Immunolocalization of anion exchanger AE2, Na(+)/H(+) exchangers NHE1 and NHE4, and vacuolar type H(+)-ATPase in rat pancreas.

We have studied the expression and localization of several H(+) and HCO(3)(-) transporters, whose presence in the rat pancreas is still unclear. The Cl(-)/HCO(3)(-) exchanger AE2, the Na(+)/H(+) exchangers NHE1 and NHE4, and the 31-kD and 70-kD vacuolar H(+)-ATPase (V-ATPase) subunits were detected by immunoblotting and immunocytochemical techniques. Immunoblotting of plasma membranes with transporter-specific antibodies revealed protein bands at approximately 160 kD for AE2, at approximately 90 kD and approximately 103 kD for NHE1 and NHE4, respectively, and at 31 kD and 70 kD for V-ATPase. NHE1 and NHE4 were further identified by amplification of isoform-specific cDNA using RT-PCR. Immunohistochemistry revealed a basolateral location of AE2, NHE1, and NHE4 in acinar cells. In ducts, NHE1 and NHE4 were basolaterally located but no AE2 expression was detected. V-ATPase was detected in zymogen granules (ZGs) by immunogold labeling, and basolaterally in duct cells by immunohistochemistry. The data indicate that NHE1 and NHE4 are co-expressed in rat pancreatic acini and ducts. Basolateral acinar AE2 could contribute to Cl(-) uptake and/or pH regulation. V-ATPase may be involved in ZG fusion/exocytosis and ductal HCO(3)(-) secretion. The molecular identity of the ductal Cl(-)/HCO(3)(-) exchanger remains unclear.     

Identification of the full-length AE2 (AE2a) isoform as the Golgi-associated anion exchanger in fibroblasts.

Na(+)-independent Cl(-)/HCO(3)(-) exchangers (AE1, AE2, AE3) are generally known as ubiquitous, multispanning plasma membrane proteins that regulate intracellular pH and transepithelial acid-base balance in animal tissues. However, previous immunological evidence has suggested that anion exchanger (AE) proteins may also be present in intracellular membranes, including membranes of the Golgi complex and mitochondria. Here we provide several lines of evidence to show that an AE protein is indeed a resident of the Golgi membranes and that this protein corresponds to the full-length AE2a isoform in fibroblasts. First, both the N- and C-terminal antibodies to AE2 (but not to AE1) detected an AE protein in the Golgi membranes. Golgi localization of this AE2 antigen was evident also in cycloheximide-treated cells, indicating that it is a true Golgi-resident protein. Second, our Northern blotting and RT-PCR analyses demonstrated the presence of only the full-length AE2a mRNA in cells that show prominent Golgi staining with antibodies to AE2. Third, antisense oligonucleotides directed against the translational initiation site of the AE2a mRNA markedly inhibited the expression of the endogenous AE2 protein in the Golgi. Finally, transient expression of the GFP-tagged full-length AE2a protein resulted in predominant accumulation of the fusion protein in the Golgi membranes in COS-7 and CHO-K1 cells. Golgi localization of the AE2a probably involves its oligomerization and/or association with the recently identified Golgi membrane skeleton, because a substantial portion of both the endogenous AE2a and the GFP-tagged fusion protein resisted detergent extraction in cold. (J Histochem Cytochem 49:259-269, 2001)     

Immunolocalization of a 25-kilodalton protein in mouse testis and epididymis.

We have recently observed that a polyclonal antibody raised against a mouse epididymal luminal fluid protein (MEP 9) recognizes a 25-kDa antigen in mouse testis and epididymis [Rankin et al., Biol Reprod 1992; 46:747-766]. This antigen was localized by light and electron microscopic immunohistochemistry. The immunoreactivity in the testis was found in the residual cytoplasm of the elongated spermatids, in the residual bodies, and in the cytoplasmic droplets of spermatozoa. In the epididymis, the epithelial principal cells were stained from the distal caput to the distal cauda. Immunogold labeling in the principal cells showed diffuse distribution without preferential accumulation in either the endocytic or the secretory apparatus of the cells. In the epididymal lumen, the immunoreactivity was restricted to the sperm cytoplasmic droplets. No membrane-specific labeling was observed in luminal spermatozoa, cytoplasmic droplets, or isolated sperm plasma membranes. Three weeks after hemicastration or severance of the efferent ducts, a normal distribution of the immunoreactive sites was found in the epididymis. Immunoreactivity, was also detected in the epididymal epithelium of immature mice as well as in that of XXSxr male mice having no spermatozoa in the epididymis. These results suggest that the immunoreactivity seen in the principal cells originates from synthesis rather than endocytosis of the testicular protein from disrupted cytoplasmic droplets. Furthermore, these results suggest that the 25-kDa protein is synthesized independently by both testis and epididymis.     

Trapping of inhibitor-induced conformational changes in the erythrocyte membrane anion exchanger AE1.

AE1, the chloride/bicarbonate anion exchanger of the erythrocyte plasma membrane, is highly sensitive to inhibition by stilbene disulfonate compounds such as DIDS (4,4'-diisothiocyanostilbene-2, 2'-disulfonate) and DNDS (4,4'-dinitrostilbene-2,2'-disulfonate). Stilbene disulfonates recruit the anion binding site to an outward-facing conformation. We sought to identify the regions of AE1 that undergo conformational changes upon noncovalent binding of DNDS. Since conformational changes induced by stilbene disulfonate binding cause anion transport inhibition, identification of the DNDS binding regions may localize the substrate binding region of the protein. Cysteine residues were introduced into 27 sites in the extracellular loop regions of an otherwise cysteineless form of AE1, called AE1C(-). The ability to label these residues with biotin maleimide [3-(N-maleimidylpropionyl)biocytin] was then measured in the absence and presence of DNDS. DNDS reduced the ability to label residues in the regions around G565, S643-M663, and S731-S742. We interpret these regions either as (i) part of the DNDS binding site or (ii) distal to the binding site but undergoing a conformational change that sequesters the region from accessibility to biotin maleimide. DNDS alters the conformation of residues outside the plane of the bilayer since the S643-M663 region was previously shown to be extramembranous. Upon binding DNDS, AE1 undergoes conformational changes that can be detected in extracellular loops at least 20 residues away from the hydrophobic core of the lipid bilayer. We conclude that the TM7-10 region of AE1 is central to the stilbene disulfonate and substrate binding region of AE1.     

Abscisic acid influx into human nucleated cells occurs through the anion exchanger AE2

Abscisic acid (ABA) is a hormone conserved from cyanobacteria to higher plants, where it regulates responses to environmental stimuli. ABA also plays a role in mammalian physiology, pointedly in inflammatory responses and in glycemic control. As the animal ABA receptor is on the intracellular side of the plasma membrane, a transporter is required for the hormone’s action. Here we demonstrate that ABA transport in human nucleated cells occurs via the anion exchanger AE2. Together with the recent demonstration that ABA influx into human erythrocytes occurs via Band 3, this result identifies the AE family members as the mammalian ABA transporters.     

AE2 anion exchanger polypeptide is a homooligomer in pig gastric membranes: a chemical cross-linking study.

Although considerable information is available on the oligomeric states of the AE1 (band 3) anion exchanger, little is known about the physiological state of the polypeptides encoded by the nonerythroid AE genes, AE2 and AE3. We have previously characterized the proteolytic susceptibility of native pig gastric AE2. In the course of studies in which pig gastric membranes were treated with the AE2 transport antagonist, DIDS, we noted evidence for cross-linking of AE2 proteolytic fragments to higher-order oligomeric forms. We have characterized the ability of DIDS and of selected N-hydroxysuccinimide cross-linking agents to increase the proportion of SDS-resistant oligomers of pig gastric AE2 and its proteolytic fragments. Cross-linking exhibited time and concentration dependence. N-Terminal protein sequencing proved that DIDS treatment created AE2 homodimers. Putative homotetramers were also observed. Protomers were cross-linked via residues within the C-terminal 40 kDa of AE2. Prior proteolytic cleavage of AE2 in membranes resulted in decreased yield of subsequently cross-linked products. AE2 cross-linking could not be detected in membranes pretreated by hypotonic wash and freeze-thaw. The results are interpreted in light of the deduced amino acid sequence of the transmembrane domain of pig AE2.     

The AE2 anion exchanger is necessary for the structural integrity of the Golgi apparatus in mammalian cells

The structural integrity of the Golgi apparatus is known to be dependent on multiple factors, including the organizational status of microtubules, actin and the ankyrin/spectrin-based Golgi membrane skeleton, as well as vesicular trafficking and pH homeostasis. In this respect, our recently identified Golgi-associated anion exchanger, AE2, may also be of importance, since it potentially acts as a Golgi pH regulator and as a novel membrane anchor for the spectrin-based Golgi membrane skeleton. Here, we show that inhibition (>75%) of AE2 expression by antisense oligonucleotides in COS-7 cells results in the fragmentation of the juxtanuclear Golgi apparatus and in structural disorganization of the Golgi stacks, the cisternae becoming generally shorter, distorted, vesiculated and/or swollen. These structural changes occurred without apparent dissociation of the Golgi membrane skeletal protein Ankyrin(195), but were accompanied by the disappearance of the well-focused microtubule-organizing center (MTOC), suggesting the involvement of microtubule reorganization. Similar changes in Golgi structure and assembly of the MTOC were also observed upon transient overexpression of the EGFP-AE2 fusion protein. These data implicate a clear structural role for the AE2 protein in the Golgi and in its cytological positioning around the MTOC.     

COX-dependent and -independent pathways in bradykinin-induced anion secretion in rat epididymis

Lysylbradykinin (LBK) added to the apical or basolateral side of cultured rat epididymal monolayers stimulated a rise in short-circuit current (Isc) due to anion secretion. The concentration-response relationships for the apical and basolateral applications have EC50 value of 0.001 microM. The responses to apical or basolateral application of LBK were blocked by WIN64338, a specific B2 receptor antagonist, but not by Des-Arg9,[Leu8]-BK, a specific B1 receptor antagonist, indicating that the LBK effects were mediated through B2 bradykinin receptors. Experiments to desensitize the B2 receptors by repeated stimulation have demonstrated that the responses to apical or basolateral LBK were due to discrete receptors on the apical or basolateral surface. In epithelia clamped in the Ussing chambers, addition of LBK to the apical or basolateral surface evoked release of PGE2 into the apical and basolateral bathing solutions over the first 10 min following hormone addition. LBK added to the basolateral side elicited a greater release than it was added to the apical side. Pretreatment of the epithelia with piroxicam (5 microM) abolished PGE2 release elicited by apical or basolateral LBK and abrogated the Isc induced by basolateral LBK. However, the rise in Isc induced by apical LBK was reduced by 31.3% only. The anion secretion response to apical LBK was not affected by MDL-12330A, an adenylate cyclase inhibitor, but greatly attenuated by thapsigargin, an inhibitor of intracellular Ca2+ release. However, the reverse effects were seen for basolateral LBK. It is concluded that distinct pathways are involved in the stimulation of anion secretion by apical or basolateral LBK. The response to basolateral LBK was COX-dependent, mediated by PGE2 and involves cAMP as second messenger. In contrast, the response to apical LBK is largely COX-independent, not mediated by PCE2 and involves Ca2+ as intracellular messenger.