AE2 Cl-/HCO3- exchanger is required for normal cAMP-stimulated anion secretion in murine proximal colon

Abstract

Anion secretion by colonic epithelium is dependent on apical CFTR-mediated anion conductance and basolateral ion transport. In many tissues, the NKCC1 Na(+)-K(+)-2Cl(-) cotransporter mediates basolateral Cl(-) uptake. However, additional evidence suggests that the AE2 Cl(-)/HCO(3)(-) exchanger, when coupled with the NHE1 Na(+)/H(+) exchanger or a Na(+)-HCO(3)(-) cotransporter (NBC), contributes to HCO(3)(-) and/or Cl(-) uptake. To analyze the secretory functions of AE2 in proximal colon, short-circuit current (I(sc)) responses to cAMP and inhibitors of basolateral anion transporters were measured in muscle-stripped wild-type (WT) and AE2-null (AE2(-/-)) proximal colon. In physiological Ringer, the magnitude of cAMP-stimulated I(sc) was the same in WT and AE2(-/-) colon. However, the I(sc) response in AE2(-/-) colon exhibited increased sensitivity to the NKCC1 inhibitor bumetanide and decreased sensitivity to the distilbene derivative SITS (which inhibits AE2 and some NBCs), indicating that loss of AE2 results in a switch to increased NKCC1-supported anion secretion. Removal of HCO(3)(-) resulted in robust cAMP-stimulated I(sc) in both AE2(-/-) and WT colon that was largely mediated by NKCC1, whereas removal of Cl(-) resulted in sharply decreased cAMP-stimulated I(sc) in AE2(-/-) colon relative to WT controls. Inhibition of NHE1 had no effect on cAMP-stimulated I(sc) in AE2(-/-) colon but caused a switch to NKCC1-supported secretion in WT colon. Thus, in AE2(-/-) colon, Cl(-) secretion supported by basolateral NKCC1 is enhanced, whereas HCO(3)(-) secretion is diminished. These results show that AE2 is a component of the basolateral ion transport mechanisms that support anion secretion in the proximal colon.

Fig. 1.

Gross anatomy and histology of cecum and colon. Photographs of cecum and colon from wild-type (WT; A) and AE2^−/− (E) mice. Note abnormal morphology of AE2^−/− cecum (E), which tended to be less elongated and to have greater fluid content. Arrows in E indicate presence of gas bubbles in AE2^−/− cecum and colon. Histological analyses of Alcian blue and periodic acid-Schiff-stained sections of cecum, proximal colon, and distal colon of WT (B, C, and D, respectively) and AE2-null (F, G, and H, respectively) mice revealed no significant differences. Goblet cells containing neutral mucins appear magenta whereas goblet cells containing sialomucins appear blue. Scale bars represent 50 μm.

Fig. 2.

pH measurements of intestinal tract contents. The contents of the small intestine (SI), cecum, proximal (p.) colon, and distal (d.) colon from 17- to 18-day-old mice (n = 3 each genotype) were suspended in water and the pH was measured. AE2^−/− small intestine, cecum, and proximal colon had significantly more alkaline contents compared with WT controls (*P < 0.05).

Fig. 3.

cAMP-stimulated anion secretion in AE2^−/− proximal colon is more dependent on NKCC1 than in WT tissue. Tissues were mounted in Ussing chambers and bathed in Ringer solution containing Cl⁻ and HCO₃⁻ (indicated in box above each panel), and changes in short circuit currents (I_sc) were measured in WT (A) and AE2^−/− (B) proximal colons (n = 4 each genotype) in response to sequential treatment with forskolin/IBMX (cAMP stimulation), bumetanide, and 4-acetoamido-4′-isothiocyanato-2–2′-stilbene disulfonic acid (SITS). Labeled arrows indicate time of drug addition; dashed line indicates baseline I_sc. C: summary of data in A and B, showing ΔI_sc responses to cAMP (plateau current 20 min after cAMP stimulation), reductions in I_sc following treatment with bumetanide and SITS (indicated by negative currents), and residual currents. *Significantly different from WT control for same treatment.

Fig. 4.

Expression of NKCC1 is not significantly changed in AE2^−/− proximal colon. A: NKCC1 protein expression was not different between WT (+/+) and AE2^−/− (−/−) proximal colon by Western blot analysis using 28 μg of whole protein lysate. B: quantification of NKCC1 expression relative to actin expression for WT and AE2^−/− proximal colon (n = 3 each genotype). AE2^−/− proximal colon exhibited a nonsignificant increase (29%) in NKCC1 expression.

Fig. 5.

AE3 protein expression is not altered in AE2^−/− colonic epithelium. Homogenates of epithelial and muscle layers from proximal colons of juvenile WT and AE2^−/− (KO) mice were analyzed by immunoblotting using anti-AE2 and anti-AE3 antibodies. A: AE2 protein expression was lost in colonic epithelium of AE2^−/− mice whereas robust expression was observed in WT samples [40 μg/lane, n = 4 mice of each genotype; the control (Co) is WT brain homogenate (30 μg/lane)]. B: AE3 protein expression was below the limits of detection in WT colonic epithelium and was not increased in AE2^−/− colonic epithelium; control total brain homogenates showed robust expression of full-length AE3. C: immunoblot analysis of epithelial (EP) and muscle (M) layers from proximal colons of juvenile WT mice revealed that AE2 protein expression was largely restricted to the colonic epithelium (n = 4). In contrast, AE3 protein expression was barely detectable in epithelial and muscle samples, even after relatively long exposure times.

Fig. 6.

Cl⁻-dependent I_sc responses to cAMP-stimulation and bumetanide are the same in WT and AE2^−/− proximal colon. Tissues were mounted in Ussing chambers, bathed in HCO₃⁻-free Ringer solution containing only Cl⁻ as a transported anion (indicated in box above each panel), and changes in I_sc were measured in WT (A) and AE2^−/− (B) proximal colons (n = 4 each genotype) in response to cAMP-stimulation and bumetanide. Labeled arrows indicate time of drug addition; dashed line indicates the baseline I_sc. C: summary of data in A and B, showing ΔI_sc responses to cAMP (plateau current 20 min after cAMP-stimulation), reductions in I_sc following treatment with bumetanide, and residual currents.

Fig. 7.

HCO₃⁻-dependent I_sc responses to cAMP-stimulation and SITS are reduced in AE2^−/− proximal colon relative to WT controls. Tissues were mounted in Ussing chambers and bathed in Cl⁻-free Ringer solution containing only HCO₃⁻ as a transported anion (indicated in box above each panel), and changes in I_sc were measured in WT (A) and AE2^−/− (B) proximal colons (n = 4 each genotype) in response to cAMP stimulation and SITS. Labeled arrows indicate time of drug addition; dashed line indicates the baseline I_sc. C: summary of data in A and B, showing ΔI_sc responses to cAMP (plateau current 20 min after cAMP stimulation), reductions in I_sc following treatment with SITS, and residual currents. *Significantly different from WT control for same treatment.

Fig. 8.

Acetazolamide has no effect on HCO₃⁻-dependent I_sc responses to cAMP-stimulation and SITS treatment in AE2^−/− proximal colon. Tissues were mounted in Ussing chambers, bathed in Cl⁻-free Ringer solution containing only HCO₃⁻ as a transported anion (indicated in box above each panel), and treated with acetazolamide to inhibit carbonic anhydrase. Changes in I_sc were measured in WT (A) and AE2^−/− (B) proximal colons (n = 4 each genotype) in response to cAMP stimulation and SITS. Labeled arrows indicate time of drug addition; dashed line indicates the baseline I_sc. C: summary of data in A and B, showing ΔI_sc responses to cAMP (plateau current 20 min after cAMP stimulation), reductions in I_sc following treatment with SITS, and residual currents. *Significantly different from WT control for same treatment.

Fig. 9.

Inhibition of the NHE1 Na⁺/H⁺ exchanger results in similar I_sc responses to cAMP, bumetanide, and SITS in WT and AE2^−/− proximal colon. Tissues were mounted in Ussing chambers and bathed in Ringer solution containing Cl⁻ and HCO₃⁻, with EIPA in the serosal solution to inhibit NHE1 (indicated in box above each panel). Changes in I_sc were measured in WT (A) and AE2^−/− (B) proximal colons (n = 4 each genotype) in response to cAMP stimulation, bumetanide, and SITS. Labeled arrows indicate time of drug addition; dashed line indicates baseline I_sc. C: summary of data in A and B, showing ΔI_sc responses to cAMP (plateau current 20 min after cAMP-stimulation), reductions in I_sc following treatment with bumetanide and SITS, and residual currents. *Significantly different from WT control for same treatment.

Fig. 10.

Expression of carbonic anhydrase isoforms 2 and 3 are not changed in AE2^−/− proximal colon. Whole colon homogenates (20 μg/lane) from WT and AE2^−/− mice were analyzed by immunoblotting using anti-Car2 (A) and anti-Car3 (B) antibodies, with actin as a loading control. C: quantification of Car2 and Car3 expression relative to actin expression for WT and AE2^−/− proximal colon (n = 3 for each genotype) revealed no significant differences in expression for either Car2 or Car3.