Differential expression and regulation of AE2 anion exchanger subtypes in rabbit parietal and mucous cells

Abstract

1. The anion exchanger isoform 2 (AE2) gene encodes three subtypes (AE2a, b and c), which have different N-termini and tissue distributions. AE2 is expressed at high levels in the stomach, where it is thought to mediate basolateral base exit during acid production. The present study investigated if the three AE2 subtypes are differentially expressed and regulated in different cell types within the gastric mucosa. 2. The cloning strategy to obtain rabbit AE2a, b and c cDNAs combined genomic PCR and RT-PCR based on primers deduced from the rat sequences. Semiquantitative RT-PCR using homologous primers revealed much higher AE2 mRNA expression in rabbit parietal cells (PCs) than in mucous cells (MCs). The subtype expression pattern was AE2b >> AE2c > or = AE2a in PCs and AE2a >AE2b >> AE2c in MCs. Sequence analysis revealed the presence of a highly conserved protein kinase C (PKC) consensus sequence in the AE2a alternative N-terminus. 3. Maximal Cl(-)-HCO(3)(-) exchange rates, measured fluorometrically in BCECF-loaded cultured gastric cells, were much higher in PCs than MCs. PKC activation by phorbol ester stimulated maximal Cl(-)-HCO(3)(-) exchange rates in MCs but not in PCs, whereas forskolin had no effect in each cell type. 4. In summary, rabbit PCs and MCs, which originate from the same gastric stem cell population, display a completely different AE2 subtype expression pattern. Therefore, AE2 subtype expression is not organ specific but cell type specific. The different regulation of anion exchange in parietal and mucous cells suggests that AE2 subtypes may be differentially regulated.

Figure 1. Poly(A⁺) RNA Northern blots (A and B) and analysis (C) of the band pattern in parietal cells, mucous cells and gastric mucosa by densitometry

Hybridization of (A) ≈5 μg poly(A⁺) RNA from rabbit kidney cortex (lane 1), colonic mucosa (lane 2), gastric mucosa (lane 3), parietal cell (lane 4), chief cell (lane 5) and mucous cell (lane 6) fractions and of (B)≈4 μg poly(A⁺) RNA of rabbit antrum mucosa (lane 1), gastric mucosa (lane 2) and parietal cells (lane 3) with a ³²P-labelled AE2 cDNA fragment (bp 2484-2742 of S45791, common to rabbit AE2a, AE2b and AE2c) as described in Methods. Exposure times (expo.) are shown in parentheses beside each blot (in hours (h) and days (d)). C, the bands were analysed by densitometry, which showed no optical saturation. Mucous cell sample: blot A, long exposure; gastric mucosa and parietal cell sample: short exposure of blot B.

Figure 2. RT-PCR analysis of AE2 mRNA variants, expressed in rabbit gastric mucosa, using specific forward primers for AE2a (exon 2), AE2b (exon 1b) and AE2c (exon 1c)

Upper panel, ethidium bromide (EtBr)-stained AE2 amplification products (28 cycles) of RNA samples from rabbit gastric mucosa (lanes 1, 4, 7), rabbit mucous cells (lanes 2, 5, 8), and parietal cells (lanes 3, 6, 9). Histone 3.3a amplimers (lane 10: gastric mucosa; lane 11: mucous cells; lane 12: parietal cells) underwent 22 cycles of amplification. Lane 13 contains the H₂O control, lane 14 the molecular weight standard. All lanes were loaded with a 13 μl aliquot of a 50 μl PCR reaction. Lower panel, the cDNAs shown in the upper panel were transferred to a nylon membrane and hybridized to a ³²P-labelled internal AE2 oligonucleotide probe (exon 6, nucleotides (nt) 673-722 of S45791) encoding a sequence present in all AE2 variants.

Figure 3. Semiquantitative RT-PCR analysis of rabbit AE2 variants

A, the ratios AE2a/histone 3.3a (representing the relative expression level of AE2a), AE2b/histone 3.3a and AE2c/histone 3.3a were plotted for gastric mucosa, the different cell fractions and kidney cortex. The shaded columns represent the expression of AE2a, AE2b and AE2c in rabbit mucous cells after subtraction of that part of the AE2 signal which is caused by a 5 % contamination of the mucous cell fraction with parietal cells (as explained in Results). n = 3 from three different cell preparations. B, representative curves illustrating the amplification of each AE2 subtype (AE2a, AE2b, AE2c) and control (histone 3.3a) cDNA fragments from rabbit parietal cell RNA. The ratio of the gene of interest/histone 3.3a was determined during the exponential phase of both reactions.

Figure 4. Membrane localization of the AE2 protein by Western blot analysis of purified rabbit gastric tubulo-vesicular and basolateral membrane fractions

Protein (100 μg) from total mucosal homogenate (lane 3), tubulo-vesicular membranes (lane 2) and basolateral membranes (lane 1) isolated from rabbit gastric mucosa were separated by 10 % SDS-PAGE, electroblotted, and stained by Ponceau S (to ensure equal loading in all lanes and to check protein quality, lower panel). The membrane shown in the upper panel was incubated with a rabbit polyclonal anti-AE2 antibody. The middle panel demonstrates the same filter, stripped, and incubated with an anti-H⁺-K⁺-ATPase α-subunit antibody.

Figure 5. AE2 immunolocalization in mouse gastric mucosa, detected by antibody to mouse C-terminal amino acids 1224-1237 in the presence of irrelevant peptide (a and c) or of peptide antigen (b and d)

AE2 is present at modest abundance in basolateral membranes of surface mucous cells (a) and at very high abundance (saturating pixel intensity) in basolateral membranes of parietal cells (panel c and panel a, lower right). The identical exposures in all panels were selected to optimize antibody signal in the surface mucous cells (a). In exposures chosen to optimize parietal cell staining (Stuart-Tilley et al. 1994), surface mucous cell staining was undetectable. Scale bar, 15 μm.

Figure 6. Proton/base flux rates in cultured parietal (A and B) and mucous cells (C and D) in the presence of CO₂-HCO₃⁻ and in the presence and absence of Cl⁻ and/or 1 mm DIDS, an anion exchange inhibitor

Parietal cells (A) and mucous cells (C) alkalized in Cl⁻ free medium. Addition of Cl⁻-containing medium was associated with a quick pH_i recovery, which was inhibited by 1 mm DIDS (A and C). pH_i recovery rates (indicated by dashed lines) and the intracellular buffering capacity at the appropriate pH_i were used to calculate the proton/base flux rates shown in B and D. Proton/base flux rates were ≈12-fold higher in parietal cells than in mucous cells, but in both cell types the addition of 1 mm DIDS blocked more than 91 % of the proton/base flux (n = 4-6 in each group, * P < 0.05, ** P < 0.01).

Figure 7. Multiple sequence alignment (MALIGN, Husar 5.0 Sequence Analysis, EMBL, Heidelberg, Germany) of the 40 N-terminal amino acids of AE2

From top to bottom: mouse (Mus musculus), rat (Rattus norvegicus), guinea-pig (Cavia porcellus), man (Homo sapiens) and rabbit (Oryctolagus cuniculus) AE2a sequence. Serine 14 is conserved in all known species. It is predicted to be phosphorylated by PKC (consensus pattern: serine or threonine-one arbitrary amino acid-one basic amino acid (lysine, arginine, or histidine); the presence of additional basic residues at the N- or C-terminus of the target amino acid enhances the V_max and K_m of the phosphorylation reaction (appendix to PROSITE, Husar 5.0 Sequence Analysis) in mouse, rat, guinea-pig and rabbit. Even in man, where the consensus pattern is atypical, a possible phosphorylation of serine 14 is not excluded. Furthermore Wang et al. (1965) described an atypical PKA consensus site at serine 10 of the rat sequence, which is conserved in mouse, rabbit and guinea-pig.

Figure 8. Proton/base flux rates (C and F) and exemplary pH_i traces (A, B, D and E) in cultured parietal (A, B and C) and mucous cells (D, E and F) in the presence of CO₂-HCO₃⁻ and in the absence and presence of forskolin (A and D) and TPA (B and E), respectively, and [¹⁴C]aminopyrine uptake rates in parietal cells (G)

A, B, D and E show pH_i traces. pH_i recovery rates (indicated by dashed lines) and the intracellular buffering capacity (data not shown) at the appropriate pH_i were used to calculate the proton/base flux rates summarized in C and F. While forskolin (an agonist of cAMP-dependent protein kinase) had no influence on proton/base flux in both studied epithelial cell types, TPA (an agonist of PKC) stimulated proton/base flux 2-fold compared to control in mucous cells, but not in parietal cells (n = 4-8 in parietal cells, n = 5-10 in mucous cells, ** P < 0.01). G shows [¹⁴C]aminopyrine uptake rates in parietal cells expressed as the n-fold of the basal (non-treated) control. Forskolin and histamine caused a marked increase in [¹⁴C]aminopyrine uptake, whereas the values of TPA alone or in combination with histamine were not statistically different from the control (n = 4-5, *P < 0.05).