Delayed expression of large conductance K⁺ channels reshaping agonist-induced currents in mouse pancreatic acinar cells

Abstract

Epithelial secretory cells display cell-specific mechanisms of fluid secretion and express large conductance voltage- and Ca²⁺-activated K⁺ (Maxi-K) channels that generate the membrane negativity for effective Cl⁻ exit to the lumen. Rat and mouse pancreatic acinar cells had been thought to be peculiar in this sense because of the previously reported lack of Maxi-K channels. However, this view is not entirely correct as evidenced in the present paper. Searching for their presence in pancreatic acinar cells in mice from 5 to 84 weeks of age with patch-clamp current measurements, we demonstrated that the expression of Maxi-K channels is regulated in an age-associated manner after birth. The expression started at approximately 12 postnatal weeks and increased steadily up to 84 weeks. In support of this, RT-PCR could not detect mSlo mRNA, the Maxi-K gene, at either 7 or 8 weeks but could at 58 and 64 postnatal weeks. These results suggest that a key steering element for fluid secretion, the Maxi-K channel, is progressively re-organized in rodent pancreas. A pancreatic secretagogue, acetylcholine, evoked Maxi-K channel current overlapping to various degrees on the previously known current response. This suggests that the rise in internal Ca²⁺ activates Maxi-K channels which reshape the mode of secretagogue-evoked current response and contribute to Cl⁻ driving in fluid secretion in an age-associated fashion.

Following birth and during weaning and adolescence, mammals undergo abrupt changes in their internal and external environments, through which they grow to maturation. The exocrine system is not a quiescent system during the whole life span, for its physiological manifestations or responses to external stimuli can vary at each life stage, as exemplified by rat pancreatic secretion. Developments occurring in the initial 8–9 postnatal weeks include: (1) the commencement of the agonist-induced enzyme discharge at 1–2 days (Chang & Jamieson, 1986); (2) the change in the content and composition of digestive enzymes that takes place during the weaning at 3–4 postnatal weeks (Henning, 1981); (3) the steep increase in the content of DNA and RNA that starts at 8–9 weeks and continues thereafter until the end of adolescence (Oates & Morgan, 1986). These fragmented and biased aspects nevertheless reflect the onsets of programmed adaptations of the system, and they are rooted in alterations of various functional proteins controlling exocrine secretion. The adult pancreas is, moreover, quite flexible in terms of growth in response to a pancreatic secretagogue, cholecystokinin, which promotes DNA synthesis (Mainz et al. 1973). Thus, the pancreatic secretory mechanisms are under constant re-organization during the whole life span.

One strategy for studying age-associated functional changes is to follow the expression of a particular cell membrane protein during development. Reports have so far shown age-associated changes in the expression of ion channels in muscles (Marijic et al. 2001), neuronal cells (Thibault & Landfield, 1996; Muller et al. 1998) and sensory cells (Marcotti et al. 2003) relevant to the respective channel numbers, properties and subtype distributions in the cell membrane. A similar protocol would be valid in exocrine acinar cells, namely by pursuing ion channel expression that is well defined functionally and genetically. Here, we look first at the functional expression of large-conductance K⁺ (Maxi-K) channels with a unitary conductance of approximately 200 pS, activated by membrane depolarizations and an increase in the internal Ca²⁺ concentration. The channel consists of the pore-forming α-subunit, a product of the single gene Slo, and modulator β-subunit. We examined concomitantly the content of mRNA of a gene encoding the mouse α-subunit of the Maxi-K channel (mSlo) by using reverse transcriptase polymerase chain reaction (RT-PCR). On the basis of these two approaches, we studied the functional development of the ion channel, one of the key triggering proteins for fluid secretion in exocrine acinar cells.

Maxi-K channels play a crucial role in NaCl-rich fluid secretion by creating the cell negative potentials that drive Cl⁻ ions into the lumen. This mechanism has been proposed to exist in salivary (Maruyama et al. 1983a), lacrimal gland acinar cells (Marty et al. 1984), and pancreatic acinar cells of pig (Maruyama et al. 1983b), all of which possess abundant Maxi-K channels in the cell membrane. However, species differences have been distinguished in the ion channels of pancreatic acinar cells. Maxi-K channels have so far appeared to be absent in rat and mouse pancreatic acinar cells (Randriamampita et al. 1988; Maruyama, 1989; Petersen, 1992; Schmid & Schulz, 1995). This puzzling issue remains to be resolved. Here, we address this issue in part by demonstrating an age-associated change in the expression of Maxi-K channels in mouse pancreatic acinar cells. They were hardly detected in animals younger than 12 weeks old but appeared progressively afterwards. The existence of this delay in the expression of the channels consequently requires that the currently argued mechanisms of pancreatic fluid secretion be reconsidered to take into account changes that occur during the development of the driving force for Cl⁻ exit.

Methods

Animals and cell preparation

Male closed colony ddy mice that had already been weaned, aged between 5 and 113 weeks, were supplied by Japan SLC Inc. (Shizuoka, Japan). They were fostered in a constant 12 h light−12 h dark cycle with free access to food and water. The mice were anaesthetized with ether and then killed by cervical dislocation. The protocol for animal use was approved by the Animal Care and Use Committee of Tohoku University Graduate School of Medicine. A small piece of the pancreatic tissue was digested with two types of enzymes at 37°C as previously described (Ong et al. 2001). Briefly, the tissue was treated first with collagenase (200 U ml⁻¹; Wako, Tokyo) for 7 min, minced with fine scissors, treated with trypsin (0.5 mg ml⁻¹; Sigma type XI) for 3 min and again with the same collagenase for 1 min. The resultant pancreatic cells were rinsed with enzyme-free standard external solution several times and suspended in the same solution until use.

Electrical setup and measurements

The single pancreatic acinar cells were subjected to standard patch-clamp whole-cell recording techniques in a chamber placed on the headstage of an inverted microscope (Nikon, Tokyo, Japan). Patch pipettes, made of glass capillaries with an inner diameter of 1.5 mm, had a tip resistance of 3–5 MΩ when filled with internal (pipette) solution. After the giga-seal formation, the patch membrane was ruptured by brief suction applied to the pipette. The series resistance was below 15 MΩ in every trial. Reagents were delivered by the standing hydrostatic pressure of 3 cmH₂O through an application glass pipette with a tip diameter of 30 μm and a shape that was otherwise similar to the patch pipette, brought to within 10 μm of the cell surface. The currents, amplified and arranged with an EPC-9 patch-clamp amplifier system (EPC9, HEKA Electronic, Lambrecht/Pfalz, Germany), were low-pass filtered at 1 kHz, sampled at 4 kHz, and monitored on a built-in oscilloscope and a digital thermal array-recorder (WR8500, GRAPHTEC, Kanagawa, Japan). The signals were stored simultaneously on a pulse code modulation (PCM) recorder (RD-130TE, TEAC, Tokyo, Japan) by which the single channel activities were evaluated after the experiments. All the experiments were carried out within 1 h after the cell isolation and performed at room temperature (22–24°C).

Solutions

The standard external solution in the recording chamber contained (mm): NaCl 140, KCl 4, MgCl₂ 2, CaCl₂ 1 and Hepes 5; pH adjusted to 7.3 with NaOH. The standard intracellular solution in the pipette contained (mm): KCl 145, MgCl₂ 2 and Hepes 5; pH adjusted to 7.2 with KOH. In some experiments the external NaCl was replaced by an equal amount of KCl, and the internal KCl by potassium glutamate. The pCa of the internal solution in most experiments was adjusted weakly to 6 by the addition of 1.0 mm EGTA and 0.82 mm CaCl₂. This combination was not enough to fix precisely the cellular pCa to 6 as the endogenous buffers in the cell including Ca²⁺ pumps supposedly operated strongly. The real cellular pCa would be close to pCa 7 because most of the cells in whole-cell recordings maintained an input resistance greater than 2 GΩ at −80 mV at which the Maxi-K channel activity was expected to be minimal. Thus, the use of this combination of EGTA and added Ca²⁺ was a mere convention to detect the Maxi-K channel activities, if any, under an unstimulated condition, and it allowed us to elicit internal Ca²⁺ increases with a Ca²⁺ mobilizing secretagogue, acetylcholine. In some experiments we fixed the pCa tightly to 6 with 10 mm EGTA and 8.29 mm CaCl₂ which were replaced by equal amounts of KCl. In the experiments relevant to Fig. 2, we fixed the pCa to a very low level with 2 mm EGTA and no added CaCl₂ in order to resolve clearly the single channel current steps under minimal background channel noise at a wide range of voltages with whole-cell recordings.

Figure 2. Single Maxi-K channel current monitored by whole-cell configuration in mouse pancreatic acinar cells.

A, single channel current traces from an acinar cell of a 52-week-old mouse. The membrane potential was clamped at voltages as indicated to the left of each trace. The pipette contained potassium glutamate solution with 2 mm EGTA and the bath contained the standard KCl solution (see Methods). The traces at each voltage were intended to show the current steps and therefore they did not exactly reflect the open probability. This particular cell contained roughly 10 Maxi-K channels and the open probabilities were approximately 0.05, 0.3 and 0.7 at −50, 10 and 50 mV, respectively. Each dashed and dotted line indicates one level of single channel opening and closure (single channel current amplitude), respectively. Each arrow indicates the direction of the channel openings. B, the unitary current amplitude plotted against the membrane potential in three different ionic gradients across the membrane. The pipette/bath ionic solution was KCl/NaCl, KCl/KCl and potassium glutamate/KCl where the pipette solution contained 2 mm EGTA. Each point in the diagram represents an average of five different cells. The straight lines in the graph show 194 pS in KCl/KCl and 209 pS in potassium glutamate/KCl. The curve was drawn by fitting the Goldman-Hodgkin-Katz current equation.

RNA extraction and RT-PCR analysis

We obtained acinar cell suspensions by enzymatic digestion as follows. Pancreas and submandibular glands were removed immediately after killing the animal. Pancreatic tissue was treated with collagenase (200 U ml⁻¹) for 10 min and trypsin (0.5 mg ml⁻¹) for 5 min. Submandibular gland tissue was treated with collagenase (600 U ml⁻¹) for 10 min, trypsin (0.5 mg ml⁻¹) for 10 min and again with collagenase (200 U ml⁻¹) for 3 min. Cells were washed and suspended in standard external solution and kept on ice. Total RNAs from freshly isolated pancreatic and submandibular gland cells were extracted with the QuickPrep Total RNA Extraction Kit (Amersham Pharmacia Biotech, Buckinghamshire, UK) according to the manufacturer's instructions. The first strand cDNAs were synthesized from total RNAs by Takara RNA LA PCR Kit (Takara Bio Inc., Shiga, Japan). Fragments of cDNA were amplified using PCR using the same kit as described above with Taq polymerase. The following primers were used: mSlo (sense, 5′-TTT AGG ATT TTT CAT CGC AAG TGA-3′; antisense, 5′-GTG AAA CAT TCC AGT GGA GTC GTA-3′) (Butler et al. 1993; Shipston et al. 1999) and β-actin (sense, 5′-TGT GAT GGT GGG AAT GGG TCAG-3′; antisense, 5′-TTT GAT GTC ACG CAC GAT TTCC-3′). RT-PCR products were 310 bp (mSlo) and 572 bp (β-actin). The amplification profile was 94°C for 15 s, 60°C for 30 s and 72°C for 90 s for 40 cycles (mSlo) or 30 cycles (β-actin). PCR products were separated on a 1% agarose gel and visualized by ethidium bromide staining.

Statistics

Values in the text were presented as means ± s.e.m.

Results

Detection of large-conductance K⁺ channels (Maxi-K) in mouse pancreatic acinar cells

The single acinar cells were subjected to patch-clamp whole-cell recordings. The voltage protocol, either with ramp or step-pulse voltages, concomitant with single channel current detection, discriminated outward currents in the mouse pancreatic acinar cells. Studying the acinar cells from mice at different ages between 5 and 113 weeks old, we noticed that one particular outward current, showing characteristics of the large-conductance, voltage-activated K⁺ channel (Maxi-K channel) and undetected in previous reports, was expressed in the membrane in an age-associated fashion. As described in a later section, we scarcely detected its presence in the pancreatic acinar cells younger than 12 weeks old, and the expression rate increased thereafter. Here, we depicted the features of the currents.

The ramp-voltage protocol running from −80 to 80 mV showed only a small non-specific ‘leakage current’ in all 5- to 11-week-old mouse pancreatic acinar cells (Fig. 1Aa). The step-voltage protocol also provided no significant outward current (Fig. 1Ab). Furthermore, we detected no clear single channel current steps at 0 mV in the expanded scales (Fig. 1Ac). When increasing the pipette EGTA concentration to 10 mm while maintaining the pipette pCa at 6, we observed an increase in specific background currents to various extents corresponding to 0.3–1 GΩ input resistance, presumably consisting of Ca²⁺-activated channel currents. In these experiments there was no indication of single channel current steps in the expanded scales examined at 0 mV, at which the background noise was minimal, and the current trace was very similar to that shown in Fig. 1Ac.

Figure 1. Whole-cell and single channel currents in pancreatic acinar cells of mice.

Voltage protocols used were a ramp voltage running from −80 to 80 mV during 400 ms and step-pulses of 400-ms duration applied from the holding potential of −80 mV up to 80 mV at each 20 mV step. The pipette contained standard KCl solution (pCa 6.0 with 1 mm EGTA), and the bath contained standard NaCl solution (see Methods). The current traces in Ac, Bc and Cc were recorded at 0 mV. Each broken line and arrow indicates the zero current level and the direction of the channel openings, respectively. The acinar cells were isolated from 5-, 39- and 51-week-old mice in A, B, C, respectively.

In contrast, when the ramp-voltage protocol was repeated in the acinar cells from older mice more than 12 weeks old, we frequently encountered the development of a huge current with a clear rectified property which became prominent at a voltage of around −30 mV (Fig. 1Ba and 1Ca). Considering the present ionic gradients across the membrane, with NaCl outside and KCl inside, the rectified current selected predominantly K⁺ ions. In addition, the expanded scales revealed single channel current steps, roughly 5 pA at 0 mV (Fig. 1Bc and 1Cc) in the same cells, suggesting that the currents contained large conductance K⁺ channels, possibly Maxi-K channels (Hille, 2001).

The outward currents revealed by the step-voltage protocol were divided roughly into two patterns. One pattern exhibited an initial rapid relaxation with a time constant of 10–20 ms and a sustained amplitude which remained undiminished for more than 300 ms (Fig. 1Cb) except for at the highest voltage step. It differed from a type of K⁺-selective channel current with a very slow activation time constant (approximately 1 s; Kim & Greger, 1999) in rat pancreatic acinar cell clusters, but was observable in recordings from salivary gland acinar cells (Maruyama et al. 1986), lacrimal gland acinar cells (Marty et al. 1984) and pig pancreatic acinar cells (Maruyama et al. 1983b), all of which possess abundant Maxi-K channels in their cell membranes. The other pattern showed a rapid decay after an initial peak current (Fig. 1Bb) and the decay was completed at 150 ms after the onset of the voltage step. It resembled one type of current characterized in mouse pancreatic acinar cells, the voltage-sensitive transient K⁺ current (Thorn & Petersen, 1994). A realistic interpretation of this current behaviour would require a summation of the transient K⁺ current and Maxi-K current, and the existence of the latter was supported by the single channel current steps, 5 pA at 0 mV in expanded scale (Fig. 1Bc).

Figure 2 shows the precise details of the single channel current steps of the large K⁺ current. Since the outward currents showed various magnitudes among the acinar cells, we selected the cells manifesting small whole-cell conductance and kept the cellular Ca²⁺ concentration low with 2 mm EGTA (without added Ca²⁺) in the pipette. These precautions were therefore suitable for both the detection of the single-channel steps at a wide range of voltages and the separation of other Ca²⁺-dependent current components (Fig. 2A). The traces at each voltage in Fig. 2A were intended to show exclusively the step-current amplitude and therefore they did not exactly reflect the channel open and closed probability. In our observations, this particular cell contained roughly 10 channels and the single-channel openings increased at increasingly positive voltages (open probability, approximately 0.05, 0.3, 0.7 at −50, 10, 50 mV, respectively). The detected single channel conductance was approximately 200 pS in the symmetrical potassium gradient across the membrane (Fig. 2B). Thus, the behaviour closely resembled that of the Maxi-K channel current reported in other exocrine acinar cells (Maruyama et al. 1983a, b).

Moreover, both the non-specific K⁺ channel current blocker tetraethylammonium (TEA; 5 mm) and the specific Maxi-K channel blockers iberiotoxin (IbTX; 100 nm) and paxilline (100 nm), applied to the surface membrane of the cells (18, 33 and 35 weeks old, respectively) exhibiting the outward current steps, inhibited or abolished the channel activities more or less in a reversible manner (Fig. 3). The application of TEA rapidly inhibited the channel activities, which recovered soon after the washout. Paxilline completely abolished the channel activities, which were restored during the subsequent washout period of 3–4 min. IbTX, generally used at a concentration of 100 nm to dissect the role of Maxi-K channels, inhibited most of the channel activities but complete abolition was never attained. In addition, we never observed a complete recovery during the washout period of 20–30 min. These distinctive features of the effects of IbTX on pancreatic Maxi-K channels may be explained by the difference in the β-subunit association with Maxi-K channels (Meera et al. 2000). The inhibitory effects of these reagents were similarly observed in another three cells (15, 30 and 50 weeks old) with TEA, seven cells (15, 26, 45, 48, 51, 94 and 113 weeks old) with IbTX and three cells (18, 33 and 45 weeks old) with paxilline.

Figure 3. Pharmacological properties of K⁺ channels in mouse pancreatic acinar cells.

A, the inhibitory effect of TEA (5 mm) on the outward currents evoked by a voltage ramp running from −80 mV to 80 mV for 400 ms. B and C, the effect of IbTX (B, 100 nm) and paxilline (C, 100 nm) on the single channel current in the whole-cell configuration. The membrane potential was 0 mV. Each dashed line indicates the zero current level. The pipette and the bath contained the standard KCl solution (pCa 6.0 with 1 mm EGTA) and the standard NaCl solution (see Methods), respectively.

Thus, we concluded that mouse pancreatic acinar cells contained functional Maxi-K channels in the plasma membrane. Accordingly, two primary questions arise. The first is whether the expression of the Maxi-K channels really depends on the age of the animal, and the second is whether the channel, if it exists, plays a role in the electrical response to the pancreatic secretagogues or in the underlying mechanism of fluid secretion.

Age-associated expression of functional Maxi-K channels in pancreatic acinar cells

We attempted to count functional Maxi-K channels in mouse pancreatic acinar cells of various ages. In this series of experiments, the standard NaCl solution was placed outside and the K⁺-rich solution in the pipette. The pCa of the pipette solution was weakly adjusted to 6 with 1 mm EGTA. We regarded the cells as possessing Maxi-K channels when we detected single channel current steps of roughly 5 pA at 0 mV or 15 pA at 80 mV. However, if they appeared to be absent according to this examination, we further ensured that they were truly absent by the application of 200 nm acetylcholine (ACh), as described in a later section relevant to Figs 6 and 7. When IbTX reduced the current amplitude elicited by the voltage pulses, we also counted these cases as Maxi-K channel-positive cells. The number of cells regarded as possessing Maxi-K channels as above, was plotted against the animal age as shown in Fig. 4B. For the estimate of the Maxi-K current density shown in Fig. 4C, we collected the current amplitude from the Maxi-K channel-positive cells at the end of the voltage pulse sustained at 80 mV and lasting for 300 ms, in which the leakage current was subtracted beforehand. During the pulse, the component of the transient K⁺ channel current if any had decayed nearly completely (Fig. 1Bb) so that the current at the end of the pulse consisted of the Maxi-K current.

Figure 6. ACh-induced current response in mouse pancreatic acinar cells with or without Maxi-K current.

IbTX was used to determine the presence of Maxi-K channels and their contribution to the Ach responses. The membrane potential was −3 mV. The reagents, ACh (200 nm), iberiotoxin (IbTX; 100 nm) or both were applied to the cell surface one by one during the period indicated by a bar in each trace. The patch pipette and the bath contained the standard KCl (pCa 6 with 1 mm EGTA) and NaCl solutions, respectively. A and B, consecutive ACh-induced responses with (lower trace) or without IbTX (upper trace) recorded in the same acinar cell from 11-week-old (A) and 94-week-old mice (B). Each dashed line indicates the closed channel level and each arrow the direction of channel openings. Each high resolution trace with or without IbTX was obtained during the period indicated by the letters a, b, c and d.

Figure 7. Mode of ACh-induced responses in the cells containing a variety of Maxi-K channels.

A–D, each trace was obtained from 6-, 52-, 58- and 58-week-old mice, respectively. ACh (200 nm) was applied during the period indicated by a bar in each trace. The patch pipette and the bath contained the standard KCl (pCa 6 with 1 mm EGTA) and NaCl solutions, respectively. Each high resolution trace was obtained during the period indicated by the letters a, b, c and d. Each dashed line indicates the closed channel level and each arrow the direction of channel openings.

Figure 4. Age-associated expression of Maxi-K channels in mouse pancreatic acinar cells.

A, body weight of animals used plotted against age. B, frequency of the expression of Maxi-K channels plotted against age from single acinar cells in whole-cell configuration. Cells were regarded as Maxi-K channel-positive if activities characteristic for the channel (see Text) were observed. Each point represented the mean rate of the presence of Maxi-K channels, calculated from three to 19 animals (25–94 cells) belonging to age-groups of 5–9, 10–14, 15–19, 20–29, 30–39, 51–59 and 80–84 weeks. C, K⁺ current densities of Maxi-K positive-cells, normalized to the cell input capacitance, and plotted against age. The current amplitude was obtained at 300 ms after the onset of pulse voltage sustained at 80 mV. Three exceptionally high values are shown separately.

Studying 428 cells from 77 mice of various ages, we noticed that the number of cells possessing Maxi-K channels, regardless of the current density, increased progressively with age (Fig. 4B). No cells from 5- to 11-week-old mice showed the presence of Maxi-K channels in the membrane (0 out of 97 cells). However, 96% of the cells from 80- to 84-week-old mice (24 out of 25 cells) showed their presence. There was no difference in the size between the cells with or without Maxi-K channels as estimated by the input capacitance of the cells. The values were 16.4 pF (16.4 ± 0.3, n = 110) for the Maxi-K channel-positive and 16.3 pF (16.3 ± 0.2, n = 242) for the Maxi-K channel-negative cells. Our mice gained body weight in due course with age (Fig. 4A) without any clear correlation with the channel expression.

The Maxi-K⁺ current density in Fig. 4C, normalized by the cell capacitance and plotted against the animal age, was distributed without any detectable pattern and seemed evenly scattered between 12 and 84 weeks. The values were mostly confined to below 50 pA pF⁻¹, and the mean density calculated was 15 pA pF⁻¹ (15 ± 2.4, n = 77) at 80 mV. Divided by the estimated single channel current amplitude (16 pA at 80 mV) and multiplied by the mean cell capacitance (16 pF), it yielded an average density of Maxi-K channels per cell of 15 channels in the cells expressing the channels. A single acinus, operating as a functional secretory unit, consists of 100–200 electrically coupled acinar cells (Findlay & Petersen, 1983). Multiplying both the number of electrically coupled cells (approximately 150 per acinus) and the mean channel density (15 per cell) by the expression rate (Fig. 4C), we found that the channel density in the acini increased progressively with age from 12 weeks. It seemed that the mode of increase in the channel density at the acinus level reflected an increase in the number of cells possessing the channels rather than the number of channels in the individual cells.

Detection of Slo transcripts by RT-PCR in mouse pancreatic acinar cells

mSlo encodes the pore-forming α-subunit of the Maxi-K channel in mice. To examine the expression of mSlo mRNA in mouse pancreas, we performed RT-PCR with total RNA preparations from the isolated mouse pancreatic acini. At the same time we examined samples from submandibular gland, in which Maxi-K channels are abundant in the membrane (Maruyama et al. 1983a). Figure 5 shows the mSlo mRNA expression separately for the pancreatic and submandibular cells in young (7 weeks old) and old (64 weeks old) mice. As we expected, submandibular cells expressed mSlo mRNA regardless of the animal age. In contrast, pancreatic cells of 7-week-old mice showed no detectable trace but those of 64-week-old mice showed a stable trace. The pattern was the same as that of two other samples from different animals (8-week-old for young and 58-week-old for old). The mRNA expression accorded with the presence or absence of functional Maxi-K channels examined with the current measurements, suggesting that the transcription of the mSlo gene is hardly activated before the 12th postnatal week.

Figure 5. Expression of mSlo in pancreatic acinar cells.

Total RNA preparation was extracted from young (7 weeks old) and old (64 weeks old) mice. RT-PCR experiments revealed a PCR product of 310 base pairs (bp) in the submandibular gland of both ages and in the pancreas of the old mouse.

Acetylcholine-induced responses with or without Maxi-K channels

The aim of the next experiments was to determine the mode of the Maxi-K channel contribution to the electrical responses induced by a Ca²⁺-mobilizing secretagogue, ACh. We set the membrane potential at −3 mV throughout under the ionic gradients maintained by the NaCl solution outside and the KCl solution inside. The pCa of the pipette solution was weakly adusted to 6 with 1 mm EGTA. This enabled easy and consistent detection of the Maxi-K channel activity, if any, during the stimulation. We manipulated three separate application pipettes each of which contain ACh (200 nm), IbTX (100 nm) or both, through which each reagent was subsequently applied to the same cell (Fig. 6A and B).

In the acinar cells younger than 12 weeks old, ACh (200 nm) simply induced the previously known patterns of responses (Petersen, 1992), manifesting inward current deflection(s) or a damped oscillatory inward response consisting of Ca²⁺-activated (Na⁺ and K⁺) non-selective cation channel (CAN channel) and Cl⁻ channel (CACl channel) currents, which are responsive to an increase in the internal calcium concentration ([Ca²⁺]_i). Figure 6A depicts such responses showing neither outward currents nor any marked current fluctuations before, during and after the application. Subsequent IbTX application was without effect and the second Ach application together with IbTX induced a reduced inward current qualitatively similar to the first response. The reduction in amplitude by the second stimulation was constantly observed regardless of the presence or absence of IbTX, and it has been interpreted as surface receptor desensitization (Maruyama, 1989) or a temporal depletion of Ca²⁺ in messenger-sensitive Ca²⁺ pools (Park et al. 2001). We observed such responses in another five acinar cells from 11-week-old mice.

Next we studied the mode of ACh-induced responses with or without IbTX in the cells showing the presence of the Maxi-K channel activity. In the upper left trace of Fig. 6B, we observed characteristic outward current fluctuations prior to the Ach stimulation. These consisted of discrete current steps, 5–6 pA in amplitude, shown in the expanded scales (Fig. 6Ba). The first Ach stimulation evoked an initial upright outward current followed by a damped outward response. The marked current fluctuation was observed during the stimulation (Fig. 6Bb), and it is interpreted as the superposition of the activated Maxi-K and other Ca²⁺-sensitive channels. After recovering from the stimulation, subsequent application of IbTX to the same cell reduced the channel activities (Fig. 6Bc), indicating that several Maxi-K channels were committed. The second Ach stimulation in the presence of IbTX induced a drastic change in the response, that is, it reversed the polarity of the response and evoked an inward current oscillation (Fig. 6B lower trace). Thus, the outward component of the ACh-induced response was nearly eliminated by IbTX. During the oscillation, we could still resolve the single channel current steps of roughly 5 pA (Fig. 6Bd) which would be ascribed to the remaining Maxi-K channel activity not completely eliminated by IbTX. The above features were observed in another seven cells from 94-week-old mice.

Thus, the characteristic outward fluctuations and outward deflections observed in unstimulated and stimulated conditions, respectively, were due to the activities of Maxi-K channels since they responded to IbTX. The result indicates that Maxi-K channels responded to ACh-stimulation and therefore an increase in [Ca²⁺]_i gave rise to the outward current component which counterbalanced the inward currents composed of CAN and CACl channels.

In Fig. 7 we show the various modes of the Maxi-K channel contribution to the ACh-induced responses with high resolution traces recorded before, during and after the Ach application. The trace in Fig. 7A was from 6-week-old acinar cells, which never contained Maxi-K channels, and the others were from cells over 50-weeks-old. In every record from old pancreatic acinar cells manifesting Maxi-K channels, it is clear that the channel activities increased during the ACh application, in agreement with the known sensitivity of Maxi-K channels to an increase in [Ca²⁺]_i induced by secretagogues (Maruyama & Petersen, 1984; Iwatsuki et al. 1985). The variety of modes and strengths of the outward component in these records was presumably due to the difference in the amount of Maxi-K channels expressed in the membrane as described in the previous section. Thus, the experiments have demonstrated that the expression of Maxi-K channels in an age-associated fashion adds various degrees of hyperpolarisation to the previously known modes of ACh-induced responses which emerge solely as depolarization in rodent pancreatic acinar cells.

Based on the ACh-induced current patterns in Figs 6 and 7, we roughly classified the modes of the ACh-induced responses with Maxi-K channels into four types as: (1) type-I, characterized by a dumped outward current as exemplified in the trace of Fig. 6B; (2) type-II, characterized by a initial upsurge current and subsequent inward current shown in Fig. 7D; (3) type-III, characterized by alternations of outward and inward currents shown in Fig. 7C; and (4) type-IV, characterized by a dominant inward current overlapped with some Maxi-K channel activities depicted in Fig. 7B. Type-I and type-IV were the two extremes for the expression of Maxi-K channels. The frequencies of these types counted in 37 cells between 44- and 113-weeks-old were 16/37, 6/37, 9/37 and 6/37 for type-I, type-II, type-III and type-IV, respectively. This indicates that once the maxi-K channels were expressed, type-I dominates in 40% of the ACh-induced responses. Their significance was evaluated in the next section after calculating the shift of the equilibrium potential in the ACh-induced responses.

The shift of equilibrium potentials in ACh-induced responses with Maxi-K channel expression

The aim of this section is to estimate the equilibrium potential of the ACh-induced responses with or without Maxi-K channels and to determine whether the channel activity creates the membrane negativity that results in Cl⁻exit into the lumen. We took the traces in Fig. 6B as an example for the estimation, and this pattern belonged to the type-I category. In the traces, the initial upright outward current, induced by ACh, reached 120 pA, on which the three types of channel currents, Maxi-K (I_K), CAN (I_CAN) and CACl current (I_CACl), were superimposed (Fig. 6B, upper trace). We disregarded the contribution of other components responsible for the resting membrane potential as they were negligibly small. However, the inward current component in the presence of IbTX reached 25 pA, and contained the latter two channels (Fig. 6B, lower trace). Thus, the amplitude of Maxi-K channel currents would be roughly 145 pA. A previous report and our observation determined a ratio of the respective I_CAN and I_CACl in the inward current as 1 : 8 (Park et al. 2001) or 1 : 4 (authors' unpublished observation) in mouse pancreatic acinar cells without Maxi-K channels. Here, we took the raio 1 : 8 in our calculation.

The ACh-induced peak current at −3 mV (Fig. 6B) is composed of the following components:

where R, E and V represent the individual channel chord resistance, equilibrium potential and the holding potential, respectively. The integrated equilibrium potential (E_e) of these pathways in parallel (Fig. 8B) is expressed as:

Without Maxi-K channels (Fig. 8A) it is:

By inserting the known values I_K = 145 pA (roughly 29 Maxi-K channels), I_CAN = 2.8 pA, I_CACl = 22 pA, V =−3 mV, E_K =−90 mV, E_CAN = 0 mV and E_CACl = 0 mV, we calculated the chord resistances as R_K = 0.6 GΩ, R_CAN = 1.1 GΩ, and R_CACl = 0.14 GΩ. Using these values, we determined the equilibrium potential, E_e, in ACh-induced responses containing the Maxi-K channel activities as −15.4 mV, while E_e of the responses composed of CAN and CACl channel pathways was 0 mV. Thus, the incorporation of Maxi-K channels can shift E_e by −15.4 mV in the peak responses as shown in Fig. 6B.

Figure 8. Diagrams for the calculation of equilibrium potentials induced by ACh with or without Maxi-K channels.

R_CAN, R_CACl and R_K represent the electrical resistance for Ca²⁺-dependent non-selective cation channels, Cl⁻ channels and Maxi-K channels, respectively. Likewise E_CAN, E_CACl and E_K represent the respective equilibrium potentials.

We re-calculated each E_e value by replacing the E_Cl value with a more realistic one, −25 mV, reported in pancreatic acini with the impalement of ion-selective electrodes (Petersen et al. 1981). The calculation showed that E_e is −33.3 mV and −22.4 mV with or without Maxi-K channels, respectively. Comparing these values with the Cl⁻ equilibrium potential (−25 mV), it is possible that Maxi-K channels can create a voltage difference, E_e–E_Cl, of −8.3 mV. When setting the ratio between I_CAN and I_CACl as 1 : 4, the value of E_e–E_Cl similarly calculated was −6.5 mV. When using the value of I_K as one-half of the previous one, 70 pA (roughly corresponding to 14 Maxi-K channels), the value of E_e–E_Cl was −3.2 mV for a I_CAN : I_CACl ratio of 1 : 8 and −0.8 mV for one of 1 : 4.

Except in the experiments using IbTX (Fig. 6B), the real contributions of Maxi-K and other channels to the ACh-induced responses contain some uncertainty because these are inseparable in the total response. Nonetheless, we can obtain a rough estimate of their contributions by assuming that the outward or inward current responses reflect Maxi-K channel currents or a mixture of CAN and CACl channel currents, respectively. The outward component in type-I, measured at the peak outward current of −3 mV, was 111 pA (111 ± 27, n = 16), and there was no inward current component. This type would be evaluated by the calculation in the foregoing paragraphs. In type-II, type-III and type-IV, the current peaks at the same voltage were 93 (93 ± 8, n = 6), 61 (61 ± 18, n = 9) and 14 pA (14 ± 4, n = 6) for the outward component, and 32 (32 ± 7, n = 6), 31(31 ± 6, n = 9) and 59 pA (59 ± 19, n = 6) for the inward component, respectively. The values converted to the chord resistance (or conductance) were 1.0 (1.0), 1.5 (0.68) and 6.7 GΩ (0.16 nS) for the outward component, and 0.1 (10), 0.1 (10) and 0.05 GΩ (20 nS) for the inward component of type-II, type-III and type-IV, respectively. Repeating the same calculation with the ratio of I_CAN: I_CACl set to 1 : 8, we obtained E_e–E_Cl (i.e. E_e+ 25 mV) of −2.3, −1.2 and +2.5 mV for the type-II, type-III and type-IV responses, respectively. Thus, E_e of type-II to type-IV responses would fluctuate critically around E_Cl and be neutral for creating the driving force of Cl⁻ exit.

Although some assumptions and values used here may deviate from other experimental conditions and not precisely reflect the physiological situation, it nevertheless seems reasonable to conclude that Maxi-K channels in type-I response can shift the equilibrium potential to below the Cl⁻ equilibrium potential by several millivolts at most, which can drive the Cl⁻ exit gently to the lumen during Ca²⁺-mobilizing stimulation.

The ACh-induced peak conductance increases collected from the cells without Maxi-K channels (we refer to them as type-0) were 25 nS on average (25 ± 6 nS, n = 12, animals between 6 and 80 weeks old). The calculated E_e of type-0 was −22.4 mV, again assuming that the ratio of CAN between CACl channels was 1 : 8 and E_Cl was −25 mV. Assuming here that the putative acinus consisted only of cells with type-0 and type-I responses, we calculated the ratio of their conductance at which total acinar E_e equals E_Cl. The ratio provides the number of cells with type-0 and type-I responses in the acinus. Using an E_e of −22.4 mV and conductance of 25 nS from type-0 responses, and E_e of −33.3 mV from the estimate of the type-I response relevant to Fig. 6B, we calculated a type-I conductance as 7.9 nS set the total acinar E_e equal to E_Cl of −25 mV. The two conductances, 25 and 7.9 nS, were proportional to the number of cells with type-0 and type-I responses, respectively. Thus, an acinus possessing 24% of cells with type-I and 76% with type-0 responses set the acinar E_e to E_Cl. If the acinus involves over 24% of type-I representing cells among the rest of the type-0 cells, a gentle Cl⁻ exit may well have occurred. When setting E_e of the type-I cells at −28 mV (in case of 14 Maxi-K channels per cell), the critical percentage of type-I cells increased to 53%. The animal age indicating 53% expression of Maxi-K channels was 35 weeks as shown in Fig. 4B. Thus, it is likely that Maxi-K channels begin to play a substantial role in fluid secretion sometime beyond this age.

Discussion

ACh-induced current response with or without Maxi-K channels and mouse pancreatic fluid secretion

The activation of Maxi-K channels, triggered by secretagogue-induced increases in [Ca²⁺]_i, has been demonstrated in most exocrine gland acinar cells except rat and mouse pancreatic acinar cells. Such activation provides a substantial electrical driving force for Cl⁻ exit into the lumen in fluid secretion. Mouse and rat pancreatic acinar cells would be peculiar in this sense because of their presumed lack of Maxi-K channels, as reported implicitly and explicitly in many previous papers. However, this belief is not valid in mice older than 12 weeks. In mouse pancreatic acinar cells, Maxi-K channels progressively develop with age. The major reason that previous reports failed to describe the presence of such channels is that they were concerned only with young pancreatic acinar cells before 12 weeks. The activation of Maxi-K channels, overlapping the period of luminal CACl-channels following agonist stimulation, is critical for creating the driving force of Cl⁻ exit. As expected from the calculation relevant to Fig. 6B, the incorporation of Maxi-K channels can generate extra negative voltage by several millivolts at most below the reported Cl⁻ equilibrium potential. The type-I response fits with this notion but type-II, type-III and type-IV do not. When we consider the Maxi-K channel expression to be significant for Cl⁻ exit, especially at the acinus level, we obviously need several assumptions as in the case of the last section of Results. It would be of interest to know at which stage the Maxi-K channel activities exceed those of CAN channels. The real significance of the Maxi-K channel expression remains for a future study to address whether the number of CAN channels change with ageing.

We observed considerable heterogeneity in the expression of Maxi-K channels among the aged pancreatic acinar cells in Fig. 4B, in which the density of the channel current did not much change with ageing (Fig. 4C). When taking this into account and considering a group of cells, an acinus, which consists of 100–200 electrically coupled acinar cells (Findlay & Petersen, 1983), the Maxi-K channel activity never reached zero after 12 weeks old, and the acinar density of Maxi-K channels increased progressively with age. It would be interesting to study in this context whether the cell coupling or gap-junction status influences the fluid secretion through the participation of Maxi-K-possessing acinar cells. The density of Maxi-K channels was scattered considerably at the same age or even in the same batch of acinar cells (Fig. 4C). This variability may be due to the individual cell status or the cell cycle. The individual cells could be developed and expressing Maxi-K channels in a way that enables an acinus or functional unit to maintain a density of channels appropriate to the need for fluid secretion at the time.

At present the so called ‘push–pull model’ of rodent pancreatic fluid secretion (Kasai & Augustine, 1990), which works without agonist-regulated K⁺ channel current, is a matter of debate. The lack of Ca²⁺-activated Cl⁻ channels in the basolateral membrane (Zdebik et al. 1997; Park et al. 2001) does not support the otherwise compelling push–pull model. We have taken into account this point when considering the mechanism of fluid secretion in young pancreatic acinar cells lacking Maxi-K channels. These pancreatic acinar cells contain CAN and CACl channels, and probably transient K⁺ channels (Thorn & Petersen, 1994, no information for animal age), inward rectifying K⁺ channels (Schmid et al. 1997, rat pancreatic acinar cells in culture; Kim et al. 2000; pancreatic acinar cells from 100- to 150-g rats), and slowly activating K⁺ channels (Kim & Greger, 1999, pancreatic anini from 100- to 150-g rats) encoded by KVLQT1 gene (Kottgen et al. 1999). These three reported K⁺ channels can in principle serve for the Cl⁻ exit as the driving force. In the present series of experiments, we deliberately concentrated on the presence or absence of the large K⁺ current but did not carry out systematic experiments particularly for these channel currents in view of the ageing of the pancreas. Nevertheless, we encountered transient K⁺ currents with a variety of amplitudes several times in both young (5, 6, 7, 9 and 10 week old) and old mouse (17, 25, 38, 59 and 80 week old) pancreatic acinar cells. It is likely that the transient K⁺ channel current exists regardless of the pancreatic age but not in all acinar cells. Accordingly, the slowly activating K⁺ channels, which could respond to pancreatic agonists ACh and secretin (Kim et al. 2001; Lee et al. 2002), would be the most probable candidate in young mouse acinar cells for creating the driving force during the stimulation. However, we do not have a very clear image of the mechanism of fluid secretion in young rodent pancreatic acinar cells compared to any other exocrine acinar cells containing Maxi-K channels.

Development of Maxi-K channel expression in pancreatic acinar cells

Before 12 weeks, we scarcely detected the Maxi-K currents in the pancreatic cells, supported by the lack of mSlo mRNA in 7- and 8-week-old mice. This suggests that there is no Maxi-K channel protein either in the cell membrane or even the organelle membrane before 7 weeks. In contrast, the RT-PCR revealed the presence of Maxi-K channel mRNA in the animals at 58 and 64 weeks, consistent with the presence of Maxi-K channel currents at these ages. Together, we conclude that the critical period for Maxi-K expression is at 11–12 weeks, from which time the activation of the gene expression started. In contrast, submandibular gland tissue contained mSlo mRNA as early as at 7 weeks and also functional channels in the cell membrane (data not shown). The reason for this onset difference of mSlo transcription between these acinar cells is unclear. Yet unknown mechanisms promoting a tissue- and stage-specific expression of mSlo, like the dSlo (Drosophila slowpoke) transcriptional control mechanism (Brenner & Atkinson, 1996; Brenner et al. 1996), may determine the difference in these two major mammalian exocrine glands.

Only 20% of the cells possessed Maxi-K channels at 20 weeks (Fig. 4B), though the frequency of the channel expression increased steadily with ageing during the entire mouse life span. We know that renewal of the adult pancreatic acinar cells takes place every 20 weeks (half-life of 70 days, Magami et al. 2002). Applying this view straightforwardly to the process of channel expression, the expression should rapidly decrease during every cell renewal so that we would never have an increasing curve like the one in Fig. 4B. Unknown intracellular or extracellular factors, effectively and selectively facilitating the channel expression, should operate during the life cycle of rodent pancreatic acinar cells.

Various phenotypes of Maxi-K channels have been reported. Such diversity is thought to result from alternative splicing and/or interaction with regulatory subunits (β-subunits) which critically determine the Ca²⁺ sensitivity and macroscopic channel kinetics (Orio et al. 2002). It would be intriguing to sort out the interactive proteins in detail in order to elucidate the fine regulation of fluid secretion through the Maxi-K channel expression. The machinery of rodent pancreatic acinar cells represents a remarkably dynamic system, especially in view of the Maxi-K channel expression and mechanism of fluid secretion. This would be reflected by the various modes of ACh-induced responses (Fig. 7). Further, it would be of interest to determine whether other elements, involving Na⁺–K⁺–Cl⁻ cotransporter, CAN and CACl channels, Ca²⁺ signalling machineries, and even the gap-junction status also undergo age-associated changes or re-organization in terms of exocrine secretion.