|
|
|
|
|
|
|
||
Address correspondence to Seth L. Alper, RW763 East Campus, Beth Israel Deaconess Medical Center, 330 Brookline Ave., Boston, MA 02215. Fax: (617) 667-8040; E-mail: salper{at}caregroup.harvard.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
0.8 pH units in the mutant (A)6342–347, but only a modest acid-shift in the mutant (A)6336–341. Two of the six (A)6 mutants retained normal pHi sensitivity of 36Cl- efflux, whereas the (A)6 mutants 318–323, 336–341, and 342–347 were not stimulated by intracellular alkalinization. We further evaluated the highly conserved region between aa 336–347 by alanine scan and other mutagenesis of single residues. Significant changes in AE2 sensitivity to pHo and to pHi were found independently and in concert. The E346A mutation acid-shifted the pHo(50) value to the same extent whether pHi was unclamped or held constant during variation of pHo. Alanine substitution of the corresponding glutamate residues in the cytoplasmic domains of related AE anion exchanger polypeptides confirmed the general importance of these residues in regulation of anion exchange by pH. Conserved, individual amino acid residues of the AE2 cytoplasmic domain contribute to independent regulation of anion exchange activity by pHo as well as pHi.
Key Words: Cl-/HCO3- exchange • weak acids • Xenopus oocytes • isotopic flux • pH-sensitive microelectrodes
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Bevensee et al., 2000; Alper et al., 2002). All AE polypeptides share a highly conserved hydrophobic, polytopic, COOH-terminal transmembrane domain of >500 amino acids (aa),* with a short COOH-terminal cytoplasmic tail capable of binding carbonic anhydrase II (Vince and Reithmeier, 2000; Sterling et al., 2002b). This transmembrane domain is preceded by a less extensively conserved hydrophilic NH2-terminal cytoplasmic domain of 400–700 aa (Alper, 1994). The COOH-terminal transmembrane domain suffices to mediate anion exchange in the absence of nearly the entire NH2-terminal cytoplasmic domain (Grinstein et al., 1978; Kopito et al., 1989). Although the cytoplasmic NH2-terminal domain of AE1 is important for its binding to multiple cytoskeletal proteins, glycolytic enzymes, and hemoglobin (Zhang et al., 2000), the functions of the cytoplasmic NH2-terminal domains of AE2 and AE3 remain less extensively investigated.
The polypeptide products of the various AE genes differ in their acute regulation by pH. Native AE1-mediated Cl-/HCO3- exchange in erythrocytes (Funder and Wieth, 1976) and heterologous AE1–mediated Cl-/Cl-exchange in Xenopus oocytes (Humphreys et al., 1994; Zhang et al., 1996) displays a broad pH versus activity profile, consistent with its primary role in facilitating CO2/HCO3- exchange between the respiring tissues and lungs (Jennings, 1992). In contrast, nonerythroid Na+-independent Cl-/HCO3- exchange in many tissue culture cells is sensitively regulated by changes in pHi (Reinertsen et al., 1988), consistent with its proposed role in recovery from alkaline loads. Similarly, the recombinant nonerythroid anion exchanger AE2 is highly sensitive to changes in pHi when expressed in tissue culture cells (Lee et al., 1991; Jiang et al., 1994) or in Xenopus oocytes (Humphreys et al., 1994; Zhang et al., 1996; Stewart et al., 2001). In contrast, recombinant AE3 expressed in 293 cells has been reported to be insensitive to changes in pHi (Sterling and Casey, 1999).
Zhang et al. (1996) compared regulation by pH of anion influx mediated by AE1 and AE2 in Xenopus oocytes, and localized to the AE2 transmembrane domain a "pH sensor site" which conferred increased pH-sensitivity of anion transport. These structure-function studies demonstrated that the extracellular proton sensitivity (pHo(50) value) of AE2-mediated 36Cl- influx (7.0 under conditions of unclamped pHi) was acid-shifted 0.7 pH units by truncation of the cytoplasmic NH2-terminal 510 aa. This result suggested a role for the cytoplasmic NH2-terminal domain in regulation of AE2 transport by changing pH, and localized between aa 99 and 510 of the 705 aa AE2 NH2-terminal cytoplasmic domain, a "pH modifier site" that modulated the pHo(50) value of the transmembrane domain.
More recently, we demonstrated that variation of pHo regulates AE2 activity in efflux assays and requires the integrity of two noncontiguous regions in the NH2-terminal cytoplasmic domain, encompassing aa 328–347 and aa 391–510 (Stewart et al., 2001). We further showed by varying pHi at constant pHo that removal of the NH2-terminal 310 aa greatly reduced AE2 regulation by pHi, but was without effect on AE2 regulation by varying pHo. In the present work we have defined individual amino acid residues within the NH2-terminal cytoplasmic domain whose mutation alters the regulation of AE2-mediated Cl- transport activity by changing pHo with constant or minimally changing pHi, and by varying pHi at constant pHo. We demonstrate that a delimited, highly conserved region of the AE2 NH2-terminal cytoplasmic domain is involved in AE2 regulation by both pHi and pHo. We further demonstrate that some of these conserved residues in corresponding regions of related anion exchanger polypeptides contribute similarly to their functional regulation by pHi.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Solutions
ND-96 medium consisted of (in mM): 96 NaCl, 2 KCl, 1.8 CaCl2, 1 MgCl2, 5 HEPES, and 2.5 sodium pyruvate, pH 7.40. Flux media lacked sodium pyruvate. pH values of 7.0, 8.0, and 8.5 in room air flux media were achieved with 5 mM HEPES. 5 mM Mes was used was used for room air flux media of pH values 5.0 and 6.0. In Cl--free solutions, NaCl was replaced isosmotically with 96 mM sodium isethionate, and equimolar K, Ca, and Mg gluconate were substituted for the corresponding Cl- salts. CO2/HCO3--buffered solutions of pH 7.4 were saturated with 5% CO2–95% air at room temperature for 1 h and differed from Cl- free ND-96 in replacement of 24 mM sodium isethionate with 24 mM NaHCO3-. pH of CO2/HCO3--buffered solutions was verified before each experiment. Addition to flux media of the weak acid salt sodium butyrate was in equimolar substitution for NaCl.
Mutant AE2 cDNAs
Murine AE2 encoded in plasmid p
X (Zhang et al., 1996) was used as template for PCR. The AE2 hexa-Ala bloc substitution mutants (A)6312–317, (A)6318–323, (A)6324–329, (A)6330–335, (A)6336–341, and (A)6342–347 were constructed by a four primer PCR method as described (Zhang et al., 1996; Chernova et al., 1997a,b; Stewart et al., 2001). Single point mutations were constructed by the same method to generate single residue missense mutants. Oligonucleotide primers were obtained from Biosynthesis; primer sequences are available upon request. Integrity of PCR products and ligation junctions was confirmed by DNA sequencing of both strands.
cRNA Expression in Xenopus Oocytes
Mature female Xenopus (NASCO) were maintained and subjected to partial ovariectomy as described (Humphreys et al., 1994). Stage V-VI oocytes were manually defolliculated following incubation of ovarian fragments with 2 mg/ml collagenase A (Boehringer) for 60 min in ND-96 solution containing 50 ng/ml gentamycin and 2.5 mM sodium pyruvate. Oocytes were injected on the same day with cRNA or with 50 nl H2O. Capped cRNA was transcribed from linearized cDNA templates with the T7 MEGAscript kit (Ambion), and resuspended in diethylpyrocarbonate-treated water. RNA integrity was confirmed by agarose gel electrophoresis in formaldehyde, and RNA concentration was estimated by A260. Injected oocytes were then maintained for 2–6 d at 19°C.
36Cl- Efflux Measurements
Individual oocytes in Cl--free ND-96 were injected with 50 nl of 130 mM Na36Cl (10,000–12,000 cpm). After a 5–10 min recovery period, the efflux assay was initiated by transfer of individual oocytes to 6 ml borosilicate glass tubes, each containing 1 ml efflux solution. At intervals of 3 min, 0.95 ml of this efflux solution was removed for scintillation counting and replaced with an equal volume of fresh efflux solution. After completion of the assay with a final efflux period in the presence of the anion transport inhibitor 4,4'-di-isothiocyanatostilbene-2,2' disulfonic acid (DIDS; 200 µM), each oocyte was lysed in 100 µl of 2% SDS. Samples were counted for 3–5 min such that the magnitude of 2SD was <5% of the sample mean.
Experimental data were plotted as ln (percentage cpm remaining in the oocyte) versus time. 36Cl- efflux rate constants were measured from linear fits to data from the last three time points sampled for each experimental condition. All single time point values for 36Cl- efflux from AE2 cRNA–injected oocytes into chloride medium exceeded 150 cpm. Efflux cpm values for water-injected oocytes (40–90 cpm) were indistinguishable from those for AE2 cRNA–injected oocytes in the presence of DIDS, and both less than threefold higher than machine background values (typically 20 cpm, peak 30 cpm). Within each experiment, water-injected and AE2 cRNA–injected oocytes from the same frog were subjected to parallel measurements. On each experimental day, activity of tested mutant AE2 polypeptides was compared with wt AE2 activity at pH 7.4. Each AE2 mutant was tested in oocytes from at least three frogs.
Individual oocytes maintained in Cl--free solution at pHo 7.4 were exposed sequentially to (Cl--containing) ND-96 at pH 5.0, 6.0, 7.0, 8.0, and 8.5, then to solution of pH 8.5 in the presence of DIDS. During this variation of bath pH between pHo 5 and 8.5, oocyte pHi varied between extreme values of 7.1 and 7.6, but usually to a lower degree (Zhang et al., 1996; Stewart et al., 2001). Rate constants measured at each pHo value for wt AE2 and for the tested AE2 mutants in each individual experiment were fit (Sigma Plot) to the following first-order logistic sigmoid equation:
 | (1) |
) and by comparison for all pairs using Tukey-Kramer analysis (JMP for Macintosh). pHo dependence of wild-type AE2-mediated 36Cl- efflux did not differ when the experiment was performed with the order of pHo change reversed (Stewart et al., 2001).
|
6 min (Stewart et al., 2001). Neither the addition of butyrate nor its subsequent removal led to changes in the intracellular-free Ca2+ concentration (unpublished data) as measured by fluorimetric ratio imaging in FURA2-AM–loaded oocytes (Vandorpe et al., 2001). pHi sensitivity of AE2-mediated 36Cl- efflux was expressed for individual oocytes as fold stimulation of AE2 activity after weak acid removal (rate constant after removal of weak acid/rate constant in presence of weak acid). We have shown previously that butyrate is neither an inhibitor of nor a substrate for AE2 (Stewart et al., 2001). To maintain constant pHi during variation of pHo concentrations of extracellular butyrate were chosen for each pHo value according to the Henderson-Hasselbalch equation to yield a uniform total intracellular butyrate concentration of 0.5 mM at steady-state pHi. pHi was assumed to be 7.0 when pHo was 7.4, and the pKa of butyrate was taken as 4.82. Sodium butyrate was substituted isosmotically with NaCl. Bath butyrate concentrations were 0.005 mM at bath pH 5.0, 0.05 mM at bath pH 6.0, 0.5 mM at pH 7.0, 5 mM at pH 8.0, and 20 mM at pH 8.5.
Measurement of Oocyte pHi
Oocyte pHi was measured during bath superfusion using pH microelectrodes as described previously (Romero et al., 1997; Stewart et al., 2001). Oocyte pHi was measured during bath superfusion using BCECF ratio fluorimetry, also as described previously (Zhang et al., 1996).
Statistics
Analysis of variance followed by Tukey-Kramer analysis was performed among groups of mutants compared with wild-type AE2. Positive values indicated means that were significantly different. Statistical significance was also assessed by Student's paired and unpaired t tests, for which level of significance was P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Our earlier studies had demonstrated the high sensitivity of AE2 function to changes in pHo (Zhang et al., 1996; Stewart et al., 2001), but under conditions in which near-steady-state pHi was unclamped and changed in parallel with pHo. Although the range of pHo change in these experiments far exceeded the range of unclamped pHi change, the ability of changing pHo alone to regulate AE2 function at constant pHi remained in question. We therefore devised conditions in which variation of pHo in tandem with variation of bath butyrate concentration might effectively clamp pHi during changes in pHo. Fig. 1 A depicts microelectrode measurements of changing pHi in a representative AE2-expressing oocyte subjected to sequential 15 min exposures to varying pHo, matching the 36Cl- efflux protocol. Exposure of a different AE2-expressing oocyte (Fig. 1 B) to the same pattern of pHo change while simultaneously varying bath butyrate concentration almost completely damped the corresponding changes in pHi.1 Fig. 1 C summarizes the pHi versus pHo relationship in the absence (top slope = 0.02) and presence of adjusted butyrate concentrations (bottom slope = 0.002). Only in the absence of changing [butyrate] did oocyte pHi change during an increase in pHo from 5.0 to 8.5 (P = 0.03, Student's unpaired t test). Parallel measurement of oocyte pHi by BCECF ratio fluorimetry in the absence (n = 4–8) or presence of step changes in graded extracellular [butyrate] (n = 6–12) similarly showed that butyrate damped pHi change in response to changing pHo (unpublished data). Thus, simultaneous variation of bath butyrate concentration and pHo allowed evaluation of AE2 regulation by changing pHo while pHi remained nearly constant (nominal "pHi clamp").
|
Regulation of AE2 NH2-terminal Hexa-alanine Substitution Mutants by Varying Bath pH
We have shown previously that the region encompassing amino acids 310–347 within the NH2-terminal cytoplasmic domain of the murine AE2 anion exchanger is required for normal regulation of AE2-mediated Cl- transport by pHo and by pHi (Stewart et al., 2001). We studied this region in greater detail through the systematic substitution of hexa-alanine blocs into consecutive six-amino acid stretches of AE2, as illustrated in the schematic of Fig. 2
A. Fig. 2 B presents the 36Cl- efflux rate constants measured at pHo 7.4 exhibited by oocytes expressing AE2 hexa-alanine bloc substitutions. With the sole exception of (A)6318–323, all mutants exhibited 36Cl- efflux activity statistically indistinguishable from that of wild-type AE2. The low 36Cl- efflux activity of the (A)6318–323 mutant at pH 7.4 precluded subsequent evaluation of its regulation by pHo (Fig. 2 E), but sufficed to evaluate its regulation by pHi (Fig. 2 G).
To define those portions of the region encompassing residues 312–347 with importance for AE2 regulation by pHo, we measured 36Cl- efflux activity of the hexa-Ala bloc mutants under conditions of varying pHo at unclamped pHi. Fig. 2 C shows a representative 36Cl- efflux trace from three AE2-expressing oocytes and from one water-injected oocyte as a function of sequentially increasing pHo. Wild-type AE2–mediated 36Cl- efflux was minimal at low pHo, increased at higher pHo values, and was inhibited by 200 µM DIDS added at the final pHo of 8.5. Whereas AE2 (A)6330–335 exhibited a pattern of 36Cl- efflux similar to that of wild-type AE2, (A)6342–347 showed reduced inhibition of 36Cl- efflux activity at lower pHo compared with wild-type AE2. Fig. 2 D profiles 36Cl- efflux activity (normalized as described in MATERIALS AND METHODS) as a function of pHo for wild-type AE2 and two representative hexa-Ala bloc mutants, (A)6330–335 and (A)6342–347. The pHo value at which the rate constant for wild-type AE2–mediated 36Cl- efflux was half-maximal [pHo(50)] was 6.87 ± 0.05 (n = 36; Fig. 2, D and E). The hexa-Ala bloc mutant (A)6330–335 exhibited pHo dependence indistinguishable from that of wild-type AE2, whereas the pHo(50) value of the AE2 mutant (A)6342–347 was shifted to a more acidic pH value of 6.11 ± 0.11 (n = 11, P < 0.05).2 Fig. 2 E summarizes the pHo(50) values measured as shown in Fig. 2 D for all hexa-Ala bloc mutants. The data suggest that amino acid residues 342–347 are particularly important for regulation of AE2 by pHo. The hexa-Ala bloc mutant (A)6336–341 displayed a pHo(50) value only marginally acid shifted from that of wild-type AE2, as assessed by all pairs Tukey-Kramer analysis. Nonetheless, the high degree of conservation among these amino acid residues (see below) warranted their inclusion among those investigated in greater detail.
Regulation of AE2 NH2-terminal Hexa-alanine Substitution Mutants by Varying pHi at Constant pHo
We next investigated regulation of the same set of AE2 mutants by changing pHi at constant pHo, using the weak acid butyrate as described previously (Stewart et al., 2001). Fig. 2 F shows a representative efflux trace in which 36Cl- efflux activities of wild-type AE2 and of AE2 (A)6330–335 are reduced at low pHi and subsequently stimulated when pHi is elevated (by butyrate removal). In contrast, AE2 (A)6342–347-mediated 36Cl- efflux activity was insensitive to changes in pHi imposed by butyrate addition and removal. Fig. 2 G summarizes similar experiments for all hexa-Ala bloc mutants, expressed as fold stimulation of 36Cl- efflux rate constant after butyrate removal. The data reveal the importance of two discrete regions encompassing aa 318–323 and 336–347 for wild-type regulation of AE2-mediated Cl- transport by varying pHi. Hexa-Ala substitutions in these regions reduced AE2 stimulation by butyrate removal from the wild-type 10-fold to below 1.5-fold (P < 0.002, Student's unpaired t test). In contrast, the mutants (A)6324–329 and (A)6330–335 retained wild-type stimulation. The (A)6312–317 mutant retained significant, though reduced, sensitivity to pHi elevation, showing 3.5-fold stimulation by butyrate removal (P < 0.002).
Determination of Individual Amino Acid Residues Important for AE2 Regulation by pHo and pHi
The hexa-Ala bloc substitution mutants implicated AE2 aa 336–347 as important for regulation of AE2 activity by pH. To define the individual residues required for this regulation, we tested the functional properties of point mutants created as part of a modified alanine scan of this region (Fig. 3
A). Threonine 339, a weak candidate PKC site, was mutated to glutamate and (not depicted) to valine. Alanine 340 was mutated to glycine. Fig. 3 B shows that none of the mutants exhibited significant reduction of 36Cl- efflux activity at pHo 7.4. Fig. 3 C compares representative pHo versus Cl- efflux activity profiles of three AE2 point mutants with that of wild-type AE2. AE2 R337A exhibited a pHo dependence indistinguishable from that of wild-type AE2, whereas the mutants E346A and E347A displayed more acidic pHo(50) values of 6.10 ± 0.07 (n = 14, P < 0.05) and 6.20 ± 0.05 (n = 11, P < 0.05), respectively. Fig. 3 D summarizes the pHo(50) values for all point mutants measured as in Fig. 2 C. The gray bars indicate the four residues (W336, R341, E346, and E347) for which alanine substitution led to significant acid-shifts in pHo(50) value relative to that of wild-type AE2. Alanine substitution of other residues did not significantly shift pHo(50) values.
|
pHi following butyrate removal was 0.56 ± 0.04 pH units with a time constant (t0.5) of 322 ± 92s (n = 4). Thus, there was a lag time between t = 9 min and t = 12–15 min, during which 36Cl- efflux accelerated to its new steady-state. Rate constants for the new steady-state were calculated between t = 18 min and 24 min (Fig. 3 E). 36Cl- efflux activity of the AE2 mutants R337A and E346A exhibited pHi sensitivity identical to that of wild-type AE2, whereas 36Cl- efflux mediated by the AE2 mutant E347A was insensitive to this change in pHi. Fig. 3 F summarizes the pHi sensitivity of the AE2 point mutants compared with wild-type AE2. Stimulation of 36Cl- efflux by butyrate removal was indistinguishable from wild-type level (black bars) in 4 of the 12 mutants, and was only slightly reduced for K344A. In contrast, 7 of the 12 alanine substitution mutants exhibited severely attenuated stimulation by butyrate removal (gray bars for W336, E338, R341, W342, I343, F345, and E347) (P < 0.05, Student's unpaired t test). The AE2 mutants W336A, R341A, and E347A were concordant in their mutant phenotypes of regulation by pHi and by pHo, whereas mutants R337A, T339E, T339V (not depicted), A340G, and K344A were concordant in their wild-type or near wild-type phenotypes. Four AE2 mutants, E338A, I343A, F345A, and W342A, were insensitive to changing pHi while maintaining wild-type or near wild-type pHo sensitivity (Fig. 3, D and F). AE2 E346A was the only mutant that displayed acid-shifted pHo sensitivity (Figs. 3 C and 2 D) together with wild-type pHi sensitivity (Fig. 3 E). Alanine substitution of the neighboring residue E347 yielded both acid-shifted pHo sensitivity and loss of pHi sensitivity. These data define several amino acid residues whose mutation can alter the wild-type regulation of AE2 activity by pH. The differential consequences for regulation by pHo and by pHi based on the point mutations studied further supports the hypothesis that AE2 activity is regulated independently by extracellular and by intracellular protons.
The Effects of Missense Mutations on AE2 Sensitivity to Changing pHo Are Not Altered Under Conditions of pHi Clamp
We compared the effect of the mutation E346A on pHo dependence of AE2 function in conditions of "unclamped" and "clamped" pHi. The pHo versus Cl- efflux activity profile in the absence of butyrate (Fig. 4
A) revealed respective pHo(50) values for wild-type and mutant AE2 of 6.76 ± 0.05 (n = 18) and 6.10 ± 0.07, n = 14, P <0.05), similar to values presented in Fig. 3 E. Fig. 4 B show that this difference between the pHo sensitivity of mutant and wild-type AE2 did not change when pHo was varied during conditions of "pHi clamp." In these conditions, the difference between the pHo(50) values of wild-type AE2 (6.97 ± 0.06, n = 7), and of AE2 E346A (6.29 ± 0.13, n = 3) was maintained (P < 0.05). Fig. 4, C and D, summarize the pHo sensitivity data for wild-type AE2 and E346A and further confirm that AE2-mediated Cl- transport can be regulated by changing pHo when pHi remains nearly constant. Moreover, at least one residue of the AE2 NH2-terminal intracellular domain among the several important for AE2 regulation by changing pHo is also important for regulation by changing pHo at near constant pHi.
|
|
|
). The acute regulation by pH of most of these anion transporters has yet to be studied. However, the evident sequence conservation suggests that these residues may make similar contributions to regulation of these AE2-related anion transporters. We tested this hypothesis for regulation by pHi of the Na+-independent anion exchangers, AE3 and AE1.
|
), and that such regulation requires integrity of residues corresponding to those required by AE2.
|
) but also by changing pHi (Fig. 9, B and C)
. In contrast, the converse chimera AE2cyto/AE1memb exhibited an AE1-like phenotype, whether in response to changing pHo (Zhang et al., 1996) or to intracellular alkalinization by butyrate removal (Fig. 9 C). The NH2-terminal AE2 truncation mutant N659 displayed similar lack of stimulation by intracellular alkalinization. Thus, stimulation by intracellular alkalinization requires defined residues from the NH2-terminal cytoplasmic domain in combination with yet-to-be defined portions of the AE2 transmembrane domain. The pHi sensitivity of the AE1cyto/AE2memb chimera thus allowed experimental test of the potential importance of the AE1 region depicted in Fig. 7 when present in the "permissive environment" of proximity to the AE2 transmembrane domain (Fig. 9 A). Fig. 9, B and C, show that introduction into the AE1cyto/AE2memb chimera of the double Ala substitution mutation E99A/E100A severely attenuated stimulation of Cl- efflux by intracellular alkalinization (n = 14, P < 0.05).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Milanick and Gunn, 1984). However, the mechanisms by which pH acutely regulates the activities of pH-regulatory ion exchangers and cotransporters remain for the most part obscure. Even regulation by pHi of the Na+/H+ exchanger, NHE1, perhaps the most extensively studied among these transporters, is not yet understood at the level of individual amino acid residues. Similarly, the molecular mechanisms by which protons regulate activity of Cl-/HCO3- exchangers of the SLC4 gene family have until recently remained uninvestigated.
The current work has shown that AE2-mediated Cl- transport is sensitive to changing pHo under conditions of constant pHi (Fig. 1), complementing our earlier demonstration that AE2 activity is also sensitive to changing pHi at constant pHo (Stewart et al., 2001). We have found two stretches of amino acid sequence in the middle of the AE2 NH2-terminal cytoplasmic domain (aa 318–323 and aa 336–347) that are required for wild-type pH sensitivity of AE2-mediated anion exchange in Xenopus oocytes (Fig. 2). These two regions are present in all five NH2-terminal variant AE2 polypeptide products arising from alternate promoter usage or splicing of the AE2 gene (Stuart-Tilley et al., 1998; Medina et al., 2000). Alanine scan mutagenesis of aa 336–347 has identified individual amino acid residues whose mutation modulates AE2 regulation by pHo or abolishes AE2 regulation by pHi under the conditions studied (Fig. 3). The AE2 amino acid residues required for wild-type regulation by pHo and pHi are not identical (Fig. 3). Although mutations in some residues altered both modes of regulation, mutations in other residues altered selectively either the response to changing pHi or the response to changing pHo (Figs. 3–5). At least one structural requirement for regulation of AE2-mediated Cl-/Cl- exchange by pHi applies equally to pHi regulation of Cl-/HCO3- exchange (Fig. 6). Regulation by pHi of Cl-/Cl- exchange mediated by the related polypeptide cAE3 (Fig. 8) and by the AE1cyto/AE2memb chimera (Fig. 9) was severely attenuated by mutations corresponding to those which alter AE2 regulation.
Defined individual amino acid residues contribute to the independent regulation of AE2-mediated Cl- transport by pHo and by pHi. Earlier studies with large deletions and chimeras (Zhang et al., 1996) and with smaller NH2-terminal and internal deletions (Stewart et al., 2001) focused attention upon AE2 residues 312–347 as a region required for wild-type AE2 regulation by independent variation of pHo and of pHi. Subsequent hexa-Ala bloc substitution mutagenesis (Fig. 1) within this NH2-terminal region focused attention upon residues 336–347. Systematic alanine scan mutagenesis of this region (Fig. 3) then identified several individual amino acid residues important for AE2 regulation by pHo only, by pHi only, or for both. Individual mutation to alanine of the four residues W336, R341, E346, and E347 significantly shifted the AE2 pHo(50) to more acidic values. Uniquely among these four mutants, AE2 E346A retained wild-type pHi sensitivity. However, individual mutation to alanine of the seven residues W336, E338, R341, W342, I343, F345, and E347 eliminated AE2 regulation by pHi in the butyrate removal assay. Among these, E338A, W342A, I343A, and F345A retained wild-type or near wild-type pHo sensitivity. These distinct effects of individual mutations upon regulation of AE2 by pHi and by pHo reinforce the hypothesis that AE2 independently senses and responds to protons on both sides of the plasma membrane.
Contribution of Cytoplasmic Amino Acid Residues to AE2 Regulation by pHo
Regulation by pHi of AE2-mediated anion transport, itself a pH-regulatory process in the physiological presence of CO2/bicarbonate buffers, is intuitively reasonable across the normal range of pHi. The NH2-terminal cytoplasmic domain of AE2 is appropriately situated to contribute to such regulation. AE2's function as an acid loader in physiological conditions suggests a rationale for inhibition of AE2-mediated anion transport not only by reduced pHi, but also by reduced pHo. AE2 itself mediates or contributes to the intracellular acidification produced by extracellular acidification. Regulation of AE2 by pHo would allow the cell to down-regulate acid loading when first faced with extracellular acidosis, preferably before the cell has begun to experience adverse effects of an intracellular acid load. Nonetheless, an important role for NH2-terminal cytoplasmic domain residues is less easily envisioned for AE2's response to changing pHo than to changing pHi.
pHo regulates AE2-mediated Cl- transport across a physiological range with a pHo(50) value of 6.9, in contrast to the AE1 pHo(50) values of <5.0 (Zhang et al., 1996) or 5.8 (as measured by a different method by Muller-Berger et al., 1995). As we show here for AE2, Muller-Berger et al. (1995) also found that AE1-mediated Cl- transport activity is inhibited by lowering pHi at constant pHo, or by lowering pHo at near-constant pHi (although at much lower pH values than for AE2).3 Defined residues of the NH2-terminal cytoplasmic domain are absolutely required for inhibition of AE2-mediated Cl- transport by acidic pHi as assessed by the butyrate removal assay. These residues must either interact with or be required to maintain integrity of the pH sensor of the AE2 transmembrane domain (Zhang et al., 1996). The postulated interaction with or conformational change in the transmembrane sensor (resulting from either mutation or deletion of these criti-cal residues of the NH2-terminal cytoplasmic domain) must sufficiently reduce the effective pK of sensor residues to render it insensitive to the 0.5 pH unit intracellular acidification produced by addition of 40 mM butyrate to a pHo 7.4 bath (Stewart et al., 2001). Among the candidate residues contributing to this transmembrane pH sensor might be E1007 (Sekler et al., 1995), the residue corresponding in AE1 to human E681 and mouse E699. In AE1, mutation of this glutamate to glutamine (Chernova et al., 1997b) or its chemical modification to hydroxynorvaline with Woodward's reagent K (Jennings, 1995) not only alters anion selectivity, but also converts H+-sulfate cotransport into proton-independent sulfate transport. This observation has led to the hypothesis that this glutamate at or near the cytoplasmic face of putative transmembrane span 8 is or contributes to the proton binding site. Such direct, pH-dependent interactions (with their presumed conformational changes) have been demonstrated in K+ channels, where it is suggested that at acidic pH, specific residues are protonated in the COOH-terminal cytoplasmic tail, leading to interaction with the NH2-terminal tail and in turn causing conformational changes in the protein (Qu et al., 2000; for review see Jiang et al., 2002).
Regulation of AE2 by pHo involves protonation of exofacial residues, likely distinct from external Cl- binding site(s) (Milanick and Gunn, 1984). These protonation(s) could alter Cl- binding affinity or Cl- translocation rate, a function of the conformational change between inward-facing and outward-facing forms of the transporter. Substitution of the entire AE2 NH2-terminal cytoplasmic domain with that of AE1 (Zhang et al., 1996; Stewart et al., 2001), or deletion or mutation of those NH2-terminal cytoplasmic domain residues essential for wild-type AE2 response to changing pHo (including many residues not subject to titration by protons near the physiological range) shift the wild-type AE2 pHo(50) to more acidic values by 0.5–0.8 pH units. This shift, representing altered proton affinity of the external pH sensor (Gunn et al., 1973), might be accomplished through loss of its direct interaction with the cytoplasmic face of the AE2 transmembrane domain, resulting in altered conformation (and pKa) of protonatable residues at AE2's extracellular face or within the aqueous vestibule thought to have E1007 at its base. Alternately, a distinct polypeptide may mediate this interaction. The effects of AE2 mutations in the mutations in the NH2-terminal cytoplasmic domain upon AE2 pHo sensitivity resemble the alterations in ligand binding affinity of transmembrane hormone receptors secondary to mutations in receptor cytoplasmic loops or tail, or to receptor interaction with cytoplasmic regulatory polypeptide(s).
Putative Structure of the pH-regulatory Region of the AE2 NH2-terminal Cytoplasmic Domain
AE2 polypeptide has been purified from porcine stomach (Zolotarev et al., 1996, 1999), and is present in gastric membranes as a dimer or higher order oligomer (Zolotarev et al., 1999), but its structure remains unknown. We have mapped the AE2 sequences implicated in regulation of AE2 by pH upon the X-ray structure of crystallized erythroid AE1 NH2-terminal cytoplasmic domain reported by Zhang et al. (2000). This alignment predicts that AE2 aa 336–347 occupies a region corresponding to most of the AE1 structure's ß strands ß2 and ß3 along with their short connecting loop. This region is adjacent to AE1 residue 356, the COOH-terminal extent of structured sequence. Although structure remains uncertain for the disordered aa 357–379 in the AE1 NH2-terminal cytoplasmic domain and the 20 subsequent amino acids that link the cytoplasmic domain to the first transmembrane span of AE1, the loop connecting ß strands ß2 and ß3 is plausibly oriented close to the cytoplasmic face of the AE1 transmembrane domain (Zhang et al., 2000). We postulate that AE2 aa 336–347, although hundreds of residues away (in linear sequence) from the first transmembrane domain of AE2, may adopt a similar conformation with respect to elements of the internal pH sensor proposed to reside in the AE2 transmembrane domain. The pattern of altered pHi-regulatory phenotype produced by the alanine scan of AE2 aa 336–347 is consistent with a structure of two short ß sheets connected by a loop, showing one short patch of alternating phenotype within each predicted ß-sheet. This pattern is distinct from that of altered pHo sensitivity across the same region.
The decreased pHi sensitivity exhibited by the (ß3 strand) double mutant E99A/E100A of the chimeric AE1cyto/AE2memb transporter (Fig. 8) further supports this hypothesis. The absence of pHi dependence for AE1-mediated Cl- efflux in the butyrate removal assay highlights the functional interaction between pH-regulatory residues in NH2-terminal cytoplasmic domain and the yet undefined pH sensor structure present in the AE2 transmembrane domain, but lacking in that of AE1.
The different effects on AE2 regulation by pHi of E346 substitution by alanine or by aspartate suggest that in this position side chain packing may be more important than charge to the structure of a functional regulatory domain. A similar result has been presented for the electrogenic SGLT1 Na+-glucose cotransporter (Quick et al., 2001). The importance of packing in the folding of this regulatory domain is further suggested by the modified pH regulation exhibited by alanine substitution of conserved tryptophan residues. In contrast, substitution of the adjacent residue E347 with Ala or with Asp each led to altered regulation by pHi (Figs. 3 F and 5 C). These two adjacent glutamates likely comprise part of a ß-sheet structure, projecting their side chains in opposite directions (Zhang et al., 2000). Thus, the different consequences of mutagenesis of these adjacent residues likely reflects their different nearest neighbors.
Important roles for glutamate residues in conferring sensitivity to changing pHo and pHi have been demonstrated previously in potassium channels (Xu et al., 2000), chloride channels (Stroffekova et al., 1998), and the capsaicin receptor (Jordt et al., 2000). Although histidine residues play major roles in regulation of several K+ channels by pHi (Coulter et al., 1995; Doi et al., 1996; Chanhevalap et al., 2000; Qu et al., 2000; Xu et al., 2000), alanine substitution of AE2 His residues 314 and 317, either individually or in tandem, was without effect on AE2 regulation by pHi (not depicted). The lack of effect of T339 mutation to Val or to Glu similarly failed to support a role for T339 phosphorylation in AE2 regulation by pH.
Conserved Amino Acids of the NH2-terminal Cytoplasmic Domain Contribute to the pHi Regulation of other Anion Exchangers
The region of aa 336–347 is among the most highly conserved of the NH2-terminal cytoplasmic domains of SLC4 gene family members. Its functional importance was confirmed in the AE2 homologue cAE3 by mutating the two glutamate residues in cAE3 corresponding to E346 and E347 of AE2. Both mutations greatly attenuated the pHi sensitivity of cAE3-mediated Cl- efflux from Xenopus oocytes. However, whereas the AE2 mutant E346A exhibited a wild-type response to butyrate removal, the corresponding mutation in cAE3 (E150) showed considerably reduced pHi sensitivity (only twofold increase upon butyrate removal) compared with the fivefold increase exhibited by wild-type cAE3. This difference between the functional consequences of mutation within this highly conserved region in AE2 and AE3 may reflect sequence differences in immediately adjacent flanking regions, but very likely reflects different effects on interacting structures. These binding sites may be at the cytoplasmic face of the AE2 and AE3 transmembrane domains, or may represent other polypeptides.
These results are consistent with the initial report of pHi-sensitive Cl-/HCO3- exchange in 293 cells transiently transfected with AE3 (Lee et al., 1991), but contrast with the pH-insensitive AE3-mediated Cl-/nitrate exchange in 293 cells reported by Sterling and Casey (1999). The latter measurements were made in high K+/nigericin pH-clamp conditions in which pHi = pHo. The activation of AE3 by intracellular alkalinization in Xenopus oocytes is consistent with its hypothesized role in cardiac myocytes in recovery from alkaline load (Leem et al., 1999). As anion exchange in guinea pig cardiomyocytes is oppositely regulated by pHi and by pHo (Sun et al., 1996; Vaughan-Jones, 1986; Leem and Vaughan-Jones, 1998), the individual effects of pHi and pHo on heterologous AE3 in transfected 293 cells (Sterling and Casey, 1999) may have been masked by the chosen experimental conditions. Interestingly, whereas the pHi dependence of anion exchange in guinea pig myocytes is similar to that exhibited by mouse AE2 expressed in Xenopus oocytes their responses to varying pHo differ. In guinea pig myocytes, low pHo stimulates and high pHo inhibits Cl-/HCO3- exchange, with a pHo(50) value of 7.03 (Wilson and Vaughan-Jones, 2000), a value similar to that of AE2-mediated Cl-/Cl- exchange oppositely regulated in oocytes by pHo. The functionally dominant anion exchanger protein of guinea pig myocytes has not yet been identified, although AE3 (Kopito et al., 1989; Kudrycki et al., 1990; Yannoukakos et al., 1994) and AE1 (Richards et al., 1999) have both been proposed.
The pH-regulatory subdomains within the AE2 NH2-terminal cytoplasmic domain whose function requires the integrity of aa 318–323 and aa 336–347 likely interact with or control the conformation of effector moieties. These may include binding site(s) at the cyto-plasmic face of the AE2 transmembrane domain not present in AE1, binding sites elsewhere within the AE2 cytoplasmic NH2-terminal domain, and/or binding surfaces on separate, regulatory polypeptides. These possibilities are currently under investigation. Although the structure of the transmembrane domain of AE1 is known only to a resolution of 18 Å, several mutations of human AE1 associated with distal renal tubular acidosis (dRTA) (Alper et al., 2002) suggest sites at the cytoplasmic face of the transmembrane domain that are sensitive to the structure of the NH2-terminal cytoplasmic domain, and so either influence or themselves might represent sites of intramolecular interaction. Thus, the AE1 mutants R589H (Bruce et al., 1997; Jarolim et al., 1998) and R901X (Karet et al., 1998) are both functional in the context of the erythroid AE1 NH2-terminal cytoplasmic domain. In contrast, within the context of the renal type A intercalated cell, in which the NH2-terminal cytoplasmic domain of kidney AE1 lacks the first 65 residues present in erythroid AE1, these mutants can exhibit impaired surface accumulation (Toye et al., 2002) and/or a dominant-negative trafficking phenotype (Quilty et al., 2002).
As judged by the butyrate removal assay, pHi regulates both Cl-/Cl- and Cl-/HCO3- exchange mediated by AE2. The recent demonstrations that Cl-/HCO3- exchange mediated by AE2 and by AE3 can be regulated by bound carbonic anhydrase II (Sterling et al., 2002a) and by carbonic anhydrase IV (Sterling et al., 2002b) adds to the potential complexity of AE regulation by pH. These AE binding proteins might in some contexts exert allosteric effects on AE activity, in addition to their presumed functions in channeling transport substrate to and from the anion translocation pathway through the polypeptide.
Conclusion
The present study has defined individual, conserved amino acid residues in the NH2-terminal cytoplasmic domain of the AE2 anion exchanger whose mutation alters the independent regulation of AE2 activity by pHo and by pHi. The corresponding conserved glu-tamate residues w
