Mutation of calmodulin-binding site renders the Na+/H+ exchanger (NHE1) highly H(+)-sensitive and Ca2+ regulation-defective.

The ubiquitous plasma membrane Na+/H+ exchanger (NHE1) is rapidly activated in response to various extracellular signals. To understand how the intracellular Ca2+ is involved in this activation process, we investigated the effect of Ca2+ ionophore ionomycin on activity of the wild-type or mutant NHE1 expressed in the exchanger-deficient fibroblasts (PS120). In wild-type transfectants, a short (up to 1 min) incubation with ionomycin induced a significant alkaline shift (approximately 0.2 pH unit) in the intracellular pH (pHi) dependence of the rate of 5-(N-ethyl-N-isopropyl) amiloride-sensitive 22Na+ uptake, without changes in the cell volume and phosphorylation state of NHE1. Mutations that prevented calmodulin (CaM) binding to a high affinity binding region (region A, amino acids 636-656) rendered NHE1 constitutively active by inducing a similar alkaline shift in pHi dependence of Na+/H+ exchange. These same mutations abolished the ionomycin-induced NHE1 activation. These data suggest that CaM-binding region A functions as an "autoinhibitory domain" and that Ca2+/CaM activates NHE1 by binding to region A and thus abolishing its inhibitory effect. Furthermore, we found that a short stimulation with thrombin and ionomycin had apparently no additive effects on the alkaline shift in the pHi dependence of Na+/H+ exchange and that deletion of region A also abolished such an alkaline shift induced by a short thrombin stimulation. The results strongly suggest that the early thrombin response and the ionomycin response share the same activation mechanism. Based on these data and the results shown in the accompanying paper (Bertrand, B., Wakabayashi, S., Ikeda, T., Pouysségur, J., and Shigekawa, M. (1994) J. Biol. Chem. 269, 13703-13709), we propose that CaM is one of the major "signal transducers" that mediate distinct extracellular signals to the "pHi sensor" of NHE1.

* This work was supported by Grant-in-aid for Scientific Research C (to S. W.) and Grant-in-aid on Priority Areas 321 from the Ministry of Education, Science, and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Supported by the Science and Technology Agency Fellowship of Japan.
The Na+/H+ exchanger (NHE1 human isoform) that has recently been cloned (6) and characterized (7-91, is a ubiquitous amiloride-sensitive electroneutral transporter that can be activated in response not only to hormones and growth factors but also to other stimuli such as hyperosmotic stress (2)(3)(4). The activation of NHEl results from increased affinity of the allosteric modifier site of the exchanger for the intracellular H+ (10)(11)(12). At present, however, little is known about the molecular mechanism of regulation of this pHi sensor. Previous studies indicate that protein kinase C is involved in this activation of NHEl (2,3). Several lines of evidence have suggested that Ca2+ also activates Na+/H+ exchange. However, the effect of intracellular Ca2+ has been controversial. Ca2+ ionophore has been reported to stimulate Na+ influx in human foreskin fibroblast (HSWP cells) (13) and increase pH, in thymic lymphocyte (141, human skin fibroblast (WS-1 cells) ( X ) , and human platelets (16). On the other hand, Ca2+ ionophore has been reported not to increase pH, in other types of cells such as 3T3 fibroblasts (171, human foreskin fibroblasts (121, and smooth muscle cells (18). These inconsistent data may result from the use of different cell types or different experimental conditions. In the accompanying paper (19), we found that NHEl is a novel member of the calmodulin (CaM)-binding proteins and that two CaM-binding sites are located in the middle of the carboxyl-terminal cytoplasmic regulatory domain. We have presented evidence that the high affinity CaM-binding regionA (aa 636-656) is involved in the activation of NHEl in response to growth factors and osmotic stress. In view of a micromolar range of intracellular CaM concentration, we thought that direct binding of Ca2+/CaM to region A may be a key event in the Ca2+-induced activation of NHE1.
In the present work, we studied the effect of a n ionomycininduced [Ca"], increase on the Na+/H+ exchange activity in PS120 fibroblasts expressing the wild-type and mutant NHE1. Here we show that the ionomycin-induced [Ca2+l, increase elicits a n alkaline shift in pH, dependence of Na+/H+ exchange without changes in the cell volume and phosphorylation state of NHE1. From the analysis with transfectants having mutations in the high affinity CaM-binding region A, we propose that region A exerts an inhibitory effect on protonation of the allosteric modifier site and that direct binding of CaZ+/CaM to it reverses its inhibition. We also present evidence that the Ca2+induced CaM binding to region A is a principal mechanism for the NHEl activation in the early phase of thrombin response.
EXPERIMENTAL. PROCEDURES Materials-The amiloride derivative, 5-(N-ethyl-N-isopropyl)amiloride (EIPA) was a gift from New Drug Research Laboratories of Kanebo, Ltd. (Osaka, Japan). W a c 1 and [7-"C]benzoic acid was purchased from DuF'ont NEN. All other chemicals were of the highest purity available.
Construction and Stable Expression of NHEl Mutant Molecules-The plasmid including cDNA coding for the NHEl human isoform with deleted of the 5'-untranslated region was described previously (plasmid designated pEAl"A5') (7). The plasmid (pEA698) coding for NHEl deleted of the carboxyl-terminal cytoplasmic tail (aa 699-815) has also been described (7). Construction of a deletion mutant (A637-656; aa 637-656 were deleted) and a charge reversal mutant (1K3R4E; Lys-641, Arg-643, Arg-645, and Arg-647 were replaced by 4 glutamic acids) and stable expression of the corresponding mutant exchangers were performed as described in the accompanying paper (19).
Measurement ofpH, Dependence of "Nu' Uptake-pH, dependence of ,,Na+ uptake was measured as described previously (11) with slight modifications. Stable transfectants grown to confluence in 24-well dishes were incubated for 5 h in a serum-free, bicarbonate-free H21 medium buffered with 20 m~ Hepes (pH 7.4) to maintain the Na+/H+ exchanger in the resting state. Cells were loaded with 0-30 rn NH,Cl for 30 min at 37 "C in NaCl standard solution (composition: 20 rm HepewTris (pH 7.4), 120 rm NaCI, 5 rn KCl, 2 m~ CaCl,, 1 m~ MgCl,, and 5 rm glucose). Cells were then rapidly washed once with choline chloride standard solution (composition: 20 rm Hepedlks (pH 7.4), 120 rm choline chloride, 2 rn CaCl,, and 1 rm MgCl,) with or without 5 p ionomycin and incubated for 40 s in the same medium. 'Wa' uptake was started by incubating cells in choline chloride standard solution containing 1 rm *%aC1(37 kBq/ml) and 1 rm ouabain with or without 5 p ionomycin. In some wells, the Wa' uptake medium also contained 0.1 rm EIPA. Forty s later, cells were rapidly washed four times with ice-cold phosphate-buffered saline to terminate '%a+ uptake. Cells were then solubilized with 0.1 N NaOH, and radioactivity was measured with a y-counter. pH, was estimated by measuring the distribution of ["Clbenzoic acid (37 kBq/ml) (21) under the same conditions as those used for w a + uptake measurement, except that the uptake medium contained ["Clbenzoic acid and nonradioactive NaCl.
Measurement of Intracellular Na' Concentration-For measurement of the intracellular Na' concentration, cells were equilibrated with 110 kBq/mI of ' 2NaCl for 4 h at 37 "C in a serum-free H21 medium buffered with 20 rn Hepes (pH 7.4). Cells were further incubated for 30 min in NaCl standard solution containing 110 kBq/ml =NaCl and 3 rm NH,Cl.
Cells were then rapidly washed once with choline chloride standard solution with or without 5 p ionomycin and incubated for another 40 s in the same medium. Cells were then washed four times with ice-cold phosphate-buffered saline to stop ,,Na+ equilibration and solubilized with 0.1 N NaOH. Radioactivity was measured with a y-counter. The intracellular Na' concentration was calculated by assuming 5 pl of cell waterlmg of protein (21).
Measurement of Cell Volume-Cell volume was measured by cell sizing using a Coulter Counter (model ZBI; Coulter Electronics) associated with a Coulter Channelyzer (model C1000). The averaged cell volume was calculated from volume distribution curves. Before cell volume measurement, cells were trypsinized and resuspended in NaCl standard solution. Cell volume did not significantly change during 2 h after trypsinization.
Other Procedures-Preparation of a rabbit polyclonal antibody (RPcd) was described in the accompanying paper (19). Protein concentration was measured by bicinchoninic assay system (Pierce Chemical Co.) using bovine serum albumin as a standard. 1 shows the time courses of EIPA-sensi- The experiment was carried out with stable transfectants expressing the wild-type NHEl as described. A, prior to ,,Na+ uptake, cells were loaded with 3 m~ NH,Cl in NaCl standard solution for 30 min. Cells were then washed once and incubated for 40 s in the Na+-free choline chloride standard solution containing 0 or 5 p ionomycin. Cells were then incubated in the uptake medium containing 1 rm "NaC1, 1 rm ouabain, and 0 or 5 p ionomycin with or without 0.1 m~ EIPA, and time courses of EIPA-sensitive ',Na+ uptake were followed. B, the same experiment as that in A was carried out in cells loaded with 30 m~ NH;. C , as in A, cells were loaded with 3 r m NH4Cl, washed, and incubated in the Na+-free solution containing 0 or 5 p ionomycin. Time courses of pH, change were then measured for up to 80 s in the absence or presence of 5 p ionomycin after incubating cells in the ["Clbenzoic acid-containing Na' uptake medium in which NaCl was used in place of '%aCl. For measurement of pH, at time of additions of choline chloride solution (-40 s) and the uptake medium (0 s), the ['4Clbenzoic acidcontaining solutions were added 40 s before. tive "Na+ uptake by PS120 cells expressing the wild-type NHE1. 'Wa+ uptake proceeded linearly at least for the first minute. When cells were preloaded with 3 m~ NH;, ionomycin stimulated the rate of EIPA-sensitive "Na+ uptake by 5-fold ( Fig. L4). In contrast, when cells were preloaded with 30 m~ NH; to produce a strong intracellular acidification, ionomycin did not stimulate 'Wa+ uptake ( Fig. U?), indicating that ionomycin did not change the maximal rate (V,,,,) of 2Wa+ uptake.

Ca2+ Ionophore Induces an Alkaline Shifi in pHi Dependence
The ionomycin-induced activation at low NH; concentration was not due to a change in the intracellular Na+ concentration, because the latter, as measured as described under "Experimental Procedures," was 9.6 2 0.8 and 10.5 f 0.4 mM (means 2 S.D., n = 3) after treatment with and without ionomycin, respectively. Under the conditions of Fig. L4, however, ionomycin produced a significant cytosolic acidification (-0.2 pH unit) (Fig. lC), which accounts for part of the observed activation of "Na+ uptake. We also found that when different batches of wild-type transfectants were loaded with 3 mM NH,, the ionomycin-induced activation of "Na+ uptake and cytosolic acidification varied from one batch of cells to the other. To avoid these complications, we measured both 'Wa+ uptake and pH, under the same experimental conditions using the same batch of cells and always analyzed zzNa+ uptake rates as a function of pH,. As shown in Fig. 2 Ionomycin-induced alkaline shift in pH, dependence of EIPA-sensitive -a+ uptake by the wild-type transfectant. The rate of "Na' uptake and pH, during uptake were measured in cells expressing the wild-type exchanger in the absence (0) or presence (0) of 5 1.1~ ionomycin as described under "Experimental Procedures." The rate of 22Na' uptake was plotted as a function of pHi.
[Ca2+], to more than 0.5 PM (data not shown), these data provided convincing evidence that the rise in [Ca2+Ii stimulates the Na+/H+ exchange activity by increasing pH, sensitivity of the exchanger.
Zonomycin Does Not Change Cell VolumeSince the ionomycin-induced increase in [Ca2+Ii has been reported to induce cell shrinkage, which may lead to exchanger activation (14,241, we checked the effect of ionomycin on the volume of cells expressing the wild-type NHE1. As shown in Fig. 3, ionomycin did not change the cell volume during the first 10 min, whereas sucrose (100 mM) induced a rapid, significant loss (10%) of cell volume. Thus, cell shrinkage does not appear to be involved in the ionomycin-induced activation of NHEl at least under our conditions.
Zonomycin Does Not Zncrease Phosphorylation of NHEl-One possible mechanism for the Ca2+-induced activation of NHEl could be phosphorylation of critical regulatory site(s) in the NHEl molecule (8,9). We checked whether ionomycin increased phosphorylation of NHE1. Cells expressing the wildtype NHEl were depleted of serum, labeled with 32Pi and then stimulated for 1 min with 5 p~ ionomycin or for 20 min with 2 unitdm1 a-thrombin as described under "Experimental Procedures." The exchanger was subsequently immunoprecipitated with a specific antibody (RP-cd). As shown in Fig. 4A, 1-min treatment of cells with ionomycin did not significantly increase phosphorylation of the exchanger, whereas thrombin markedly increased it after the 20-min treatment; 32P radioactivity incorporated into the exchanger, relative to the control value, was 0.99 2 0.17 and 1.60 2 0.44 (means 2 S.D., n = 3) for the ionomycin-and thrombin-stimulated cells, respectively. It is important to point out that during the early phase (up to 1 min) of thrombin stimulation, phosphorylation of the exchanger did not increase significantly (data not shown), which is consistent with the previous observation (8). Fig. 4B shows the result of the phosphopeptide mapping analysis. As described previously (9, 22 and 23), four major phosphopeptides (Pl-P4) together with additional minor spots (a, b, and c) appeared in the maps. The minor spots may represent the partially digested fragments as described previously (23). 32P Incorporation into two phosphopeptides (P3 and P4) relative to those into P1 and P2 increased in response to thrombin. However, ionomycin neither increased 32P incorporation into these major phosphopeptides nor produced any additional major phosphopeptide. These findings strongly suggest that phosphorylation of NHEl did not occur during the ionomycin-induced rapid activation of NHE1.
Mutations in CaM-binding Region A Abolish Zonomycin-induced Activation of NHEl-In the accompanying paper (19)  and a mutant with charge reversal mutation of this region (1K3R4E), which lost the CaM-binding ability (19). As shown in Fig. 5 ( A and B ), these mutations almost completely abolished the ionomycin-induced alkaline shift in pH, dependence of the rate of EIPA-sensitive "Na+ uptake. In addition, deletion of the cytoplasmic tail (aa 636-8151 (plasmid A635 (23)) also produced the same effect (data not shown).2 In sharp contrast, a mutant deleted of the carboxyl-terminal tail (aa 699-815) (plasmid pEA698 (7)) that retained CaM-binding regions, preserved the ionomycin-induced alkaline shift (Fig. 5C). These findings suggest that direct binding of Ca2+lCaM to region A stimulated the exchange activity by increasing pHi sensitivity.

Mutations in CaM-binding Region A Abolish Thrombin-induced
Activation of NHEl-In the previous study (ll), a brief stimulation of Chinese hamster lung fibroblastic cells (CCL39, a parental cell line of PS120) with a combination of a-thrombin and insulin has been shown to induce an alkaline shift in pH, dependence of the rate of amiloride-sensitive Na' uptake. As shown in Fig. 6 A , we also found that a brief (-1 min) stimulation with thrombin induced a similar alkaline shift (-0.2 pH unit) in pH, dependence of 'Wa+ uptake in cells expressing the wild-type NHE1. Interestingly, when cells were stimulated with a combination of thrombin and ionomycin, the effects of these agents are not additive (data not shown). Furthermore, deletion of CaM-binding region A almost completely abolished the rapid thrombin-induced alkaline shift in pH, dependence of 'Wa+ uptake (Fig. 6B 1. Therefore, the early phase of thrombin effect and the ionomycin effect are likely to be controlled by the same mechanism. It should be noted, however, that the same mutations of CaM-binding region A reduced the cytoplasmic alkalinization by only 50% when cells were stimulated with thrombin for 15 min (19). Thus, the early and part of late phase of the thrombin response appear to be mediated via different mechanisms.
Mutations of CaM-binding Region A Increase pH, Sensitivity of NHEl-When pH, dependence curves of the zzNa+ uptake rate were compared among four different transfectants, it was pH, dependence of *%a+ uptake similar to that of the wild-type. Re-' In a previous paper (23), we reported that A635 mutant exhibits a cently, we extensively repeated the same experiments under slightly different conditions and found that A635 has a definitively higher pH, sensitivity (by -0.2 pH unit), as compared to the wild-type. noted that apparent pK values for two transfectants (wild-type and A6981 are lower than those for the other two (A637456 and 1K3R4E) (compare Figs. 2, 5, and 6). Such a difference became clearer when the rates of z2Na+ uptake were normalized by the V, , value obtained in each experiment (Fig. 7). Interestingly, when the former two transfectants were stimulated with ionomycin or for a short time (-1 min) with thrombin, their pH, dependence curves became similar to those for the latter ones. Thus, mutations defective in CaM binding to region A caused a alkaline shift of approximately 0.2-0.3 pH unit in the pH, dependence of Z%a+ uptake.
To confirm that the pHi sensitivity of Na+/H+ exchange was indeed different among the wild-type and mutant transfectants, we measured a change in the resting value of pH, in cells placed at various pH,. As shown in Fig. 8, pH,  values of Na+/H+ exchange. Therefore, the data strongly suggest that mutants of CaM-binding region A possess higher pH, sensitivity, compared to the wild-type exchanger. These findings demonstrate that mutations of C a " binding region A constitutively activate the exchanger by increasing its pH, sensitivity.
It should be noted that despite constitutive activation of mutants, their maximal activities were lower than that of the wild-type NHEl (cf. Figs. 2, 5, and 6). This difference appears to be due to the reduced number of NHEl expressed in the plasma membrane of mutant transfectants (data not shown). High copy-numbered mutant cells with high constitutive activity could be eliminated during the "H+-killing" selection because of the toxicity of high pH, to cell growth.

DISCUSSION
The NHEl molecule can be separated into two distinct domains, an amino-terminal transport domain (T) that catalyzes amiloride-sensitive Na+/H+ exchange with a built-in modifier site (pH sensor) and a carboxyl-terminal cytoplasmic regulatory domain that determines the set point value of the modifier site (7) (cf. Fig. 9). Protonation of the modifier site is generally considered to activate the exchanger. In the present work, we studied the mechanism by which [Ca2+], regulates activity of Mes whose pH was adjusted to the indicated value of pH,. Cells were then incubated for 5 min in the same medium containing 18 kBq/ml ['4Clbenzoic acid to measure pHi the Na+/H+ exchanger. We found that Ca2+ ionophore ionomycin rapidly (within 40 s) stimulated NHEl activity by increasing its pH, sensitivity (Fig. 2). Mutations that prevented CaM binding to a high affinity binding region (region A, aa 636-656 of NHE1) of the cytoplasmic domain rendered NHEl constitutively active by inducing an alkaline shift in the pH, dependence (Fig. 7). These same mutations abolished the ionomycininduced activation of NHEl (Fig. 5). Therefore the high affinity CaM-binding region A exerts an inhibitory effect on protonation of the modifier site of the exchanger. We propose that in quiescent cells CaM-binding region A, when unbound by CaM, functions as an autoinhibitory domain and that the ionomycininduced increase in [Ca2+Ii activates the exchanger by permitting CaM to bind to region A, thus preventing it from exerting the inhibitory effect. The autoinhibitory effect of the CaM-binding site has been postulated for other CaM-binding proteins such as the plasma membrane Ca2' pump (25). On the other hand, it has been suggested that the CaM-binding domain does not fully overlap with the autoinhibitory domain in Ca2+/CaMdependent protein kinase I1 (26). It is possible that in NHEl the autoinhibitory domain overlaps only partially with CaMbinding region A. It is also possible that CaM-binding region B, which has intermediate afinity for CaM and is located next to region A, may form part of the inhibitory domain. Future work is needed to clarify these points.
One could argue that the effects caused by mutations of region A may be due to a long range structural distortion affecting the function of the whole cytoplasmic domain. This seems not to be the case, however, because mutations of region A "constitutively" activate the exchanger, as opposed to the inhibition expected from the overall structural distortion. In addition, a large deletion of the cytoplasmic tail (aa 699-815) did not influence the NHEl functions including the pH, sensitivity of Na+/H+ exchange and the ionomycin-or growth factorinduced alkaline shift in the pH, dependence (see "Results" and Ref. 7). Therefore, we think that the observed mutation-induced effects are attributable to the loss of function of a limited segment (region A) of the cytoplasmic domain.
One could also argue that the ionomycin-induced activation of NHEl may be due to cell shrinkage caused by an increase in [Ca2+], (14, 24). However, cell shrinkage did not occur in response to ionomycin under our experimental conditions, although osmotic response to a hypertonic solution occurred normally (Fig. 3). Therefore, cell shrinkage did not contribute to the observed ionomycin-induced activation of NHE1. It is also possible that an increase in [Ca2+], will activate protein kinase(s) such as Ca2+/CaM-dependent protein kinase or protein kinase C, which may eventually lead to phosphorylation and activation of NHE1. However, we detected neither a significant increase in the phosphorylation of NHEl within 1 min after addition of ionomycin (Fig. 4A) nor ionomycin-induced formation of new phosphopeptides in the peptide map analysis (Fig.  4B). Thus direct phosphorylation of NHEl is not involved in the rapid response to ionomycin. Receptor occupancy by growth factors and hormones induces a rapid increase in [Ca2+],. We examined how a short stimulation with a-thrombin affected the NHEl activity. We found t h a t (i) thrombin and ionomycin had apparently no additive effects on the alkaline shift in the pH, dependence of Na+/H+ exchange (see "Results"), (ii) deletion of CaM-binding region A completely abolished the thrombin-induced alkalinization (Fig.  6), and (iii) thrombin did not increase phosphorylation of NHEl after a short stimulation (1 min) (see "Results"). These data suggest that direct interaction of Ca2+/CaM with regionA is also the main mechanism for the activation of NHEl in the early phase of thrombin response.
It is important to note that the mutations abolishing CaM binding to region A (aa 636-656) reduced the cytoplasmic alkalinization induced by a long (10-20 min) stimulation with thrombin only by 50% (see Fig. 7 in the accompanying paper (Ref. 19); see also Ref. 23). On the other hand, deletion of aa 567-635 (plasmid A567-635) resulted in a complete loss of such thrombin-induced alkalinization and a drastic acidic shift in the pH, dependence of Na+/H+ exchange (7,23). The latter result suggests that aa 567-635 of the cytoplasmic domain is required for the maintenance of high pH, sensitivity, as well as for the full activation of the exchanger by thrombin. Therefore, activation of NHEl in response to growth factors seems to be mediated via at least two independent mechanisms involving interaction of two different regions of the cytoplasmic domain with the modifier site (pH, sensor) within the NH,-terminal transporter domain (T), as illustrated in our working model (Fig. 9). Amino acids 636-656 are involved in Ca2+/CaM-dependent, rapid activation of NHE1, whereas aa 567-635 are involved in the slow and long lasting activation induced by growth factors.
In our previous paper, we proposed that aa 567-635 mediates growth factor signals through the effect of a putative regulatory ancillary protein (R) (23). Protein phosphorylation is involved in this R-mediated regulation, because a potent phosphatase inhibitor okadaic acid or protein kinase C activator phorbol ester has been shown to induce a long lasting activation of NHEl (2,3,9,27), even when the cytoplasmic tail (aa 63-15) was deleted (23). 'The R-mediated mechanism may involve phosphorylation of R itself, because all major phosphorylation sites in the NHEl molecule has been shown to map to the cytoplasmic tail (aa 636-815), yet deletion of this region (plasmid A6351 still preserves 50% of cytoplasmic alkalinization in response to growth factors including thrombin (23). Thus growth factors seem to regulate activity of NHEl by the Ca2+/ CaM-dependent and the R-mediated mechanisms, both of which do not require direct phosphorylation of the exchanger molecule. In addition to these, direct phosphorylation of the NHEl cytoplasmic tail could provide another mechanism for activation of the exchanger. The latter mechanism will require further investigation.
In this study, we investigated the Ca2+-dependent activation of NHE1. It is interesting to compare the amino acid sequence of NHEl with those of three other recently cloned NHE isoforms (NHE2, NHE3, and NHE4) (6,2831). The amino-terminal transport domain with 12 (or 11) transmembrane helices are well conserved among these isoforms (5040% overall sequence homology). Interestingly, two transmembrane helices (M5a and M5b) and the intracellular loop intervening these helices contain several conserved negatively charged residues, some of which may form part of the modifier site (32). It is possible that these residues interact with positively charged residues within CaM-binding region A of NHEl or corresponding regions of other isoforms. In the amino-terminal portion (aa 470-680) of the cytoplasmic domain, many of the amino acids of CaM-binding region A of NHEl are conserved in the corresponding regions of NHE2 and NHE4, although relatively high overall sequence homology (-60%) is found only between NHE2 and NHE4 (30). Thus CaM may also bind to NHE2 and NHE4. On the other hand, a relatively low overall homology (20-308) is found in the carboxyl-terminal portions of the cytoplasmic domain of four isoforms. The recently cloned trout exchanger (trout P-NHE), which does not have a sequence homologous to region A of NHE1, has been shown to exhibit a higher pH, sensitivity, as compared to the human NHEl (33). In contrast, an epithelial isoform (rat NHE3) has been reported to possess a lower pH, sensitivity (34). Such variation in the pH, sensitivity could be due to the variation in the sequence of the specific cytoplasmic segments that interact with the modifier site within the amino-terminal transport domain.
In conclusion, the present data support the notion that the NHEl activation induced by a relative short incubation with Ca2+ ionophore or thrombin occurs by reversing the inhibition by the CaM-binding autoinhibitory domain through the direct binding of Ca2+/CaM. These data suggest that CaM is one of major signal transducers that mediate distinct extracellular signals to pH, sensor of NHE1. The regulation of NHEl via Ca2+/CaM binding could be particularly important for muscle cells such as cardiac myocytes. These cells will require rapid extrusion of H+ produced in response to increases in [Ca2+l, and metabolic activity associated with contraction. The Na+/H+ exchange stimulated rapidly through the direct binding of Ca2'1 CaM would serve as a potent H' extrusion system during contraction.