- (2001) Volume 2, Issue 4
| Seth L Alper*, Marina N Chernova, Andrew K Stewart Departments of Medicine and Cell Biology, Harvard Medical School; Molecular Medicine and Renal Units, Beth Israel Deaconess Medical Center. Boston, MA, USA |
| Corresponding Author Seth L Alper Department of Medicine (Cell Biology) Harvard Medical School Molecular Medicine and Renal Units RW763 Beth Israel Deaconess Med. Ctr. East Campus 330 Brookline Ave Boston, MA 02215 USA Phone +1-617.667.2930; +1-617.667.2918 Fax +1-617.667.8040; +1-617.667.2913 E-mail salper@caregroup.harvard.edu |
Among human bicarbonate transporters, two major gene families encode Na-independent Cl- /HCO3 - exchangers: the SLC4 anion exchanger (AE) family [1], and the SLC26 “sulfate permease” anion transporter family [2]. The SLC4 AE family contains at least three genes, and comprises a subfamily within the larger and phylogenetically more ancient bicarbonate transporter superfamily that includes the Na bicarbonate cotransporters (NBC) and the Na-driven Cl/base exchangers. Mutations in the human AE1 gene cause autosomal dominant spherocytic anemia and distal renal tubular acidosis of both dominant and recessive forms. Anemia is also associated with AE1 mutations in mouse, cow, and zebrafish. Naturally occuring mutations in the human AE2 and AE3 genes have not been detected. The SLC26 family in humans consists of at least 10 members, and includes anion exchangers which exchange chloride for bicarbonate, hydroxyl, sulfate, formate, iodide, and/or oxalate. Mutations in three of these genes cause hereditary disease, including chondrodysplasia (SLC26A2, DTD), diarrhea (A3, down-regulated in adenoma/chloride-losing diarrhea protein: DRA/CLD), and goiter/deafness syndrome (A4, pendrin). Little is known about the acute regulation of these modulators of intracellular and compartmental pH and volume.
Keywords |
| Acidosis, Renal Tubular; Band 3 Protein; Spherocytosis, Hereditary |
Abbreviations |
| AE: anion exchanger; CLD: chloride-losing diarrhea; DTD: diastrophic dysplasia; DIDS: 4,4'-di-isothiocyanostilbene-2,2'- disulfonic acid; DRA: down-regulated in adenoma; NBC: Na bicarbonate cotransporter; NHE: Na+/H+ exchanger; PDZ: postsynaptic density-95, discs large, zonula occuldens-1; SLC: HUGO designation for "solute carrier" gene family |
| We have studied the structural elements required for short-term regulation of the SLC4 Cl-/HCO3 - exchangers AE1, AE2, and AE3, and have initiated similar studies of SLC26A3, the downregulated in adenoma/chloride-losing diarrhea protein (DRA/CLD). We have detected regulation of some of these proteins by pH, tonicity, and NH4 +. Small regions or single amino acid residues crucial to these regulations have been delineated and, in the case of AE2, mapped onto the published X-ray structure of the AE1 N-terminal cytoplasmic domain. Whereas AE1 shows little sensitivity to pH over the physiological range, intracellular and extracellular acidification both inhibit AE2. The particular structural requirements of AE2 inhibition by protons depend on whether only pHi is varied, or pHi change is caused by an imposed change in pHo. |
| AE1 is expressed at highest levels in erythrocyte plasma membrane (eAE1) and in the basolateral plasma membrane of Type A intercalated cells of the renal cortical and medullary collecting ducts. kAE1 of intercalated cells lacks the N-terminal 65 (human) or 79 (mouse) residues of eAE1. AE1 is expressed at lower levels in heart, distal colon, and other tissues. AE2 is the most widely expressed form of AE, and in nearly all epithelial cells resides in the basolateral plasma membrane. A Golgi form of AE2 has been detected in situ and in transfected cells, and exhibits distinct detergent extraction characteristics [3, 4]. Mammalian AE2 variant transcripts encode 5 distinct N-terminal amino acid sequences of as yet unknown functions [5, 6]. AE3 is expressed at highest level in excitable tissues, throughout the GI tract, and at lower levels in other tissues [7]. Mammalian AE3 transcripts encode two N-terminal amino acid sequences (bAE3 and cAE3) [8]. In epithelial cells of the kidney, bAE3 is basolateral, whereas in epithelial cells of the gut, bAE3 is detectable on both basolateral and apical surfaces. In contrast, cAE3 in kidney and gut preferentially localizes to apical membranes, where it could potentially contribute to apical bicarbonate secretion [9]. Mammalian AE3 transcripts also encode two C-terminal amino acid sequences. One of these (AE3-14nt) is the only AE to lack a transmembrane anion transport domain [10]. Although AE3-14nt is relatively abundantly expressed as mRNA and protein, its function remains unknown. cAE3-14nt does not exhibit a dominant negative phenotype when co-expressed in Xenopus oocytes with cAE3 [Papageorgiou, Shmukler, Alper: unpublished data]. The recently reported AE4 [11] is more closely related in sequence to the NBC Na-bicarbonate cotransporters than to AE anion exchangers, although reported to mediate Na+-independent Cl- /HCO3 - exchange. |
| We have focused our studies on AE1 and AE2. These two anion exchangers exhibit similar maximal transport rates when expressed in Xenopus oocytes. Their anion selectivities and apparent affinities for extracellular anions are similar. AE2 is less sensitive than AE1 to inhibition by a wide range of drugs, with the notable exception (at least in Xenopus oocytes) of select polyaminosterol sulfates [12]. Notably, the AE2 anion exchanger expressed in Xenopus oocytes exhibits several modes of regulation not displayed by AE1. AE2 is strongly inhibited by protons, whereas AE1 is minimally regulated by protons within the nearphysiological range [13]. AE2 is activated by hypertonicity [14] and can mediate regulatory volume increase in oocytes [15], but AE1 lacks these properties. AE2 is paradoxically activated by low concentrations of ammonium ion despite ammonium-mediated acidification in Xenopus oocytes, whereas AE1 is unaffected by ammonium [16]. We have set out to define the structural elements of AE2 which mediate these isoformspecific modes of regulation. Although we have measured AE function by fluorometric indices of Cl-/HCO3 - [17] and Cl-/nitrate exchange in transfected mammalian cells and in Xenopus oocytes [unpublished data], we have performed most studies on regulation using unidirectional 36Clflux studies in Xenopus oocytes. These studies have been supplemented by fluorometric and microelectrode measurements of oocyte pHi. Initial studies were performed by varying bath pH (pHo) and allowing intracellular pH (pHi) to achieve steady-state before measuring flux [13, 18]. More recent studies utilize weak acids to vary pHi at constant pHo. Current efforts are standardizing conditions to vary pHo at constant pHi. |
| Study of chimeric proteins assigned distinct pHregulatory functions to the 705 aa N-terminal cytoplasmic domain and the 533 aa C-terminal transmembrane domain of AE2 [16]. When pHo was varied between pH 9 and pH 5 in these studies, pHi varied between pH 7.61 and 7.13, only 0.13 pHi unit per pHo unit. Across this range, AE2 was steeply regulated with a sigmoidal pattern characterized by a pHo(50) value of 6.93, whereas AE1 activity was not significantly altered. In these conditions, the AE2cyto/AE1memb chimera exhibited an AE1 phenotype. In contrast, the AE1cyto/AE2memb chimera exhibited an AE2 phenotype, modified only in its acid-shifted pHo(50) value of 6.3. Thus the sigmoidal pHresponse of AE2 activity required the transmembrane domain, and the AE2 cytoplasmic domain served to increase sensitivity to protons of the transmembrane domain. The transmembrane domain site was named a “sensor”, and the Nterminal cytoplasmic domain site was named a “modifier”, by analogy to similar studies with the NHE1 Na+/H+ exchanger [19]. |
| Further studies have focused on the molecular definition of the “modifier site” in the N-terminal cytoplasmic domain of AE2. Removal of 99 Nterminal amino acids had no effect, whereas removal of 510 N-terminal residues produced a pH vs. activity curve identical in shape and pHo(50) to that of the AE1cyto/AE2memb chimera. These studies, originally performed with 36Cl- influx experiments, have recently been reproduced with 36Cl- efflux experiments in which single oocytes were subjected to the full range of pHo values. Stepwise analysis of additional N-terminal truncations with the efflux assay further narrowed the required region to residues between aa 328 and 347, and between aa 391 and 510. A hexa- Ala bloc scan restricted to the former stretch revealed that, even within this small region, the residues required to exhibit the wildtype pHo(50) value are non-contiguous in the primary sequence. The above structure-function experiments could not reveal the role of pHi in AE2 regulation. Another approach to measurement of AE2 regulation by pH has used nigericin high K+ to clamp pHi nominally equal to pHo during measurement of Cl-/nitrate exchange [20]. However, these experiments were similarly unable to determine the exclusive role of pHi in AE regulation. Therefore, we standardized conditions of exposure to weak acids to vary pHi at constant pHo. In order to apply this method to study of AE function, we showed that weak acids were neither inhibitors of AE2-mediated 36Cl- transport, nor were they themselves substrates for AE2. With this procedure, we re-evaluated the regulation of Cl- /Cl- exchange activity by pHi at constant pHo. The AE2 regions required for regulation by pHi overlapped with, but were not identical to those required for regulation by pHo (with consequent variation of pHi). In particular, removal of the Nterminal 310 aa reduced regulation by pHi but had no effect on that by pHo. Removal of the Nterminal 328 or 391 aa, which retained a nearly normal regulation by pHo, abolished regulation by pHi. |
| To delineate further the regions required for AE2 regulation by pHi, we performed hexa-alanine bloc substitutions between residues 310 and 347. In contrast to the N-terminal truncation experiments, only Ala substitution for residues between 336-341 and 342-347 abolished regulation by pHi. Once more, the 6 amino acid blocks required for wildtype pHo(50) value were not identical to those required for regulation by pHi. This region is among the most highly conserved in the N-terminal cytoplasmic domain of the bicarbonate transporter superfamily. Within this region, point mutations have been identified which modify only AE2 regulation by pHi, or which modify regulation by both pHi and pHo. The region involved aligns with a segment of the AE1 N-terminal cytoplasmic domain X-ray structure predicted to abut or to reside nearby the endofacial surface of the AE transmembrane domain at the cytoplasmic surface of the plasma membrane [21]. In this position, the candidate modifier region would be well situated to interact directly with the AE2 pH sensor(s) in the membrane-spanning domain. Future studies will test the consequences to Cl-/HCO3 - exchange of mutants shown to modify regulation of Cl-/Clexchange by pHi and pHo. |
| We have initiated comparison of the anion exchange processes and mechanisms of the two distinct structural classes of anion exchanger, the band 3-related AE (SLC4) and the “sulfate permease” SLC26 family [22]. We have found that Cl-/Cl- exchange mediated by the Cl- /formate/OH-/HCO3 exchanger, pendrin [23, 24] is insensitive to bath pH change, and that mediated by the Cl-/HCO3 - exchanger DRA [25] is minimally sensitive to bath pH change. Whereas pendrin is also insensitive to intracellular acidification at constant pHo, DRA is inhibited by intracellular acidification and activated by intracellular alkalinization. DRA is also stimulated by ammonium and, less robustly, by hypertonicity. Although insensitive to the stilbene inhibitor DIDS, DRA is rather potently inhibited by the antiinflammatory drugs, tenidap and niflumate. The carboxy-terminal 9 aa of DRA (including a candidate PDZ domain recognition sequence) are dispensable for wildtype Cl-/Cl- exchange function in the Xenopus oocyte. However, truncation of four additional residues greatly reduces DRAmediated Cl- transport. |
| Comparison of mechanism, regulation, and structure-function relationships between SLC4 and SLC26 anion exchangers should enhance understanding of how two distinct sets of structures mediate very similar ion exchange reactions across lipid bilayers. The availability of microbial homologs of the SLC26 gene family may prove particularly helpful for structural study of this class of proteins. |
