Nature & Wildlife

9 pages
12 views

A pH-dependent conformational change of NhaA Na(+)/H(+) antiporter of Escherichia coli involves loop VIII-IX, plays a role in the pH response of the protein, and is maintained by the pure protein in dodecyl maltoside

of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
A pH-dependent conformational change of NhaA Na(+)/H(+) antiporter of Escherichia coli involves loop VIII-IX, plays a role in the pH response of the protein, and is maintained by the pure protein in dodecyl maltoside
Transcript
   A pH-dependent Conformational Change of NhaA Na   /H   Antiporterof   Escherichia coli  Involves Loop VIII–IX, Plays a Role in the pHResponse of the Protein, and Is Maintained by the Pure Protein inDodecyl Maltoside* (Received for publication, March 22, 1999, and in revised form, June 1, 1999)  Yoram Gerchman, Abraham Rimon, and Etana Padan‡  From the Division of Microbial and Molecular Ecology, Institute of Life Sciences, Hebrew University of Jerusalem,91904 Jerusalem, Israel Digestion with trypsin of purified His-tagged NhaA ina solution of dodecyl maltoside yields two fragments atalkaline pH but only one fragment at acidic pH. Deter-mination of the amino acid sequence of the N terminusof the cleavage products show that the pH-sensitivecleavage site of NhaA, both in isolated everted mem-brane vesicles as well as in the pure protein in deter-gent, is Lys-249 in loop VIII–IX, which connects trans-membrane segment VIII to IX. Interestingly, the twopolypeptideproductsofthesplitantiporterremaincom-plexed and co-purify on Ni 2  -NTA column. Loop VIII–IX hasalsobeenfoundtoplayaroleinthepHregulationof NhaA;threemutationsintroducedintotheloopshiftthepHprofileoftheNa   /H  antiporteractivityasmeasuredin everted membrane vesicles. An insertion mutationintroducing Ile-Glu-Gly between residues Lys-249 and Arg-250 (K249-IEG-R250) and Cys replacement of either Val-254 (V254C) or Glu-241 (E241C) cause acidic shift of the pH profile of the antiporter by 0.5, 1, and 0.3 pHunits, respectively. Interestingly, the double mutantE241C/V254C introduces a basic shift of more than 1 pHunit with respect to the single mutation V254C. Takentogether these results imply the involvement of loop VIII–IX in the pH-induced conformational change,which leads to activation of NhaA at alkaline pH. Sodium proton antiporters are ubiquitous membrane pro-teins found in the cytoplasmic and organelle membranes of cells of many different srcins, including plants, animals andmicroorganisms. They are involved in cell energetics and playprimary roles in the regulation of intracellular pH, cellularNa  content, and cell volume (reviewed in Refs. 1–4).  Escherichia coli  has two antiporters, NhaA (5) and NhaB (6),which specifically exchange Na  or Li  for H  (3).  nha  A isindispensable for adaptation to high salinity, for challenging Li  toxicity, and for growth at alkaline pH (in the presence of Na  ) (7). Accordingly, expression of   nha  A, which is dependenton NhaR, a positive regulator, is induced by Na  , in a pH-de-pendent manner (8–10).  nha B by itself confers a limited so-dium tolerance to the cells, but becomes essential when thelack of NhaA activity limits growth (11).Both the NhaA and NhaB are electrogenic antiporters thathave been purified to homogeneity and reconstituted in a func-tional form in proteoliposomes (12–14). The H   /Na  stoichiom-etry of NhaA is 2H   /Na  and that of NhaB 3H   /2Na  . NhaBbut not NhaA is sensitive to amiloride derivatives, and the rateof activity of NhaA but not of NhaB is drastically dependent onpH, changing its  V  max  over 3 orders of magnitude from pH 7 topH 8 (12).Interestingly, a strong pH sensitivity is characteristic of antiporters as well as other transporters that are involved inpH regulation (reviewed in Ref. 4). Identifying the amino acidresidues involved in the pH sensitivity of these proteins isimportant for understanding the mechanism of pH regulation.NhaA contains eight histidines, none of which were foundessential for the Na   /H  antiporter activity of NhaA (15).However, replacement of histidine 225 by Arg (H225R) sug-gested that His-225 has an important role in the pH sensitivityof the antiporter. Whereas the activation of the wild-type NhaA occurs between pH 7.5 and pH 8, that of H225R antiporteroccurs between pH 6.5 and pH 7.5. In addition, while thewild-type antiporter remains almost fully active, at least up topH 8.5, H225R is reversibly inactivated above pH 7.5, retaining only 10–20% of the maximal activity at pH 8.5 (15). Further-more, replacement of His-225 with either cysteine (H225C) orserine (H225S) but not alanine (H225A) yielded an antiporterwith a wild-type pH-sensitive phenotype, implying that polar-ity and/or hydrogen bonding, the common properties shared byHis, Cys, and Ser, are essential at position 225 for pH regula-tion of NhaA (16). Glycine 338 affects the pH response of NhaA;its replacement with serine (G338S in TMS 1  XI) produced atransporter, which in contrast to the wild-type protein lacks pHcontrol; it is active between pH 6.5 and 8.5 (17).Recently, we have found that NhaA undergoes a conforma-tion change upon its activation by pH which can be probed bytrypsin (18). At acidic pH the protein in everted membrane vesicles is completely resistant to trypsin, while at alkaline pHit is digested in a pattern reflecting the pH profile of theantiporter activity. Furthermore, two mutants with a modifiedpH profile are susceptible to trypsin in isolated membrane vesicles only at the pH range, where they are active and re-flecting the level of activity (18). H225R, the mutant with a pHprofile shifted toward acidic pH, is digested at the pH where itis active; G338S, which lost pH control, is active and exposed to * This work was supported by the Bundes Ministeriumf forschung,Wissenschaft, Technologie and Erzieihung, the International Bureau of the Bundes Ministerium fu¨r Forschung, Wissenschaft, Technologie andErzieihung at the Deutscher Luftraum Zentrum (German-Israeli Pro- ject Cooperation on Future-oriented Topics), and the Moshe Shilo Cen-ter for Biogeochemistry. The costs of publication of this article weredefrayed in part by the payment of page charges. This article musttherefore be hereby marked “ advertisement ” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.‡ To whom correspondence should be addressed. Tel.: 972-2-6585094;Fax: 972-2-6586947; E-mail: etana@vms.huji.ac.il. 1 The abbreviations used are: TMS, transmembrane domain; DM, n -dodecyl   - D -maltoside; PAGE, polyacrylamide gel electrophoresis;Tricine,  N  -[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; Bis-Trispropane, 1,3-bis[tris(hydroxymethyl)methylamino]propane. T HE  J OURNAL OF  B IOLOGICAL  C HEMISTRY   Vol. 274, No. 35, Issue of August 27, pp. 24617–24624, 1999 © 1999 by The American Society for Biochemistry and Molecular Biology, Inc.  Printed in U.S.A. This paper is available on line at http://www.jbc.org  24617   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  6 h  t   t   p :  /   /   w w w . j   b  c  . or  g /  D o wnl   o a  d  e  d f  r  om   trypsin throughout the entire pH range of activity. Although NhaA has many trypsin-cleavable sites, only twomain fragments were observed following digestion of isolatedmembrane vesicles at alkaline pH. This observation suggeststhat only one cleavage site is exposed by pH while all the othersites are masked. It was therefore inferred that the trypsincleavage site is located in, and therefore serves as a tag of, thatpart of the protein which undergoes a conformational change inresponse to pH. Identification and study of this site was there-fore undertaken in this study. The results show that loop VIII–IX is important for the pH regulation of NhaA and bearsthe trypsin cleavage site, which is involved in the pH-depend-ent conformational change of NhaA. This change is maintainedby the pure protein in DM. EXPERIMENTAL PROCEDURES  Bacterial Strains and Culture Conditions— EP432 is an  E. coli  K12derivative, which is  mel BLid,   nha  A1:: kan ,   nha B1:: cat ,   lac ZY, thr1(11). TA16 is  nha  A   nha B  lac I Q and otherwise isogenic to EP432 (12).DH5   (U. S. Biochemical Corp.) and JM109 were used as hosts forconstruction of plasmids. Cells were grown in modified L broth in whichNaCl was replaced by KCl (Ref. 7; 87 m M , pH 7.5). Where indicated, themedium was buffered by 60 m M  Bis-Tris propane, and pH was titratedwith HCl. Cells were also grown in minimal medium A without sodiumcitrate (19) with either glycerol (0.5%) or melibiose (10 m M ) as a carbonsource. Thiamine (2.5   g/ml) was added to all minimal media. Forplates, 1.5% agar was used. Antibiotics were 100   g/ml ampicillin,and/or 50   g/ml kanamycin, and/or 12   g/ml chloramphenicol, and/or12.5   g/ml tetracycline. Resistance to Li  and Na  was tested asdescribed previously (15).  Plasmids— pGM36 and pGMAR100 are pBR322 derivatives (17, 20);the first bears nha  A and most of the nha R gene. The latter carries nha  A and C-terminal truncated  nha R. pAR100 is a pACYC184 derivative,which carries an insert identical to that of pGMAR100 (17). pI Q is apACYC184 derivative, which carries  lac I Q (kindly provided by E. Bibi,Weizman Institute of Science, Rehovot, Israel).pAXH is a plasmid carrying Xa-His-tagged NhaA. It was constructedpreviously (21) and contains NhaA fused at its N terminus to the  tac promoter for over expression and at its C terminus to a sequenceencoding in tandem two factor Xa protease cleavage sites and 6 His.pMXH was constructed by digestion of pAXH by  Mlu I and  Xho I, fol-lowed by end filling with T4 DNA polymerase and self-ligation of the4960-base pair fragment. The resulting plasmid carries  nha  A fusionencoding NhaA, of which RLRPSV C terminus is replaced byEHHHHHH.  Mutagenesis— Site-directed mutagenesis was conducted following apolymerase chain reaction-based protocol (22). DNA of pGMAR100 wasused as a DNA template. The end primers and the mutagenic primersare described in Table I.In the case of E241C and V254C, the resulting mutagenized DNA (1295 base pairs) was digested with  Nhe I and  Mlu I, yielding a fragmentof 879 base pairs, which was ligated either to the 4436-base pair  Nhe I-  Mlu I fragment of pGMAR100 to yield plasmids pV254C orpE241C or to the 4139-base pair  Nhe I-  Mlu I fragment of pAXH to yieldplasmids p(V254C)-XH or p(E241C)-XH. In the case of    (Lys-242–His-253), the resulting 1259-base pair polymerase chain reaction fragmentwas digested as above to yield an 820-base pair fragment and cloned asabove to yield pGMAR100 derivative p  (Lys-242–His-253). In the caseof the IEG insert between Lys-249 and Arg-250 (K249-IEG-R250), the1304-base pairs polymerase chain reaction fragment was digested asabove to yield an 888-base pair fragment and cloned as above into bothpGMAR100 and pAXH to yield plasmids p(K249-IEG-R250) (5324 basepairs) or p(Lys-249-IEG-R250A)-XH (5027 base pairs), respectively.  IsolationofMembraneVesicles,AssayofNa  /H    AntiporterActivity,and Quantitation of NhaA in the Membranes—  Assays of Na   /H  anti-port activity were conducted on everted membrane vesicles (23). Theassay of antiport activity was based upon the measurement of Na  (orLi  )-induced changes in the   pH as described (5, 24).High pressure membrane vesicles were prepared essentially aseverted membrane vesicles but the pressure used was 20,000 p.s.i.(French pressure cell press; SLM Aminco).Quantitation of the NhaA and its mutated derivative in membraneswas determined by Western analysis as described previously (16). Overexpression and Affinity Purification of His-tagged Antiport- ers— To overexpress the wild-type and the mutated antiporters, plas-mids overexpressing the His-tagged proteins in TA16 cells were used.The transformed cells were grown in minimal medium to  A 600  0.6,induced with 0.5 m M  isopropyl-1-thio-  - D -galactopyranoside, grown for2 h to  A 600  1–1.2, harvested (12), and used for preparation of highpressure membranes either after storage overnight at 4 °C or afterfreezing at   20 °C. Xa-His-tagged NhaA or His-tagged NhaA wereaffinity-purified on Ni 2  -NTA-agarose column (Qiagen, Hilden, Ger-many) as described (17).  DigestionbyTrypsin— Purified antiporter was subjected to trypsin ina 30-  l reaction mixture containing 10  g of antiporter protein, 30 ng of trypsin (type III from bovine pancreas, Sigma T-8253), 0.1% DM, 8.3m M  potassium acetate, 200 m M  KCl, 6.5% glycerol, 0.7 m M  Na/EDTA, 20m M  Hepes/Tris (pH 8, if not indicated otherwise), 1 m M  CaCl 2 . Incuba-tion was for 30 min at 37 °C. The reaction was terminated by theaddition of 100 ng of trypsin inhibitor (type II-S from soybean, SigmaT-9128). Samples of 5   g of protein were run on SDS-PAGE (17).Digestion of membrane vesicles (200   g of protein, 4   g of trypsin)was carried out in 100  l of reaction mixture containing 140 m M  KCl, 10m M  Tricine (pH 8), and 5 m M  MgCl and incubation conducted for 1 h at37 °C. This reaction mixture, which is also used for the Na   /H  anti-porter activity assay, gave similar digestion pattern and was occasion-ally used also with pure proteins as indicated.  Separation and Isolation of NhaA Tryptic Fragments and Determi-nation of the Amino Acid Sequence of Their N Termini— The proteinsample (50   g) was resuspended in SDS-PAGE sampling buffer andloaded on 12.5% bisacrylamide gel. After separation the polypeptideswere transferred (400 mA for 60 min) to polyvinylidene difluoride type(Millipore Immobilon TM -P) transfer membranes in transfer buffer con-taining 25 m M  Tris, 192 m M  glycine, 10% methanol, 0.025% SDS (pH8.4). The filters were then washed in distilled water, stained for 5 minin 0.1% Coomassie (R-250) in 50% methanol, destrained for 5–10 min in10% acetic acid in 50% methanol, and washed for 5–10 min in distilledwater. The stained bands were cut and subjected to N-terminal se-T  ABLE  I  Primers used for construction of NhaA mutants Mutation Mutagenic primer a Location b Codon change New restriction site E241C  TTCCTTTGAAAtgtAAGCATGGGCG  710-735 GAG 3  TGT None V254C  CTGGAGCATtgcTTGCACCCATGGGTGG  751-778 GTG 3  TGC  Sty  (K242-H253)  TTATTCCCCTGAAAGAG c 707-723and 759-781Codons to aminoacids  Bsg I,  Eco NI,  Ppu MI,  Sty I GTCCTGCACCCTTGGGTGGCGT  242-253 deletedK249-IEG-R250 CGCCGGCGAAGattgaaggtCGACTCGAGCATG  737-747and 748-760Insert codons foramino acids  SaI  I,  Sgr  AIForward IEG between K249and R250K249-IEG-R250 GCTCGAGTCGaccttcaatCTTCGCCCGGCGAAC  757-748and 747-734 As above  Sal I,  Xho IReverseStart primer  TTTAACGATGATTCGTGGCGG   67 to   47 NoneEnd primer  GCTCATTTCTCTCCCTGATAAC  1298-1276 None a  All primers start at 5  . Mutated codons are sown in lowercase. The mutated codons introducing new restriction site are underlined andindicated consecutively from the 5   end. b Locations are relative to the initiation codon. The nhaA sequence appears in the GenBank™ data base (accession no. J03879). c Base pairs 724-758 were deleted.  pH-induced Conformational Change in NhaA 24618   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  6 h  t   t   p :  /   /   w w w . j   b  c  . or  g /  D o wnl   o a  d  e  d f  r  om   quencing in a Perkin-Elmer (Applied Biosystem Division) 492 (Procise)microsequencer system.  Protein Determination— Protein was determined according toRef. 25.  DNA Sequence— Sequencing of DNA was conducted by an automatedDNA sequencer (ABI PRISM TM 377, Perkin-Elmer). RESULTS The pH Profile of Trypsin Digestion of Pure Xa-His-tagged NhaA in DM— For affinity purification of NhaA, we have pre- viously engineered NhaA fused at its C terminus to two factor Xa cleavage sites and 6 His residues (designated henceforth Xa-His-tagged NhaA instead of pYG10; Ref. 21). Trypsin diges-tion of both native NhaA (18) and Xa-His-tagged NhaA (17) ineverted membrane vesicles showed identical pH profile, whichis similar to that of the activity of the antiporter, and thus hasbeen suggested to reflect a native active conformational changeof the antiporter. Both activities are shut off at acidic pH andincrease above pH 7 to reach the maximum at pH 8.5. Thedigestion products of each protein are two main fragmentssimilar in size. Given that the trypsin digestion pattern probesa native conformation of the protein, it was interesting to usetrypsin digestion to test whether this conformation of NhaA ismaintained by the pure protein in DM. Xa-His-tagged NhaA was affinity-purified in DM, subjectedto digestion by trypsin at various pH values, and the productsseparated on SDS-PAGE (Fig. 1  A ). Upon alkalinization, twoprotein fragments appear in a pH-dependent fashion, similarin size to that previously obtained by trypsin digestion of iso-lated membrane vesicles (17). In the acidic pH digest, a singlefragment was observed, suggesting that the cleavage site,which splits the pure protein at alkaline pH, is masked. How-ever, when the separation on SDS-PAGE was prolonged (Fig.1  B ), the single band obtained at acidic pH was found to beslightly shorter than the control protein, implying the existenceof an additional exposed cleavage site(s).Since the His tag was fused at the C terminus of NhaA,binding to Ni 2  -NTA column was used to test whether the Cterminus is intact. Hence, Xa-His-tagged NhaA was treatedwith trypsin either at acidic or alkaline pH and the capacity of the products to affinity-purify on Ni 2  -NTA tested. Neither theacidic nor the basic pH digestion products bound to the column,implying that at both pH values the C terminus together withthe His tag was trimmed off. Similar results were obtainedwhen the trypsin treatment was conducted on everted mem-brane vesicles overexpressing Xa-His-tagged NhaA (data notshown).In contrast to Xa-His-tagged NhaA, native NhaA in isolatedmembrane vesicles did not seem to be cleaved by trypsin at itsC terminus (18); the size of the protein following acidic diges-tion did not change; Western analysis with a site-directed poly-clonal antibody against the C terminus recognized the appar-ently uncut protein obtained at acidic pH and the light trypticfragment obtained at alkaline pH. This antibody did not rec-ognize the intact Xa-His-tagged protein, most probably due tosteric hindrance of the tags. We thus assumed that the trypsincleavage site at the C terminus of Xa-His-tagged NhaA wasintroduced by the Xa tag. Indeed, as shown below, deleting thefactor Xa sequences yielded His-tagged NhaA, which is notclipped by trypsin at the C terminus.The N terminus remained intact following trypsin digestionof pure Xa-His-tagged protein; as shown below, the large frag-ment obtained at alkaline pH starts with the srcinal N-termi-nal sequence of the native protein (Fig. 2). The Unique Trypsin Cleavage Site Exposed by pH in Pure Xa-His-tagged NhaA Is in Loop VIII–IX— To identify the tryp-sin cleavage site that is exposed by pH and splits NhaA, each of the two fragments (heavy and light) obtained from the trypsindigest of Xa-His-NhaA at alkaline pH (Fig. 1) were isolatedfrom the gels and subjected to N-terminal sequencing (Fig. 2).The N terminus of the heavy fragment was found (with lessthan 1% contaminations) identical to that of the native protein.The light fragment fraction contained mainly (about 90%) afragment with a N-terminal sequence that overlaps a sequencebetween Arg-250 and Leu-264 (Arg-250–Leu-264) of loop VIII–IX. In addition, it contained small amounts (about 5% each) of two additional peptides with sequences overlapping Ser-246– Ala-259 in loop VIII–IX and Val-50–Asn-64 in loop I-II, respec-tively. Hence, the main cleavage site of trypsin that is exposedat alkaline pH is in Lys-249 of loop VIII–IX.  Lys-249 in Loop VIII–IX Is Also the Site Exposed to Trypsinin Situ at Alkaline pH in Isolated Membrane Vesicles— Thesimilarity in the pH profile of the trypsin digestion and the sizeof the products have suggested that the cleavage site  in situ  ineverted membrane vesicles is identical to that of the pureprotein in DM. However, proving this suggestion was difficultsince it is a very laborious task to purify the products that lostthe His tag during the  in situ  digestion by trypsin, from themembranous fraction. We therefore constructed a plasmid(pMXH) from pAXH, which encodes His-tagged NhaA with nofactor Xa cleavage sites. Everted membrane vesicles isolatedfrom cells overexpressing this protein exhibit a Na   /H  anti-porteractivityandapHprofileidenticaltothatofthewild-typeprotein (data not shown).These everted membrane vesicles were exposed to trypsinboth at acidic (data not shown) and alkaline pH (Fig. 3), thetreated membranes solubilized in DM, the solubilized fractionaffinity-purified on Ni 2  -NTA column. and the eluted polypep-tides separated on SDS-PAGE (Fig. 3). Following treatment at F IG . 1.  Digestion of purified Xa-His-tagged NhaA by trypsin as a functionof pH.  Xa-His-tagged NhaA was affinity-purified in DM (0.1%), subjected to tryp-sin at the indicated pH ( lanes b–g ), and5-  g protein samples run on SDS-PAGEfor 1 h in  A .  Lane a , molecular size mark-ers.  Lane h , the trypsin inhibitor added attime 0.  H.F. , heavy fragment;  L.F. , lightfragment.  B ,  lane a , molecular size mark-ers; lanesb and c containedsamplesofthetrypsin digest at pH 6.5, but the formerobtained the trypsin inhibitor at time 0.The SDS-PAGE was run for 90 min.  pH-induced Conformational Change in NhaA  24619   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  6 h  t   t   p :  /   /   w w w . j   b  c  . or  g /  D o wnl   o a  d  e  d f  r  om   acidic pH, the protein purified on the column behaved in SDS-PAGE in a fashion similar to that of the undigested control: oneband, of a size identical to the untreated control. These resultssuggest that following trypsinolysis the His tag remains intactin this recombinant protein and allows the affinity purificationof the protein. The results also show that indeed the trypsincleavage site at the C terminus of the Xa-His-tagged NhaA reside in the factor Xa cleavage site. When alkaline pH digestof His-tagged NhaA was applied to the column, two fragmentswere affinity-purified on the Ni 2  -NTA (Fig. 3,  lanes b  and  c ):a heavy fragment similar in size to that observed following treatment of Xa-His-tagged NhaA, and a slightly shorter lightfragment as expected on the basis of the difference between thetwo C termini of Xa-His-tagged and His-tagged NhaA. In con-trast, none of the tryptic polypeptides derived from Xa-His-tagged NhaA subjected to the same treatment were purified bythe column (Fig. 3,  lane e ). It is remarkable that despite thetrypsin split in His-tagged NhaA two fragments co-purify viathe His tag: the C-terminal fragment with His tag and theN-terminal fragment without it. This result implies that bothC-terminal and N-terminal fragments formed by the trypsinsplit are bound to each other and do not separate in DM.To verify that the trypsin cleavage site of His-tagged NhaA  in situ  (in the membrane) is identical to that of the pureprotein, the light C-terminal fragments obtained from the tryp-sin digest of His-tagged NhaA membranes were isolated bySDS-PAGE and subjected to N-terminal sequencing. The re-sults show that indeed the trypsin cleavage site that is exposed in situ  at alkaline pH is identical to that identified in the pureprotein in DM, namely Lys-249 of loop VIII–IX. Also similar tothe pure protein is a minor split occurring   in situ  at Arg-245.  Mutations in Loop VIII–IX Affect the pH Regulation but Notthe Activity of NhaA— Since loop VIII–IX changes its conforma-tion with pH, the question arises as to whether loop IV-IX playsany role in the activity of NhaA or its regulation by pH.To answer this question, three types of mutations have beenintroduced to loop VIII–IX.The first type is a deletion mutation lacking 12 amino acidsfrom Lys-242 to His-253 (  Lys-242–His-253). This mutant wasexpressed to a very low level (2% of the expression of the wildtype; Table II) and did not grow in the presence of 0.6  M  NaCl T  ABLE  II  Expression of loop VIII-IX NhaA mutants  nha  A   nha B cells transformed with the indicated plasmids weregrown to  A 600    0.7 in LBK. The membrane fraction, prepared bysonication (16), was treated with 6  M  urea and then used for Westernanalysis. To compare the level of expression, densitometry was used;pGMAR100, which harbors wild-type  nha  A, and all other plasmidscarrying mutated  nha  A are pBR322 derivatives. Plasmid Expression % pGMAR100 (wild-type) 100pE241C 120pV254C 110pE241CV254C 100pK249-IEG-R250 100p  (K242-H253) 2p  (K242-H253)P257S 4F IG . 2.  The unique cleavage sites exposed at alkaline pH in purified Xa-His-tagged NhaA.  A model of the secondary structure of NhaA with the proposed 12 transmembrane segments (28). The major ( large arrow ) and the minor ( small arrow ) trypsin cleavage sites are shown. Thecorresponding determined N-terminal sequences are  shaded . Residues involved in pH regulation are shown in  enlarged bold  characters.  pH-induced Conformational Change in NhaA 24620   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  6 h  t   t   p :  /   /   w w w . j   b  c  . or  g /  D o wnl   o a  d  e  d f  r  om   either at pH 7 or at pH 8.3. As measured in isolated membrane vesicles, the mutant did not show any Na   /H  antiporter ac-tivity at pH 7 but at pH 8.5 a very low but reproducible activitywas monitored (Fig. 4). Whereas, with the wild-type mem-branes, 100% of dequenching of the fluorescence was obtainedwithin 30 s, 30% of this activity was obtained by the mutantonly after 10 min. This low activity was ascribed to the mutant-NhaA since the control membranes derived from cells trans-formed with the vector (pBR322) did not show any activity. A spontaneous suppressor mutation was obtained by grow-ing    Lys-242–His-253 at pH 7 in the presence of 0.6  M  NaCl.The mutation was cloned and identified as P257S in  Lys-242–His-253 (  (Lys-242–His-253) P257S). The suppression wasonly partial since the second-site mutation restored growth inthe presence of Na  only at neutral pH but not at pH 8.3. Theexpression of the   (Lys-242–His-253) P257S  nha  A was 2-foldhigher as compared with the srcinal mutant (Table II). Al-though similar to mutant   (Lys-242–His-253), mutant   (Lys-242–His-253) P257S did not show any antiporter activity ineverted membrane vesicles at pH 7; its activity at pH 8.5 wassubstantially higher in rate but not in extent as compared withthe srcinal mutant (Fig. 4). This phenotype most probablyaccounts for the improved growth of the double mutant at pH 7in the presence of Na  as compared with the srcinal mutant,which did not grow. Interestingly, despite their low activity,both mutants show pH sensitivity being inactive at pH 7 andactive at pH 8.5 (Fig. 4).The second type of mutation was an insertion mutation;amino acids IEG were inserted between Lys-249 and Arg-250 of loop VIII–IX, creating factor Xa protease cleavage site (IEGR).This insertion mutation, designated K249-IEG-R250, was ex-pressed as good as the wild-type (Table II) and allowed us tocreate very specifically a single split in NhaA in loop VIII–IX with factor Xa protease at the same site as trypsin. The size of the fragments obtained following factor Xa digestion were in-deed as expected for a unique split in loop VIII–IX (data notshown).The Na   /H  antiporter activity in everted membrane vesi-cles containing K249-IEG-R250 NhaA was similar in its max-imalactivitytothewild-typecontrol(Fig.5  A ).However,thepHprofile of the activity of the mutant was shifted by about half apH unit toward acidic pH (Fig. 5  A ). Hence, the insertion mu-tation into loop VIII–IX affects the pH sensitivity of NhaA.The third mutation was three point mutations, each in aseparate plasmid; these were introduced to loop VIII–IX of NhaA by site-directed mutagenesis: E241C, V254C, and F IG . 4.  Na   /H  antiporter activityof deletion mutations in loop VIII–IX. Everted membrane vesicles were isolatedfrom EP432 transformed with the vectorplasmid (pBR322) or plasmids harboring wild-type NhaA ( w.t .) or   Lys-242–His-253 or   (Lys-242–His-253)P257S muta-tions as indicated. The   pH was moni-tored with acridine orange (0.5   M ) inmedium containing 140 m M  KCl, 10 m M Tricine (at the indicated pH), 5 m M MgCl 2 , and membrane vesicles (50–100  g of protein). At the onset of the experi-ment, Tris  D -lactate (2 m M ) was added( closed arrow ), and the fluorescencequenching was recorded until a steadystate level of   pH has been reached. NaCl(10 m M ) was then added ( openarrow ), andthe new steady state of fluorescence ob-tained (dequenching) was monitored. Allexperiments were repeated at least twice,and the results were essentially identical.F IG . 3.  The pH-sensitive trypsincleavage site in everted membranevesicles  in situ  resides in loop VIII–IX.  Everted membrane vesicles were iso-lated from EP432 cells transformed withplasmid pI Q and either pMXH or pAXHoverexpressing His-tagged or Xa-His-tagged NhaA, respectively. The mem-branes (2 mg/ml protein) were suspendedin 140 m M  KCl, 10 m M  Tricine (pH 8), 5m M  MgCl 2 , and 40   g/ml trypsin and in-cubated for 1 h at 37 °C. Then the pro-teins were affinity-purified on Ni 2  -NTA column and run on SDS-PAGE.  Lane a ,molecular size markers;  lanes b  and  c ,samples from reaction mixtures contain-ing His-tagged NhaA;  lanes d  and  e , Xa-His-tagged NhaA. In  lanes b  and  d , thetrypsin inhibitor was added at time 0.  pH-induced Conformational Change in NhaA  24621   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  6 h  t   t   p :  /   /   w w w . j   b  c  . or  g /  D o wnl   o a  d  e  d f  r  om 
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks