (bandwidth, dichroic, barrier; Chroma Technology, Brattleboro, VT) was used for fluorescence excitation and emission of fluo 3. The emitted fluorescent light was monitored using a photomultiplier (R56000U, Hamamatsu, Shizuoka-ken, Japan) attached to one of the microscope outlets. The signal from the photomultiplier was fed into a current-voltage converter, and the potential was amplified (10×) and low-pass filtered (3 Hz) using a direct current amplifier (Princeton Applied Research, Princeton, NJ). The light signal, together with the current or potential signal, was acquired and stored in a computer using the pCLAMP acquisition system together with the Axotape 2.0 software (Axon Instruments).

Chemicals. All chemicals were purchased, unless otherwise indicated, from Sigma Chemical (St. Louis, MO).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

General properties. The whole cell capacitance can be used as a measure of cell size. The cells in these experiments had a capacitance of 11.5 to 23.8 pF, corresponding to a cell diameter of 19.1-27.5 µm, assuming a spherical cell and a specific membrane capacitance of 1 µF/cm². The mean diameter as measured from 23 cells was 22.6 ± 9.3 µm (means ± SD).

Membrane potential changes. The resting membrane potential in 23 cells was 69.2 ± 12.4 mV (means ± SD). When the cells were exposed to 100 µM CCh, they depolarized to a mean value of 46.4 ± 8.6 mV (means ± SD, n = 23). This initial depolarization was, in 16 of 23 cells, followed by a pattern of membrane potential oscillations, i.e., the membrane potential oscillated between the resting value and a depolarized value. Figure 1 shows six cells with different patterns of oscillation. Some cells oscillated to the same depolarized potential during the entire CCh exposure, whereas others showed oscillating peaks that subsided with time. The frequency of the oscillations varied between two and eight per minute. The depolarized value was close to the equilibrium potential for Cl (48 mV as calculated with the Nernst equation, pipette solution in Fig. 1). To further investigate if this depolarization was caused by a Cl permeability, the following sets of experiments were performed.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Oscillations of membrane potential (Em) when cells were exposed to 100 µM carbachol (CCh; as indicated by bars). Six cells exhibited different patterns of oscillation (represented as A-F). Cell D oscillates to same potential (approximately 45 mV) during entire exposure to CCh, whereas cell B shows only an initial depolarizing peak to 41 mV and depolarizing peaks then decrease in amplitude. Equilibrium potential for Cl () = 48 mV in these experiments. Note also that effect of CCh can be reproduced in a single cell (B). Pipette solution (in mM): 76 K2SO4, 10 KCl, 10 NaCl, 1 MgCl2, and 10 HEPES, adjusted to pH 7.25 with KOH. Junction potential correction 9 mV.

Whole cell voltage-clamp experiments. The same type of oscillations was also seen in 10 of 11 current recordings. The frequency varied between one and six per minute. Figure 2 shows two cells voltage clamped froma holding potential of 60 mV to the equilibrium potentials for K⁺ and Cl (90 and 0 mV, respectively, see inset in Fig. 2) (20). In this subset of experiments, the pipette solution contained (in mM) 145 KCl, 1.5 CaCl₂, 1 MgCl₂, and 10 HEPES to further widen the gap between the equilibrium potentials for K⁺ and Cl. Figure 2 shows that there is both a Cl and a K⁺ current (at 90 and 0 mV, respectively). Measured from 11 cells, the mean peak Cl current was 1,540 ± 174 pA, and the K⁺ current was 195 ± 32 pA. The onset of the two components differed with a mean time to peak current of 5.7 ± 3.1 s (Cl current) and 18.5 ± 3.9 s (K⁺ current).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Membrane current response of 2 cells exposed to CCh (100 µM). Cells were voltage clamped between a holding potential (60 mV) and equilibrium potentials for K+ and Cl (90 mV and 0 mV, respectively). Pulse sequence (inset) was delivered once every second. There is both an inward Cl current and an outward K+ current, the former being dominant, indicating a net outflow of Cl. Left cell shows an oscillating pattern of holding (60 mV) current similar in frequency to potential oscillations (see Fig. 1). Pipette solution (in mM): 144 KCl, 1.5 CaCl2, 1 MgCl2, and 10 HEPES adjusted to pH 7.25 with KOH. Junction potential correction was not made in these experiments.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Current-voltage relationship during CCh (100 µM) exposure. Cell was clamped from a holding potential of 69 mV in 10-mV steps from 129 to 11 mV. Current was measured 40 ms from start of pulse. A: during CCh exposure, cells exhibited a dramatic increase in current. A slow inactivation can be seen during the first negative voltage steps. B: current-voltage curves in control solution () and in CCh (). Slope conductance between 80 and 40 mV increased from 200 pS in control solution to 10.4 nS in CCh solution. Im, membrane current. Pipette solution same as described in Fig. 1 legend. Junction potential correction 9 mV.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effect of low external Cl on CCh-induced membrane currents. A: cell was clamped to different potentials and exposed to 100 µM CCh for 20 s in control solution (149 mM Cl) and low- Cl solution (22 mM Cl, equal to pipette concentration; Cl replaced with glucuronate), as indicated. B: current-voltage relationships for evoked currents from 9 cells in normal extracellular solution () and 7 cells in low-Cl solution (). In normal extracellular solution, current reverses at 41.6 ± 2.2 mV, close to . In low-Cl solution, reversal potential changed to 3.2 ± 2.0 mV. Pipette solution same as described in Fig. 1 legend. Junction potential corrections were 9 mV (control solution) and 3 mV (low-Cl solution).

Calcium measurements. Previous studies in HT-29 cell monolayers have shown that CCh causes an increase in intracellular calcium. When calcium was omitted in the extracellular solution during CCh exposure, [Ca²⁺]_i decreased (5, 16). The same type of response in monolayers was seen in the present study. Cells incubated with the fluorescent agent fluo 3 exhibited a dramatic increase in fluorescence at a wavelength of 525 nm, consistent with an increase in intracellular calcium when exposed to 100 µM CCh (n > 50, not shown). In the majority of cells exposed to CCh investigated using fluo 3, different types of oscillation in [Ca²⁺]_i were observed. We were able to make simultaneous measurement of the membrane potential and the intracellular calcium in six cells. There was a close correlation between the changes in fluorescent-calcium signal and membrane potential (Fig. 5A). Close scrutiny of the recordings showed that the increase in fluorescence preceded the depolarization by a few seconds (Fig. 5B). These results suggest that the underlying mechanism of the oscillating currents-potential changes could be openings and closures of calcium-dependent Cl channels triggered by CCh-induced calcium oscillations in the cytoplasm.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Simultaneous measurements of Em and fluorescent signal from fluo 3 (F525; an indicator of intracellular Ca2+) from an isolated cell during exposure to CCh. A: when exposed to 10 µM CCh (indicated by bar), fluorescent-Ca2+ signal and Em oscillate with same frequency. B: figure shows first depolarizing peak (from A) on an expanded time scale. It can be seen that the Ca2+ signal precedes the potential change by a few seconds. Pipette solution same as described in Fig. 1 legend. Junction potential correction 9 mV.

Cell-attached experiments. In this mode, and with standard extracellular solution in the pipette, CCh induced membrane currents in 11 of 16 patches. To determine the nature of this current, further experiments were performed with different pipette solutions. Figure 6, A and B, shows two cells exhibiting CCh-induced currents. In the first, the pipette contained normal extracellular solution, and it can be seen that the current reverses at ~0 mV. With a low-Cl solution in the pipette (15 mM), the reversal potential should, ideally, change 58 mV in the positive direction. However, in these experiments, the reversal potential changed to 46.2 ± 5.3 mV. The use of 144 mM KCl in the pipette, on the other hand, did not change the reversal potential. It can also be seen that the patch current oscillates with approximately the same frequency as in the whole cell experiments (cf. Figs. 1-2). Figure 6C summarizes the results using a different pipette solution. The data is consistent with a Cl current being activated by CCh in these patches.

View larger version (13K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Fig. 6.   Cell-attached recordings during CCh (100 µM) exposure. A: with standard extracellular solution in pipette, current reversed at a pipette potential of about 0 mV. B: with a low-Cl solution in pipette (15 mM Cl, Cl replaced with glucuronate), current reversed at a relative Em (Vrel) of ~45 mV. Increasing K+ (144 mM) in pipette solution did not change reversal potential significantly (P  0.05). C: diagram summarizes findings with different pipette solutions. Vrel = Vm  Vrest, where Vm is patch membrane potential and Vrest is resting membrane potential. Junction potential correction was 11 mV (15 mM Cl) or 4 mV (144 mM K+).

Single-channel experiments. To directly study the presence of calcium-activated Cl channels, we performed experiments on inside-out patches. In excised patches, channel activity could be seen in 19 of 20 patches. Figure 7 shows current from one of four patches exposed to solutions with different calcium concentrations. The patch contained at least four channels that opened and closed randomly when exposed to the normal bathing solution. With the patch being clamped to 67 mV, the current direction is positive, indicating a flow of chloride ions from the pipette to the bath solution. When the calcium concentration was changed to 50 nM [intracellular resting calcium level (5, 16)], the channels closed and stayed closed until the calcium was again increased to 1,000 nM. Five hundred nanomolar calcium was not sufficient to activate the channels. The current-voltage relationship for the channels observed in the four patches exposed to low calcium (as shown in Fig. 8), together with three patches not exposed to low calcium but examined at several potentials, gave a channel conductance of 19 pS (regression analysis, r = 0.99). The current reversed at 2 mV, which is close to but statistically not equal (P < .05) to the equilibrium potential for Cl (6 mV) using these solutions. The equilibrium potential for K⁺ was 90 mV.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Continuous current trace of an inside-out patch with several Cl channels. Membrane was clamped to 67 mV. Lowering Ca2+ (to 50 nM) in normal extracellular solution facing inside of membrane closed channels. Channels started to open again when Ca2+ was increased to 1,000 nM. Inset: amplitude histogram from 1st part of 4th trace. Pipette solution (in mM): 110 KCl, 30 KOH, 1.5 CaCl2, 1 MgCl2, 11 EGTA, and 10 HEPES, adjusted to pH 7.25 with KOH. Junction potential correction 7 mV.

View larger version (11K): [in this window] [in a new window]   Fig. 8.   Current-voltage diagram of unitary currents from 7 patches with Cl channels. Open symbols, patches exposed to normal Ca2+-containing extracellular solution only. Filled symbols, 4 patches exposed to different Ca2+ solutions (see Fig. 7). Conductance 19 pS. Bath contained normal extracellular solution. Pipette solution same as described in Fig. 7 legend. Junction potential correction 7 mV.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Secretion in the large intestine is thought to be maintained by an exit of chloride ions through Cl channels in the apical membrane of the epithelial cells. The Cl channels can be stimulated by a variety of secretagouges, e.g., acetylcholine and vasoactive intestinal peptide. These substances bind to receptors that control intracellular second messengers such as calcium or cAMP through G protein systems. The second messengers subsequently control the Cl channels.

HT-29 has been shown to contain several of the components necessary for secretion, e.g., different types of Cl channels (6-9, 14, 15, 18, 19), basolateral Na⁺-K⁺-Cl exchange mechanisms and a Na⁺-K⁺ pump (11). It has further been shown to contain muscarinic M3 receptors (13) to respond with an increase in intracellular calcium and to depolarize when exposed to the parasympatheticomimetic drug CCh (5, 16, 17).

When the cells in this study were exposed to CCh, the membrane potential started to oscillate (in 16 of 23 cells) between the resting value and a value close to the (48 mV). The type of oscillation varied between different cells (Fig. 1), but no systematic correlation between membrane potential, cell resistance, or cell capacitance and type of oscillation could be revealed. One reason for the difference in behavior could be that cells do not follow a uniform differentiation and that the cholinergic receptor and/or ion channels have different properties. Intracellular stores of calcium might also differ between cells.

Oscillations of membrane potential have been reported in another colonic tumor cell line, T84 (4). When exposed to CCh, T84 exhibited oscillations, although in the hyperpolarizing direction, which was caused by a calcium-induced K⁺ current.

The same authors could in a subsequent paper (3), using the perforated patch-clamp technique, show that CCh actually activated three different currents: an outward K⁺ current, an outward nonselective cation current and, an inward Cl current.

In this study, we have shown that the membrane potential oscillations in the HT-29 cells were generated by a Cl conductance (depolarizing phase) and a K⁺ conductance (repolarizing phase), as shown in Figs. 2 and 3. The reversal potential for the CCh current (Figs. 3 and 4) was close to the equilibrium potential for Cl (48 mV). When the extracellular Cl was lowered to 22 mM (equal to the pipette solution), the reversal potential changed to a value close to 0 mV (Fig. 4), in agreement with the Nernst equation for a cell with a predominant Cl conductance. Using the cell-attached mode with different pipette solutions, we could confirm the macroscopic current in membrane patches. However, results from cell-attached patches should be interpreted with caution, since the membrane potential cannot be controlled and the intracellular ionic composition is unknown. Bearing these cautions in mind, we conclude that the currents seen in these patches are indeed carried by chloride ions, since the reversal potential for the current changed ~45 mV when the pipette concentration of Cl was lowered (from 149 to 15 mM) but did not change when the pipette concentration of K⁺ was increased (from 4 to 144 mM). In most of the patches, it was difficult to see any distinct conductance levels, either due to the existence of several different channels being active at the same time, activation of low-conductance channels, rapid transition between the open and closed state, or the existence of multistate channels. Furthermore, as mentioned above, the membrane potential cannot be controlled and any change in the membrane potential will influence the current through the channels. At the resting membrane potential, one would expect a larger current, since the electrical driving force for chloride ions is greater than during the depolarized peaks when there is a very small net driving force.

In isolated patch experiments, with inside-out patches containing multiple channels, the channels showed a strong dependence on the calcium concentration with no activity at all in low calcium (50-500 nM) solutions. The conductance of the Cl channel was 19 pS (Fig. 8), and the reversal potential was 2 mV. The expected reversal potential is 6 mV. The reason for this discrepancy is not clear. One possible, but rather unlikely reason, is that the channel is not completely selective for Cl, but, on the other hand, there are no other anions present. Another possible explanation to this deviation is that the junction potential correction used (based on the Henderson equation) is not satisfactory, but at present we have no alternative method. A third explanation is the presence of a nonselective cation channel, but such a channel has, to our knowledge, not been reported in HT-29.

It cannot be excluded that several calcium-dependent Cl channels are present in this cell line. In earlier papers, a number of different Cl channels have been reported. Hayslett et al. (9) found two types of Cl channels in DBcAMP-stimulated cells. A small channel with a conductance of 15 pS was found as well as a large rectifying channel with a conductance of 50 pS at positive clamp voltages and 32 pS at negative clamp voltages. Kunzelmann et al. (15) reported the presence of a small (4 pS) Cl channel, activated by calcium, hypotonic cell swelling, and 8-BrcGMP. Morris and Frizzell (18, 19) described a 15-pS Cl channel as the basis for the whole cell-activated Cl current. They also found the presence of a large multistate channel in excised patches that was only poorly affected by calcium. The channel presented in this study is probably the same as the one described by Morris and Frizzell (18, 19).

In earlier work on HT-29 cells, it has been shown that CCh, at concentrations similar to the ones used in this study, induces a transient calcium peak followed by a plateau (16), but oscillations have, to our knowledge, not been reported before. It was concluded in the earlier work that the transient peak was caused by release from intracellular stores and the plateau level was dependent on an influx from the extracellular solution (5, 16). However, in those experiments, measurements were usually made on cell monolayers, and, in such a situation, many cells contribute to the fluorescent signal, and oscillations in an individual cell might be lost in the overall fluorescence. Calcium oscillations have been demonstrated in T84 cells by Devor et al. (2). They showed that oscillations were indeed seen in isolated cells but not in cells in a subconfluent monolayer. They suggested that this could be due to reorganization of cellular organelles influencing the spatial relationship of the sinks and sources within the cell or alterations in the availability of second messengers during monolayer formation. Apart from HT-29 and T84 cells, calcium oscillations are known to occur in other tissue types, e.g., mouse pancreatic acinar cells (22) and rat hepatocytes (24). In experiments in which we simultaneously measured membrane potential and intracellular calcium, it could be demonstrated that these two changed concomitantly, the calcium signal preceding the potential signal by a few seconds, linking the changes in intracellular calcium to membrane potential.

The functional importance of the calcium oscillations in these cells are not clear. In a study by Woods et al. (24), calcium oscillations were observed in rat hepatocytes. They suggested that the frequency of the calcium transients could determine the cellular response to the hormone that initiated the calcium signal. From an energetic point of view, calcium transients, compared with a long-lasting response, decrease the intracellular calcium load and thus decrease the need for ATP-consuming calcium pumps in the membrane of calcium-storing organelles to be activated. It is also well known that sustained increases of [Ca²⁺]_i can be damaging to the cell (21).

In summary, we have found that CCh causes oscillations of the membrane current in both the whole cell mode as well as in the cell-attached mode. With a fluorescent technique we could also show that intracellular calcium oscillated synchronously with the membrane potential. The current had two components, an inward Cl current and an outward K⁺ current, the former being the predominant one. Calcium-dependent Cl channels (19 pS), the ion channel probably responsible for the Cl current, could be seen in isolated patches.