Role of the BK channel (K_Ca1.1) during activation of electrogenic K⁺ secretion in guinea pig distal colon

Abstract

Secretagogues acting at a variety of receptor types activate electrogenic K⁺ secretion in guinea pig distal colon, often accompanied by Cl⁻ secretion. Distinct blockers of K_Ca1.1 (BK, Kcnma1), iberiotoxin (IbTx), and paxilline inhibited the negative short-circuit current (I_sc) associated with K⁺ secretion. Mucosal addition of IbTx inhibited epinephrine-activated I_sc (^epiI_sc) and transepithelial conductance (^epiG_t) consistent with K⁺ secretion occurring via apical membrane K_Ca1.1. The concentration dependence of IbTx inhibition of ^epiI_sc yielded an IC₅₀ of 193 nM, with a maximal inhibition of 51%. Similarly, IbTx inhibited ^epiG_t with an IC₅₀ of 220 nM and maximal inhibition of 48%. Mucosally added paxilline (10 μM) inhibited ^epiI_sc and ^epiG_t by ∼50%. IbTx and paxilline also inhibited I_sc activated by mucosal ATP, supporting apical K_Ca1.1 as a requirement for this K⁺ secretagogue. Responses to IbTx and paxilline indicated that a component of K⁺ secretion occurred during activation of Cl⁻ secretion by prostaglandin-E₂ and cholinergic stimulation. Analysis of K_Ca1.1α mRNA expression in distal colonic epithelial cells indicated the presence of the ZERO splice variant and three splice variants for the COOH terminus. The presence of the regulatory β-subunits K_Caβ1 and K_Caβ4 also was demonstrated. Immunolocalization supported the presence of K_Ca1.1α in apical and basolateral membranes of surface and crypt cells. Together these results support a cellular mechanism for electrogenic K⁺ secretion involving apical membrane K_Ca1.1 during activation by several secretagogue types, but the observed K⁺ secretion likely required the activity of additional K⁺ channel types in the apical membrane.

electrogenic K⁺ secretion by colonic epithelial cells contributes to the high K⁺ concentration in the lumen (30, 82). Various secretagogues stimulate the rate of this K⁺ secretion while Na⁺ absorption diminishes the luminal Na⁺ content (2, 64). In this way the colonic epithelium alters the fluid arriving from the small intestine into a composition with high K⁺ concentration and low Na⁺ concentration. Those epithelial cells responsible for producing this K⁺ secretion use the active Cl⁻ secretory mechanism (16, 17) with the inclusion of apical membrane K⁺ channels (21). The extent of secretagogue activation determines the final colonic fluid and electrolyte balance and emanates from several sources including the enteric and sympathetic nervous systems as well as paracrine systems in the mucosa (8, 58, 87).

Excessive rates of colonic fluid secretion often lead to hypokalemia due to the attendant K⁺ secretion. Massive losses occur during invasion by virulent bacteria as with cholera (12, 63, 78) or with some types of villous adenomas (11, 46), but subtler imbalances over time also can lead to severe hypokalemia. Central to the regulation of luminal composition is the range of relative K⁺ and Cl⁻ secretory rates possible with various secretagogues (64, 87). The profound enteric nervous system discharge created during cholera (44) produces the ultimate synergistic fluid secretion with Cl⁻ secretion predominant over K⁺ secretion during this high fluid flow. In comparison, the sympathetic/parasympathetic imbalance occurring with acute colonic pseudo-obstruction results in luminal K⁺ concentrations near 150 mM (69, 80), likely due to the β-adrenergic activation of K⁺ secretion without active Cl⁻ secretion (64, 86). In contrast to the catastrophic K⁺ losses exhibited in these pathological regulatory conditions, a well-managed K⁺ loss by stimulating increased colonic K⁺ secretion might aid patients during end-stage renal disease (ESRD) such that total body K⁺ balance could be maintained (27, 48, 67).

Transepithelial fluid flows occur as the result of osmotic gradients developed by net solute absorption or secretion with the Na⁺/K⁺-ATPase in the basolateral membrane generally setting the electrochemical gradients (2, 30). Simply opening apical membrane K⁺ channels allows these fluid-transporting cells to also secrete K⁺ by a mechanism involving pump uptake and channel exit. Regulatory control of these apical K⁺ channels provides the means to match K⁺ secretory rates to any physiological process, such that the range of K⁺ channel proteins available allows each cell type to use those channels that fit best into the cellular signaling initiating fluid flow.

Although numerous studies support the general cellular mechanism for electrogenic K⁺ secretion (30), the molecular identity of the apical membrane K⁺ channels required during secretagogue activation is not definitively established, with both K_Ca1.1 (BK, Kcnma1) and K_Ca3.1 (Kcnn4) implicated in apical K⁺ exit during colonic K⁺ secretion (34, 49, 72, 74). In particular, distal colon from K_Ca1.1-null mice exhibits attenuated electrogenic K⁺ secretion (70, 75, 76). For these K_Ca1.1-null mice, the P2Y₂/P2Y₄-receptor-mediated K⁺ secretory response is eliminated (70) and responses to both β-adrenergic and aldosterone stimulation are substantially reduced (75, 76). Together these results support K_Ca1.1 as an important component of the apical membrane K⁺ conductance responsible for K⁺ exit into the colonic lumen during K⁺ secretion.

Activation of K_Ca1.1 depends on both intracellular Ca²⁺ and membrane voltage, and also can be altered via phosphorylation including that by protein kinase A (PKA) (3, 10, 32, 81). Specifically, membrane depolarization opens K_Ca1.1, with Ca²⁺ increases shifting this voltage sensitivity to more negative membrane voltages. Splicing of K_Ca1.1 proteins to include the alternative exon STREX converts the PKA enhancement of activation into inhibition (5, 79). Combining Ca²⁺ and PKA sensitivity likely provides the means for opening K_Ca1.1 during secretagogue activation. However, the typical voltage sensitivity is such that cytosolic Ca²⁺ would have to reach values near 10 μM for K_Ca1.1 to activate given the apical membrane voltage present in colonic epithelial cells (42). Since K_Ca1.1 forms channel complexes with auxillary subunits K_Caβ (Kcnmb), the ability of K_Caβ to alter activation kinetics (3, 9, 40, 60) may be responsible in part for producing the channel activity necessary to support K⁺ secretion. The experiments in this study examine involvement of K_Ca1.1 using specific inhibitors during stimulation with physiological secretagogues, as well as K_Ca1.1 localization within the colonic epithelium.

METHODS

Male guinea pigs (500–700 g body wt, Hartley strain, Hilltop Lab Animals, Scottdale, PA) received standard chow and water ad libitum and were housed on site at least 2 wk before experiments. Guinea pigs were euthanized with an animal decapitator (Harvard Apparatus, Holliston, MA) in accordance with a protocol approved by the Wright State University Laboratory Animal Care and Use Committee. Colonic mucosa was isolated as described previously (86). These isolated colonic mucosal sheets were used for measurement of transepithelial electrical parameters, mRNA detection by RT-PCR, protein detection by immunoblot, and confocal immunolocalization.

Transepithelial current measurement.

Isolated mucosal sheets were used for measurement of transepithelial current and conductance as described previously (39). Briefly, mucosal sheets were mounted in Ussing chambers (0.64-cm² aperture), supported on the serosal face by nuclepore filters (∼10-μm thick, 5-μm pore diameter; Whatman, Clifton, NJ). Bathing solutions (10 ml) were circulated by gas-lift through water-jacketed reservoirs (38°C). Standard Ringer's solution contained (in mM) 145 Na⁺, 5.0 K⁺, 1.4 Ca²⁺, 1.2 Mg²⁺, 125 Cl⁻, 25 HCO₃⁻, 4.0 H_(3-X)PO₄^X−, and 10 d-glucose. Solutions were continually gassed with 95% O₂-5% CO₂, which maintained solution pH at 7.4. Automatic voltage clamps (Physiologic Instruments, San Diego, CA) permitted measurement of short-circuit current (I_sc) and calculation of transepithelial conductance (G_t) from current responses to voltage pulses imposed across the mucosa ( ± 5 mV, 3-s duration, 60-s intervals). I_sc was referred to as positive for cation flow across the epithelium from mucosal to serosal side (K⁺ secretion, negative; Cl⁻ secretion, positive).

Mucosal responses to physiological secretagogues and to inhibitors were examined after producing a quiescent basal condition by suppressing neural and paracrine activators persisting in the isolated colonic mucosa (87). Removal of all muscle layers limited influences from nerves in this mucosal preparation. Compounds released from the mucosa into the bathing solutions were reduced in concentration (∼8,000-fold) by replacing the solutions three times after mounting the mucosa. Prostanoid production within the isolated mucosa was suppressed by the COx-1 inhibitor SC560 (1 μM) and the COx-2 inhibitor CAY10404 (1 μM) added to both bathing solutions. The action of peptide-YY and neuropeptide-Y released from mucosal cells was inhibited using the neuropeptide-receptor-Y2 antagonist BIIE0246 (1 μM) added to the serosal bath. Amiloride (10 μM) added to the mucosal bath inhibited electrogenic Na⁺ absorption. Sequential addition of the secretagogues epinephrine, prostaglandin E₂ (PGE₂), and carbachol (CCh) allowed examination of the full range in secretory responses from modulatory to flushing to synergistic. Prior addition of epinephrine does not alter subsequent PGE₂ responses (23, 24, 39), and combined addition of PGE₂ and CCh produces the synergistic mode of secretion (87).

Drugs were added in small volumes from concentrated stock solutions. CAY10404, prostaglandin E₂, and SC560 were obtained from Cayman Chemical (Ann Arbor, MI); charybdotoxin (ChTx) from Abcam Biochemicals (Cambridge, MA); epinephrine from Hospira (Lake Forest, IL); and BIIE0246, iberiotoxin (IbTx), ICI-118551, and paxilline from Tocris Bioscience (Ellisville, MO). All other chemicals were obtained from Sigma Chemical (St. Louis, MO).

Detection of mRNA and proteins.

Total RNA was extracted by RNeasy-Mini-Kit (Qiagen, Valencia, CA) from isolated mucosa and EDTA-released epithelial cells (38). Individual crypts were dissected under 40× magnification with Dumont no. 5 forceps (22). For isolated crypts, a guanidinium thiocyanate-phenol-chloroform extraction (TRIzol) method was used to obtain total RNA. Briefly (86), after reverse transcription of mRNA, cDNA was amplified by PCR. Primers specific for K_Ca1.1α (Kcnma1) and K_Caβ (Kcnmb) (Table 1) were based on previous design (33, 52) and alignment of nucleotide sequences for human, mouse, and rat: initial denaturing 95°C (10 min), 40 cycles denaturation 94°C (1 min), annealing 55°C (1 min), extension 72°C (1 min), final extension 72°C (7 min). Primers for Kcnma1 amplified the exon 16–20 segment, which included ZERO/STREX splicing, as well as the alternative COOH-termini from the exon 26–27 segment and the exon 26–28 segment. Since multiple exons were transcribed for Kcnma1 and for Kcnmb, contamination with genomic DNA would have resulted in much larger products than those predicted and obtained. Splicing variants were verified by inspection of sequencing chromatograms. The two primer sets used for Kcnmb3 (61, 68) were forward 5′-ggt-ttg-cca-tga-tgg-gct-tct-c-3′ with reverse 5′-aca-gac-atc-tga-agg-cca-gca-c-3′ and forward 5′-cat-cgc-cat-gat-ggc-ctc-ct-3′ with reverse 5′-tca-gag-cgc-ctc-cca-gca-at-3′, but neither generated a product with homology to Kcnmb3.

Table 1.

Primers for K_Ca1.1α and K_Caβ

		Forward	Reverse
KCa1.1α Kcnma1	ex16–ex20 (STREX)	5′-agc-cat-tga-gta-caa-gtc-tg-3′	5′-gga-gtc-cat-gtt-gtc-aat-ct-3′
	ex26–ex27	5′-tgg-tga-tct-gtt-ctg-caa-agc-3′	B 5′-agc-cgc-tct-tcc-tgc-acg-tac-ttc-3′
	ex26–ex28 ↓	↓	E 5′-ctg-ctt-gtg-gat-tta-tat-tgg-3′
			F 5′-tcc-tgg-gag-tca-aca-ttc-atc-3′
			G 5′-tag-ttc-tgg-tct-cct-ggg-agt-3′
			H 5′-tcc-acc-agt-gat-tca-gca-tgt-3′
KCaβ1 Kcnmb1		5′-tgt-gct-gtc-atc-acc-tac-t-3′	5′-cat-ggc-aat-aat-gag-gag-3′
KCaβ2 Kcnmb2		5′-gac-tgg-cta-tga-tgg-tgt-g-3′	5′-ggt-cag-aat-agc-agg-aga-ag-3′
KCaβ4 Kcnmb4		5′-cgt-gtc-gct-ctt-cat-ctt-3′	5′-caa-tgc-agg-agg-aca-atc-3′

Reverse primers in exon 27 (B) and exon 28 (E, F, G, H) were used with the forward primer positioned in exon 26. (numbering based on constitutive exons for K_Ca1.1α, 15).

Proteins were isolated from colonic epithelial cells (86). Briefly, after disruption by sonication in a buffered solution containing protease inhibitors, samples were centrifuged to obtain a membrane sample. Following SDS-PAGE and transfer to polyvinylidene difluoride membranes, incubation with K_Ca1.1 (Kcnma1) specific primary antibody (1:1,000, NeuroMab clone L6/60 of mouse slo1, 75–022; UC Davis/NIH NeuroMab Facility, Davis, CA), and then with horseradish peroxidase-conjugated secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, PA) allowed detection of protein.

Tissue fixation and immunolocalizaton.

Colonic tissues were fixed after isolation, as described previously (24). Briefly, isolated mucosal sheets were chemically fixed followed by dehydration, sectioning, and mounting on gelatin-coated slides. Sections were permeabilized, blocked, and then incubated for 48 h (4°C) with primary antibody. Antibodies for K_Ca1.1 (Kcnma1) were obtained from UC Davis/NIH NeuroMab Facility (Davis, CA) [NeuroMab clone L6/60 (75–022), COOH terminus of mouse slo1; 1:50] and Santa Cruz Biotechnology (Santa Cruz, CA) [polyclonal rabbit-anti-MaxiKα (H300), COOH-terminal residues 937–1236 of human Kcnma1; 1:50]. Antiserum for Kcnma1 (53) was generously provided by A. J. Hudspeth at the Howard Hughes Medical Institute and Laboratory of Sensory Neuroscience, The Rockefeller University, New York, NY (1:200; rabbit antiserum-K_Ca1.1, residues KYVQEDRL). Secondary antibodies to detect immunoreactivity (2 h, room temp) were obtained from Invitrogen (Carlsbad, CA): donkey-anti-mouse IgG antibody, conjugated to AlexaFluor488 (4 ng/μl) and donkey-anti-rabbit IgG antibody, conjugated to AlexaFluor488 (4 ng/μl). Labeling of actin was obtained with phalloidin conjugated to AlexaFluor568 (0.005 units/μl). Sections were washed and mounted in Vectashield (Vector Labs, Burlingame, CA). Fluorescence of double-labeled sections was visualized with an Olympus FluoView FV300 confocal microscope in the Microscopy Core Facility of the WSU and PHP Neuroscience Institute. Images were acquired using identical confocal aperture, background, and gain settings.

Data analysis.

Responses of I_sc and G_t to secretagogues and inhibitors were obtained from adjacent mucosae in each colon to permit direct comparisons. Recordings of I_sc were digitized at 10-s intervals to examine the secretory time course. Concentration responses of I_sc and G_t were fit by Henri-Michaelis-Menten binding curves using a nonlinear least-squares procedure. Results were reported as means and standard error of the mean (SE) with the number of animals (n) indicated. Statistical comparisons were made using two-tailed Student's t-test for paired responses, with significant difference accepted at P < 0.05.

RESULTS

The guinea pig distal colon produces high rates of electrogenic K⁺ secretion in response to many physiological neurotransmitters, hormones, and paracrine factors (23, 64, 87). Involvement of K_Ca1.1 during these agonist responses was examined by sensitivity to selective inhibitors.

K_Ca1.1 blockers inhibit β-adrenergic K⁺ secretion.

Basal I_sc was negative, consistent with a low rate of ongoing K⁺ secretion in the ex vivo distal colonic mucosa (23). Addition of iberiotoxin (IbTx), an α-K channel toxin peptide blocker of K_Ca1.1 (18, 56), to the mucosal solution during this basal condition significantly reduced I_sc toward zero (Fig. 1A, Table 2) and decreased G_t by 0.60 ± 0.15 mS/cm² (n = 5, P = 0.017), supporting a contribution of K_Ca1.1 to apical membrane conductance during basal K⁺ secretion. Paxilline, an alkaloid blocker of K_Ca1.1 (56, 66, 73), also reduced basal I_sc toward zero (Table 2) but with a slower time course.

Fig. 1.

Sensitivity of epinephrine activated secretion to K_Ca1.1 blockers. Isolated mucosae were stimulated by epinephrine addition to the serosal solution from the standard basal condition with electrogenic secretion monitored by short-circuit current (I_sc) (see methods). A: I_sc was measured in adjacent mucosae during basal conditions with addition of IbTx (300 nM) to the mucosal solution of one (dotted line). B: adjacent mucosae were stimulated by epinephrine (5 μM) with one pretreated by mucosally added IbTx (200 nM) 60 min prior to epinephrine (dotted line). C: a third mucosa adjacent to those shown in panel B was treated with three successive concentrations of IbTx (10, 100, and 200 nM) beginning after epinephrine stimulation (dotted line), control (solid line), and IbTx pretreatment (dashed line). D and E: responses of epinephrine-activated I_sc (^epiI_sc) and transepithelial conductance (^epiΔG_t) to 3 concentrations of IbTx were measured with an adjacent mucosa as time control. Control ^epiI_sc and ^epiΔG_t were −125.8 ± 15.2 μA/cm² and +6.13 ± 1.34 mS/cm² (○, n = 3), respectively. Fractional inhibition by IbTx was corrected for any time-dependent changes, averaged, and scaled to the control ^epiI_sc or ^epiΔG_t (●). Values significantly different from control (P < 0.05) are indicated by an asterisk (*). A fit of Henri-Michaelis-Menten kinetics with a single binding site was made to the concentration dependence from each mucosa and the average (n = 3) shown as a dashed line: ^epiI_sc, ^IbTxIC₅₀ = 193 ± 41 nM, maximal fractional inhibition = 0.513 ± 0.121; ^epiΔG_t, ^IbTxIC₅₀ = 220 ± 77 nM, maximal fractional inhibition = 0.476 ± 0.040. An additional fit was made assuming that IbTx could inhibit ^epiI_sc or ^epiΔG_t completely (dotted line). F: adjacent mucosae were stimulated by epinephrine (1.0 μM) with one pretreated by paxilline (1.0 μM) addition to both mucosal and serosal solutions 35 min prior to epinephrine (dotted line).

Table 2.

Action of K_Ca1.1 channel inhibitors on K⁺ secretion

	IbTx			Paxilline
	IbTxΔΔIsc, μA/cm2	IbTxΔΔIsc/secΔIsc	n	paxΔΔIsc, μA/cm2	paxΔΔIsc/secΔIsc	n
Basal	+9.9 ± 0.6* (0.0001)	−0.258 ± 0.033* (0.001)	5	+4.1 ± 1.2* (0.017)	−0.145 ± 0.059* (0.043)	8
Epinephrine	+20.8 ± 5.1* (0.027)	−0.281 ± 0.047* (0.010)	5	+16.4 ± 5.0* (0.013)	−0.216 ± 0.054* (0.005)	8
ATP(transient)	+11.6 ± 2.6* (0.047)	−0.654 ± 0.209 (0.089)	3	+9.6 ± 1.4* (0.002)	−0.454 ± 0.038* (0.0003)	5
ATP(sustained)	+3.1 ± 0.9 (0.074)	−0.293 ± 0.069* (0.052)	3	+3.9 ± 1.9 (0.104)	−0.317 ± 0.102* (0.036)	5

Values are means ± SE (P values). Secretagogue-activated short-circuit current (^secΔI_sc = secretagogue − control) was measured for epinephrine (Fig. 1) and ATP (Fig. 3), and inhibitor-induced differences in ^secΔI_sc (^inhibΔΔI_sc = inhibitor − control) were compared as well as the fractional inhibition (^inhibΔΔI_sc/^secΔI_sc). Inhibitors were added prior to secretagogue activation: paxilline (mucosal and serosal) 1.0 μM; IbTx (mucosal) 300 nM for basal and epinephrine, 200 nM for ATP. Significant differences from zero are indicated by asterisks (P values).

Epinephrine activates a transient Cl⁻ secretory response followed by sustained K⁺ secretion (^epiI_sc, 25). Mucosally added IbTx inhibited the negative I_sc resulting from epinephrine activation of K⁺ secretion (Fig. 1B, Table 2) without altering the Cl⁻ secretory I_sc transient. Inhibition of the epinephrine-activated G_t (^epiΔG_t) by 1.21 ± 0.30 mS/cm² (n = 5, P = 0.015) during IbTx treatment further supported a contribution by apical membrane K_Ca1.1 to epinephrine-activated electrogenic K⁺ secretion.

Sensitivity of ^epiI_sc to IbTx was quantified further by consecutive addition of increasing concentration to the mucosal solution during ongoing epinephrine stimulation (Fig. 1C). IbTx inhibited ∼50% of ^epiI_sc with an IC₅₀ of 193 nM (Fig. 1D). Similarly, IbTx inhibited ∼50% of ^epiΔG_t with an IC₅₀ of 220 nM (Fig. 1E). This low sensitivity of electrogenic K⁺ secretion to IbTx supported involvement of β-subunits, K_Caβ (Kcnmb), which would interfere with toxin binding (40, 51, 60). Use of higher IbTx concentrations to further substantiate the proportion of block would be confounded by increasing nonspecificity, because IbTx blocks other K⁺ channels at concentrations above 300 nM, particularly K_V1.3, which is present in colonocytes (18, 20).

Paxilline (1.0 μM) also inhibited ^epiI_sc (Fig. 1F, Table 2) and ^epiΔG_t (−1.32 ± 0.40 mS/cm², n = 5, P = 0.014) consistent with a contribution of K_Ca1.1 to producing electrogenic K⁺ secretion. Paxilline at concentrations above 1.0 μM also can inhibit Ca²⁺ ATPase and the inositol 1,4,5-trisphosphate receptor (4, 43). The extent of paxilline action on K_Ca1.1 was examined by restricting addition to the mucosal bathing solution and focusing on the early portion of the epinephrine response to obviate nonspecific actions. Addition of paxilline to only the mucosal solution (1.0 μM) produced a similar inhibition of ^epiI_sc (Table 3) and ^epiΔG_t (−1.27 ± 0.36 mS/cm², n = 10, P = 0.006) as two sided addition, supporting an apical location for the K_Ca1.1 involved. Increasing the mucosal paxilline concentration to 10 μM enhanced inhibition of ^epiI_sc (Table 3) with a fractional inhibition of ∼50%, similar to the maximal value estimated for IbTx (Fig. 1D). Further inhibition at longer incubation may be confounded by inhibition of other proteins involved in Ca²⁺ signaling.

Table 3.

Action of mucosal paxilline on epinephrine-activated K⁺ secretion

	Paxilline (1.0 μM)			Paxilline (10 μM)
	pax-1ΔΔIsc, μA/cm2	pax-1ΔΔIsc/secΔIsc	n	pax-10ΔΔIsc, μA/cm2	pax-10ΔΔIsc/secΔIsc	n
Basal	+1.4 ± 1.4 (0.33)	−0.020 ± 0.033 (0.57)	10	+6.7 ± 1.1* (0.0001)	−0.155 ± 0.059* (0.0001)	12
Epinephrine	+16.0 ± 4.5* (0.006)	−0.170 ± 0.043* (0.004)	10	+43.3 ± 3.9* (0.0001)	−0.495 ± 0.042* (0.0001)	12

Values are means ± SE (P values). Secretagogue-activated I_sc (^secΔI_sc = secretagogue − control) was measured for epinephrine (Fig. 1), and inhibitor induced differences in ^secΔI_sc (^inhibΔΔI_sc = inhibitor − control) were compared as well as the fractional inhibition (^inhibΔΔI_sc/^secΔI_sc). Paxilline (mucosal) was added prior to secretagogue activation. Significant differences from zero are indicated by asterisks (P values).

Since paxilline promotes binding of IbTx and ChTx to K_Ca1.1 (26, 35, 55), ^epiI_sc was measured in the combined presence of paxilline and toxin. Treatment with the β2-adrenergic receptor antagonist ICI-118551 suppressed the Cl⁻ secretory transient (25, 87) which revealed the full K⁺ secretory time course. IbTx (300 nM) added during epinephrine stimulation produced greater inhibition of ^epiI_sc with paxilline (1.0 μM) present, and approached the fractional inhibition produced by 10 μM paxilline alone (Fig. 2A). ChTx also exhibited a similar synergistic inhibition of ^epiI_sc (Fig. 2B). These interactions between K_Ca1.1 inhibitors that bind to distinct sites in the channel (60, 88) provided additional support for the involvement of K_Ca1.1 in epinephrine-activated K⁺ secretion.

Fig. 2.

Sensitivity of epinephrine-activated K⁺ secretion to combined K⁺ channel blockers. Adjacent mucosae were pretreated with the β2-adrenergic receptor antagonist ICI-118551 (1.0 μM) added to the serosal solution 30 min prior to epinephrine (5 μM) stimulation. A: 2 of 4 mucosae were pretreated with mucosally added paxilline 55 min prior to epinephrine, at either 1.0 μM (dotted line) or 10 μM (dashed line). After maximal activation, IbTx (300 nM) was added mucosally to control (gray solid line) and 1.0 μM paxilline pretreated mucosae (asterisks). A control mucosa without channel inhibitors is shown (black solid line). Combined IbTx/pax fractional inhibition was 0.77 compared with maximal inhibition (10 μM paxilline), exceeding the expectation of 0.67 for independent inhibitor actions. B: 2 of 3 mucosae were pretreated mucosally with paxilline 60 min prior to epinephrine, at either 1.0 μM (dotted line) or 10 μM (dashed line). After maximal activation ChTx (300 nM) was added mucosally to control (solid line) and 1.0 μM paxilline-pretreated mucosae (asterisk). Combined ChTx/pax fractional inhibition was 0.74 compared with 10 μM paxilline, exceeding the expectation of 0.49 for independent inhibitor actions. C and D: 2 of 3 mucosae were pretreated with paxilline 50 min prior to epinephrine, at either 1.0 μM mucosal and serosal (dotted line) or 10 μM mucosal (dashed line). After maximal activation, either Tram34 (100 μM) or BaCl₂ (10 mM) was added mucosally to control (solid line) and paxilline-pretreated mucosae (asterisks).

The contribution of K_Ca3.1 (Kcnn4) to apical membrane K⁺ conductance was examined by sensitivity of ^epiI_sc to Tram34 (1, 24). The minor Tram34 inhibition observed in the presence of paxilline (Fig. 2C) indicated that the remaining ^epiI_sc was not due to a large involvement of K_Ca3.1 or the other K⁺ channels blocked by this high concentration of Tram34 (83). The relatively nonselective K⁺ channel blocker Ba²⁺ inhibited ^epiI_sc in the presence of paxilline (Fig. 2D), illustrating further the reliance of K⁺ secretion on apical membrane K⁺ channels (21, 30, 64). The incomplete inhibition of control ^epiI_sc by Ba²⁺ suggested that K_Ca1.1 was relatively resistant to this Ba²⁺ block.

K_Ca1.1 blockers inhibit purinergic K⁺ secretion.

ATP activates a transient electrogenic K⁺ secretory response followed by slower activating sustained K⁺ secretion (87). IbTx and paxilline inhibited both the transient and sustained components of purinergic K⁺ secretion (Fig. 3, Table 2). The fractional inhibition of the sustained response was similar to that for epinephrine activated K⁺ secretion, suggesting a ∼50% contribution by K_Ca1.1 to the apical membrane K⁺ conductance required for secretion. In contrast, the transient ATP response appeared to require K_Ca1.1 as the sole apical K⁺ channel, although the fractional inhibition was not statistically different from that for the sustained response.

Fig. 3.

Sensitivity of ATP-activated K⁺ secretion to K_Ca1.1 blockers. Isolated mucosae were stimulated by ATP (100 μM) addition to the mucosal solution from the standard basal condition with electrogenic secretion monitored by I_sc. A: adjacent mucosae were stimulated by ATP with one pretreated by IbTx (200 nM) addition to the mucosal solution 25 min prior to ATP (dotted line). B: adjacent mucosae were stimulated by ATP with one pretreated by paxilline (1.0 μM) addition to both mucosal and serosal solutions 20 min prior to ATP (dotted line).

K_Ca1.1 blockers inhibit K⁺ secretion during Cl⁻ secretory activation.

Electrogenic K⁺ secretion also accompanies the Cl⁻ secretion stimulated by prostaglandin E₂ (PGE₂) and carbachol (CCh) (64, 87). During these secretory modes, the rate of Cl⁻ secretion exceeds that for K⁺ such that the resulting I_sc is positive. As might have been anticipated, ^PGE2I_sc (Fig. 4, A and B) was more positive in the presence of paxilline (+31.8 ± 8.4 μA/cm², n = 4, P = 0.032); however, ^PGE2I_sc was not consistently altered with IbTx (−9.4 ± 18.0 μA/cm², n = 5, P = 0.63). By comparing the PGE₂-induced difference in I_sc from the prior steady-state epinephrine activation, the change in I_sc (^PGE2ΔI_sc) represented increases in Cl⁻ secretory rate together with any differences in K⁺ secretion between the epinephrine condition and the PGE₂ condition. In the presence of paxilline, ^PGE2ΔI_sc was essentially identical with control ^PGE2ΔI_sc (Table 4), indicating that the transition to this secretory mode by PGE₂ occurred primarily via increasing Cl⁻ secretion and likely leaving the K⁺ secretory rate similar to the rate activated by epinephrine. The significantly lower ^PGE2ΔI_sc in the presence of IbTx (Table 4) could result from decreased Cl⁻ secretion; but more likely given the result with paxilline, K⁺ secretion became more resistant to IbTx during PGE₂ activation.

Fig. 4.

Sensitivity of prostanoid and synergistic secretion to K_Ca1.1 blockers. Isolated mucosae were stimulated consecutively by serosal addition of epinephrine (5 μM), PGE₂ (3 μM), and CCh (10 μM). A: during the prostanoid secretion initiated by PGE₂, I_sc was measured in adjacent mucosae with one pretreated by IbTx (300 nM) addition to the mucosal solution 170 min prior to PGE₂ (dashed line). B: during PGE₂-initiated secretion, I_sc was measured in adjacent mucosae with one pretreated by paxilline (1.0 μM) addition to both mucosal and serosal solutions 60 min prior to PGE₂ (dashed line). C: during the synergistic secretion stimulated by CCh, I_sc was measured in adjacent mucosae with one pretreated mucosally by IbTx (100 nM) addition 170 min prior to CCh (dashed line). A third adjacent mucosa was treated mucosally with IbTx (100 nM) during CCh activation (dotted line). D: during CCh-stimulated synergistic secretion, I_sc was measured in adjacent mucosae with one pretreated by paxilline (1.0 μM) addition to both mucosal and serosal solutions 80 min prior to CCh (dashed line).

Table 4.

Action of K_Ca1.1 channel inhibitors on Cl⁻ and K⁺ secretion

	IbTx			Paxilline
	IbTxΔΔIsc, μA/cm2	IbTxΔΔIsc/secΔIsc	n	paxΔΔIsc, μA/cm2	paxΔΔIsc/secΔIsc	n
PGE2	−45.0 ± 15.3* (0.042)	−0.218 ± 0.058* (0.020)	5	+0.4 ± 25.1 (0.99)	+0.055 ± 0.127 (0.69)	4
CCh/PGE2	+90.4 ± 25.1* (0.023)	+0.499 ± 0.222 (0.088)	5	+98.3 ± 19.0* (0.014)	+0.467 ± 0.206 (0.108)	4

Values are means ± SE (P values). Secretagogue-activated I_sc (^secΔI_sc = secretagogue − control) was measured for PGE₂ and CCh/PGE₂ (Fig. 4), and inhibitor-induced differences in ^secΔI_sc (^inhibΔΔI_sc = inhibitor − control) were compared as well as the fractional inhibition (^inhibΔΔI_sc/^secΔI_sc). PGE₂ (3 μM) was added during steady-state epinephrine activation and CCh (10 μM) was added after PGE₂. Inhibitors were added prior to secretagogue activation: paxilline (mucosal and serosal) 1.0 μM; IbTx (mucosal) 300 nM. Significant differences from zero are indicated by asterisks (P values).

Activation of the synergistic secretory mode by stimulating with CCh in the presence of PGE₂ (25, 87) produced a large positive I_sc consistent with a high Cl⁻ secretory rate (Fig. 4, C and D). Inhibition of K_Ca1.1 led to a more positive ^CChI_sc with either IbTx (+81.0 ± 19.4 μA/cm², n = 5, P = 0.014) or paxilline (+138.8 ± 27.0 μA/cm², n = 4, P = 0.014). Addition of IbTx during the synergistic secretory response quickly led to a more positive I_sc consistent with inhibiting a synergistic component of K⁺ secretion. Comparing ^CChΔI_sc in the presence and absence of inhibitors indicated that IbTx and paxilline had similar actions on the synergistic mode (Table 4), likely suppressing the K⁺ secretory component of the response. The response to 10 μM paxilline added mucosally was not significantly different from that to 1 μM added to both mucosal and serosal baths (data not shown).

Distal colonic epithelial cells express K_Ca1.1.

Immunoblotting with a monoclonal antibody allowed detection of protein expression for K_Ca1.1α in colonic epithelial cells (Fig. 5). RT-PCR of mRNA from colonic epithelial cells using three primer sets (Fig. 6A) confirmed the presence of K_Ca1.1α and demonstrated splicing events. Sequences for guinea pig K_Ca1.1α have been deposited with GenBank (accession numbers JQ241358, JQ241359, JQ241360). The product from the exon-16 to exon-20 region indicated the absence of the insertion exon STREX, which is termed the ZERO variant (84). Inclusion of STREX would have resulted in a larger product. Primers for detecting the alternative terminations for K_Ca1.1α yielded products (Fig. 6A) consistent with termination in exon-27 or an early termination in exon-28 (Fig. 6B). The termination in exon-27 resulted in an amino acid sequence ending of QEERL, whereas the termination in exon-28 resulted in an amino acid sequence ending of EMVYR (15). The splicing of exon-28 occurred with two variants that differed by an omission of 3 bp at the beginning of exon-28. Insertions and omissions expected for shifts in the reading frame that lead to the other two known endings (53, 65, 85) were not detected in the PCR products.

Fig. 5.

Detection of K_Ca1.1α in colonic epithelial cells. Protein isolated from the membrane fraction of distal colonic epithelial cells was immunoblotted with an antibody against the K_Ca1.1 protein (Kcnma1) NeuroMab clone L6/60. An immunoreactive band consistent with K_Ca1.1 occurred at ∼100 kDa (arrowhead) using this monoclonal antibody. Use of the secondary antibody alone eliminated all bands (data not shown), indicating that the primary antibody was necessary for the observed result.

Fig. 6.

Expression of mRNA for K_Ca1.1α. A: RT-PCR of colonic epithelial cell mRNA amplified K_Ca1.1(Kcnma1) products with sizes predicted by the position of the primers, 357 base pairs for the segment from exon-16 to exon-20 (534 bp including the insertion exon STREX), 351 base pairs for exon-26 to exon-27, and 445 base pairs for exon-26 to exon-28 (numbering based on constitutive exons, 15). Sequencing of products confirmed identity with K_Ca1.1 (homology with human 95%, 95%, 97%). Amplification of a GAPDH product (555 bp) served as positive control for RNA isolation. Absence of product when not including reverse transcriptase (RT) indicated further the lack of genomic DNA contamination. B: sequences for rat, mouse, human, and guinea pig are shown (differences from human indicated by shading) together with the splice site for the alternative COOH termini of K_Ca1.1 (alternative reading frames illustrated by the distinct amino acid sequences). The beginning of exon-28 is indicated (◇) as well as amino acid variations resulting from sequence differences (=). QEERL ending: [27] exon-27(stop), human NM_001161352, mouse NM_001253364; EMVYR endings: [28a] exon-28(stop), human NM_001014797, mouse NM_010610; [28a*] exon-28(skip 1st 3 bp; stop), mouse NM_001253363, rat NM_031828. VEDEC endings: [28b] insert(1 bp) + exon-28(skip 1st 2 bp; stop), rabbit AB009312; [28b*] insert(3n + 2 bp) (n = 9, 3′-end exon-27) + exon-28(stop), mouse NM_001253358; FKSSP endings: insert(3n + 1 bp) + exon-28(stop); [28c] n = 0 rat U93052 or [28c*] n = 15 rat U55995. The transcripts in guinea pig distal colonic epithelial cells included three alternate sequences: QEERL[27], EMVYR[28a], and EMVYR[28a*].

The β-subunit for K_Ca1.1, K_Caβ (Kcnmb), modifies activation kinetics and inhibitor sensitivity (3, 9, 40, 51, 60). RT-PCR of mRNA from colonic epithelial cells indicated the presence of K_Caβ1 and K_Caβ4 (Fig. 7). Expression of K_Caβ2 mRNA was detected in colonic mucosal samples, serving as a positive control for the apparent absence of K_Caβ2 in colonic epithelial cells and indicating expression in an interstitial cell type only. Sequences for guinea pig K_Caβ1, K_Caβ2, and K_Caβ4 have been deposited with GenBank (accession numbers JQ241361, JQ241362, JQ241363). Primers designed from either the human or rat sequence (61, 68) failed to result in products consistent with K_Caβ3, leaving the presence of K_Caβ3 in guinea pig distal colonic epithelial cells unresolved.

Fig. 7.

Expression of mRNA for K_Caβ. A: RT-PCR of colonic epithelial cell mRNA and colonic mucosal mRNA amplified Kcnmb products with sizes predicted by the position of the primers, 477 base pairs for Kcnmb1 (K_Caβ1), 405 base pairs for Kcnmb2 (K_Caβ2), and 456 base pairs for Kcnmb4 (K_Caβ4). Sequencing of products confirmed identity with Kcnmb (homology with human 90%, 95%, 95%). Amplification of a GAPDH product (555 bp) served as positive control for RNA isolation. B: absence of product when not including reverse transcriptase indicated the lack of genomic DNA contamination.

Localization of K_Ca1.1 in distal colonic epithelial cells.

Numerous studies have indicated the presence of K_Ca1.1 in colonic epithelial cells within the apical and basolateral membranes at both surface and crypt locations (14, 28, 47, 70, 71, 75, 77). Expression of K_Ca1.1α mRNA was examined by RT-PCR for samples from crypts isolated by manual dissection. Crypts were sectioned to remove any surface cells and included only the lower two-thirds to one-half of each crypt. Portions of intact mucosa were processed as well as portions of mucosa after the crypts had been removed entirely. A product consistent with that found in larger mucosal samples (Fig. 6) was detected in crypt samples but not from surface epithelium (Fig. 8).

Fig. 8.

Distribution of K_Ca1.1α in the colonic epithelium. RT-PCR of colonic crypt mRNA amplified a 357 base pair K_Ca1.1 product for the segment from exon-16 to exon-20. Samples included 9 crypts, 7 crypts, mucosal piece containing 6 crypts, and surface epithelium after complete removal of 10 crypts from a mucosal piece. GAPDH products indicated isolation of RNA from each sample.

Immunolocalization (Fig. 9) exhibited K_Ca1.1α immunoreactivity (K_Ca1.1α^IR) in colonic epithelial cells from both the surface and crypt regions. The use of three antibodies validated in several cell types (6, 53, 54, 59) supported the identity of this localization as K_Ca1.1α. For both surface and crypt cells, K_Ca1.1α^IR occurred as puncta along the cell margins consistent with a presence in lateral membranes (Fig. 9). Apical localization of K_Ca1.1α was apparent in the brush border of surface cells (Fig. 9, A and B) and in the apical pole of crypt cells (Fig. 9, D–G). The termination in exon-27 results in a protein ending with the amino acids, QEERL. Using an anti-serum specific for the chicken COOH-terminal sequence (QEDRL), K_Ca1.1α^IR was detected in the brush border and lateral membranes of surface cells (Fig. 9C). K_Ca1.1α^IR-(QEDRL) was not apparent in the apical pole of crypt cells, but rather only in cell margins close to the base of crypt cells (Fig. 9H). A longitudinal profile along the edge of a crypt showed this K_Ca1.1α^IR-(QEDRL) as the outline of cell margins (Fig. 9I). Perhaps an alternate COOH terminus such as EMVYR allowed trafficking to the apical location within crypt cells.

Fig. 9.

K_Ca1.1α localization in the colonic epithelium. K_Ca1.1α was detected in distal colonic mucosa by immunoreactivity to 3 antibodies (K_Ca1.1α^IR; green; b). Actin detected with labeled phalloidin revealed the morphology of the epithelium (red; c) for cellular localization of K_Ca1.1α^IR (a). Prominent K_Ca1.1α^IR was apparent in surface cells along the lateral cell margins (arrowhead) and the brush-border membrane (arrow). Labeling was similar with each antibody, anti-K_Ca1.1α-L6/60 (A), anti-K_Ca1.1α-H300 (B), and anti-K_Ca1.1α-QEDRL (C). Scale bars, 5 μm. Crypt cross-section and longitudinal profiles are shown (D–I). Prominent K_Ca1.1α^IR was apparent in crypt cells along the lateral cell margins (arrowhead). Apical poles also exhibited punctate K_Ca1.1α^IR (arrow). Labeling was similar with anti-K_Ca1.1α-L6/60 (D and E) and anti-K_Ca1.1α-H300 (F and G) antibodies. Anti-K_Ca1.1α-QEDRL (H and I) exhibited K_Ca1.1α^IR only along the lateral margins at the basal end of the crypt cells (arrowhead); luminal margin is indicated by an arrow. Scale bars, 10 μm. Use of the secondary antibody alone led to loss of labeling (data not shown), indicating that the primary antibodies were necessary for the labeling observed.

DISCUSSION

Sympathetic activation of K⁺ secretion.

β-Adrenergic activation of electrogenic K⁺ secretion conforms to the cellular model for Cl⁻ secretion with the inclusion of apical membrane K⁺ channels (21, 29). Sensitivity to IbTx and paxilline supported a reliance on K_Ca1.1 for ∼50% of apical membrane K⁺ exit (Fig. 1, D and E, Table 3). The relatively high IC₅₀ for IbTx (∼200 nM) may relate to the involvement of K_Caβ subunits that reduce blocker affinity. A splice variant of K_Ca3.1 (Kcnn4) with low sensitivity to Tram34 occurs in the apical membrane of rat colon (1), but the only minor inhibition by Tram34 at high concentration (24, Fig. 2C) indicates that K_Ca3.1 was likely not responsible for the remaining apical K⁺ exit during epinephrine activated K⁺ secretion.

Previous studies comparing epinephrine responses in distal colon from K_Ca1.1-null mice and wild-type mice clearly demonstrate a requirement for the K_Ca1.1 channel in producing electrogenic K⁺ secretion (76). However, because of the calculations necessary to account for the concomitant Cl⁻ secretion during the early time course examined, the exact proportion of the β-adrenergic K⁺ secretion dependent on K_Ca1.1 was not determined definitively. The higher rate of β-adrenergic K⁺ secretion in guinea pig distal colon (∼5-fold) made the determination of the contribution by K_Ca1.1 more precise (Fig. 1D), but it may be that the guinea pig generated a higher K⁺ secretion than mouse simply by using additional types of K⁺ channels in the apical membrane. If these K⁺ channel types activate with distinct time courses, then sensitivity during the early and sustained phases would differ. Overall, K_Ca1.1 provided a significant if not sole contribution to the apical membrane K⁺ conductance activated by epinephrine in mammalian distal colon.

Purinergic activation of K⁺ secretion.

Addition of ATP to the luminal side of the distal colon produces a small transient electrogenic K⁺ secretion that is absent in K_Ca1.1-null mice (70). This absolute dependence on K_Ca1.1 was supported by the inhibitor sensitivity of this transient ATP response in guinea pig distal colon (Fig. 3, Table 2). The sustained component of the ATP response involving adenosine (87) had an inhibitor sensitivity supporting a smaller contribution by K_Ca1.1 during adenosine receptor stimulation, similar to that for epinephrine-activated K⁺ secretion (Table 2). Thus each K⁺ secretagogue activates a specific set of K⁺ channels with K_Ca1.1 taking a smaller or larger role depending on the character of the physiological response.

Cl⁻ secretagogues.

Activation of Cl⁻ secretion in colon generally includes an attendant electrogenic K⁺ secretion (21, 34, 64). The relative proportion of K⁺ to Cl⁻ secretion varies with secretagogue type and time after activation. Inhibition of apical membrane K⁺ channels increases the measured I_sc (as K⁺ secretion diminishes) whereas inhibition of basolateral membrane K⁺ channels decreases I_sc (as Cl⁻ secretion diminishes). The generally positive I_sc during Cl⁻ secretagogue activation in the presence of K_Ca1.1 blockers (Fig. 4, B and D), or in colonic mucosa from K_Ca1.1-null mice (50), suggested that K_Ca1.1 present in the basolateral membrane were of minor importance for sustaining Cl⁻ secretion. Instead, apical membrane K_Ca1.1 contributed to K⁺ secretion during prostanoid (PGE₂) and synergistic (PGE₂/CCh) stimulation of Cl⁻ secretion.

Apical membrane K⁺ channels provide a route for K⁺ exit during K⁺ secretion, but also contribute to setting the membrane electrical potential difference and thereby to defining the electrochemical gradients for ion flow. Together with basolateral K⁺ channels, apical K⁺ channels maintain a negative membrane potential that drives Cl⁻ exit. As presented for rat distal colon (36), apical K⁺ channels could ensure apical Cl⁻ exit and perhaps have the largest influence during periods of similar K⁺ and Cl⁻ secretory rates when apical K⁺ conductance would be substantial. Inhibition of the PGE₂ response by mucosally added IbTx (Table 4) seemed to support this concept of apical K⁺ channels maintaining apical Cl⁻ exit, except that block with paxilline left the PGE₂ response unaltered (Table 4). This discrepancy of action for the two K_Ca1.1 blockers supported an alternative proposal in which PGE₂ increased K⁺ secretion above the epinephrine-stimulated rate by activating K_Ca1.1 even more resistant to IbTx, possibly by recruiting K_Ca1.1α with different K_Caβ's. The similarity of blocker sensitivity with PGE₂/CCh stimulation (Table 4) supported a return of moderate IbTx sensitivity during synergistic activation.

Epithelial localization of K_Ca1.1.

An epithelial presence for K_Ca1.1 permits a contribution to various transepithelial flows. Apical localization provides for K⁺ exit into the lumen in support of K⁺ secretion, while basolateral localization supports other flows dependent on the basolateral Na⁺/K⁺-ATPase (2, 30). Both surface and crypt cells of guinea pig distal colon exhibited K_Ca1.1α^IR in apical and basolateral membranes (Fig. 9). K_Ca1.1α^IR in crypts suggested localization in apical vesicles perhaps representing a pool of channels ready for insertion. Previous studies support a more restrictive K_Ca1.1α localization. Immunolabeling and IbTx binding indicate a predominant presence in apical and basolateral membranes of surface cells but with an increase in crypt labeling during various stresses (14, 28, 47, 68, 71). Other studies localize K_Ca1.1α^IR only to crypt apical membranes (70, 75). K_Ca1.1α mRNA distribution also supported the presence in colonic crypts (77, Fig. 8). The source of these discrepancies in localization may include species differences, physiological state of the animals, and sensitivity of detection methods. The high K⁺ secretory rate exhibited by guinea pig distal colon indicated a normal physiological state with a large contingent of apical membrane K⁺ channels, thus making detection easier.

Alternative splicing of K_Ca1.1α mRNA alters many properties of the resulting channels including regulation and membrane trafficking (3, 10). Inclusion of the STREX exon converts K_Ca1.1 sensitivity from cAMP activation to inhibition (5, 79). The mouse colon expresses both the ZERO and STREX variants (13, 76) with stimulation by aldosterone making the ZERO variant predominant (76) thereby strengthening cAMP activation. The apparently exclusive expression of the ZERO variant in guinea pig distal colon (Fig. 6A) presumably contributed to the large capacity for K⁺ secretion by favoring K_Ca1.1 activation.

Trafficking of K_Ca1.1 to apical and basolateral membranes determines its role either as a component of the K⁺ secretory path or as a supporting element for Cl⁻ secretion or Na⁺ absorption. COOH-terminal splicing of K_Ca1.1 alters its membrane expression with QEERL and EMVYR variants trafficking to the plasma membrane (7, 45). The lack of K_Ca1.1α^IR for the QEERL variant in the crypt apical pole (Fig. 9H) suggested that the COOH terminus contributes to the routing signals for the apical and basolateral membranes (37). Since surface cell apical membranes exhibited K_Ca1.1α^IR for the QEERL variant (Fig. 9C), this apical membrane targeting distinction appeared to be specific for crypt cells, which would require the EMVYR variant.

Epithelial regeneration in the colon occurs via stem cells in the base of the crypt producing cells that move along the crypt structure to positions in the surface epithelium (30). During this progression K_Ca1.1α^IR appeared in both apical and basolateral membranes (Fig. 9), such that subunit composition of apically located K_Ca1.1 changed as cells moved from crypt to surface (Fig. 9, C and H). K_Ca1.1α mRNA expression also decreases in the transition from crypt to surface of mouse colon (77). The apparent lack of mRNA for K_Ca1.1α in guinea pig surface cells supported this finding (Fig. 8) despite the surface presence of K_Ca1.1α^IR (Fig. 9, A–C). Presumably K_Ca1.1α persisted during the ∼1 day duration of surface cells without replenishment by protein synthesis, suggesting a relatively long life-time for the K_Ca1.1α protein.

Auxillary K_Ca1.1 subunits.

The K_Caβ subunits (Kcnmb) contribute to cell type-specific activity of K_Ca1.1 (3, 9, 40, 60). Of the behaviors promoted by these subunits, K_Caβ1 would best suit sustained K⁺ secretion by its increasing of Ca²⁺ sensitivity and shifting of voltage activation toward negative values. The slower activation and lower Ca²⁺ sensitivity produced with K_Caβ4 also might serve K⁺ secretory requirements, but the rapid inactivation conferred by K_Caβ2 and some splice variants of K_Caβ3 would limit the utility of K_Ca1.1 for sustained K⁺ secretion. Previous studies indicate expression of K_Caβ mRNA in colonic epithelium with conflicting results. In mouse colon, K_Caβ1, K_Caβ2, and K_Caβ4 were detected (14, 62), but another study detected only K_Caβ2 (75). Human sigmoid colon expresses all four K_Caβ's (68). Perhaps reflecting species differences, guinea pig distal colonic epithelial cells expressed only K_Caβ1 and K_Caβ4 (Fig. 7).

K_Caβ subunits also confer differential sensitivities to the toxin blockers IbTx and ChTx (51, 60) permitting a functional test of which K_Caβ's combine with K_Ca1.1 to facilitate K⁺ secretion. Interaction with K_Caβ1 or K_Caβ4 lowers sensitivity of K_Ca1.1 to IbTx, whereas K_Caβ2 or K_Caβ4 lowers sensitivity to ChTx. The low sensitivity of K⁺ secretion in guinea pig distal colon to both IbTx and ChTx (Figs. 1D and 2B, Table 2) supported the involvement of K_Caβ4 together with K_Ca1.1 in a secretory K⁺ channel complex. Since K_Ca1.1 and K_Caβ combine in a 1:1 manner, the tetrameric K_Ca1.1 channel complex would include four individual K_Caβ subunits. The extremely low sensitivity for IbTx conferred by K_Caβ4 (57) suggested that the observed K⁺ secretory sensitivity to IbTx resulted from a K_Ca1.1 complex including both K_Caβ1 and K_Caβ4 such that ChTx sensitivity was reduced.

K⁺ secretory activation.

A variety of secretagogues stimulate colonic K⁺ secretion apparently via cells capable of Cl⁻ secretion (29, 64) with an output ranging from primarily K⁺ secretion (Figs. 1 and 3) to mostly Cl⁻ secretion (Fig. 4). Salivary glands also secrete fluid driven by electrogenic Cl⁻ secretion that includes K⁺ secretion. The apical K⁺ exit involves K_Ca1.1α together with K_Caβ1 and K_Caβ4 (57), similar to guinea pig distal colonic K⁺ secretion. Recent studies demonstrate that flow-dependent K⁺ secretion in the kidney emanates from intercalated cells using apical membrane K_Ca1.1 and basolateral Na⁺:K⁺:2Cl⁻-cotransporters (41). Localization of K_Caβ's in the kidney indicates that K_Caβ4 occurs in the intercalated cells, and K_Caβ1 segregates to neighboring principal cells (19). Thus intercalated cells secrete K⁺ via K_Ca1.1/K_Caβ4, together with acid secretion, whereas principal cells secrete K⁺ via K_Ca1.1/K_Caβ1, together with Na⁺ absorption (31). A common theme for K⁺ secretory cells is the matching of K⁺ channel complexes to the physiological demands of the organ; and distal colonic epithelium will use K_Ca1.1/K_Caβ complexes that best serve the physiological requirements of this distal-most gastrointestinal site.

GRANTS

This study was supported by a grant from the National Institutes of Health and National Institute of Diabetes and Digestive and Kidney Diseases (DK-65845).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: J.Z., S.T.H., and D.R.H. conception and design of research; J.Z., S.T.H., and D.R.H. performed experiments; J.Z., S.T.H., and D.R.H. analyzed data; J.Z., S.T.H., and D.R.H. interpreted results of experiments; J.Z. and D.R.H. drafted manuscript; S.T.H. and D.R.H. edited and revised manuscript; D.R.H. prepared figures; D.R.H. approved final version of manuscript.