Isolation of Prokaryotic VoVl-ATPase from a Thermophilic Eubacterium ...

THEJOURNAL OF BIOLOOICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 16, Issue of April 22, pp. 12248-12253, 1994 Printed in U.S.A.

Isolation of Prokaryotic VoVl-ATPasefrom a Thermophilic Eubacterium Thermus thermophilus* (Received for publication, November 29, 1993,and in revised form, January 21, 1994)

Ken Yokoyama,Yasuhiro AkabaneS, Noriyuki Ishii§, and Masasuke Yoshidan From the Research Laboratory of Resources Utilization, Ibkyo Institute of Technology, Nagatsuta 4259, Yokohama 227, $Surface Science Research Center, Lion Corporation, Hirai 7-13-12, Edogawa, Tokyo 132,and the $Institute of Physical and Chemical Research, Hirosawa 2-1,Wako 351-01,Japan

The soluble ATPasepurified from an aerobic thermophilic eubacterium, Thermus thermophilus, was not a usual F,-ATPase but a Vl-ATPase,a peripheral section of plasma membrane V-type ATPase (Yokoyama, K., Oshima, T., and Yoshida, M. (1990) J. Biol. Chem. 265, 21946-21950). Here, we report thepurification of V-type ATPase from the same bacterium (V,V,-ATPase) which consists of VI-ATPase and a membrane-integrated section, V,. The VoVl-ATPase, either in the Triton X-100-solubilized membrane fraction or in the purified state, migrates as a single band in a non-denaturing polyacrylamide gel electrophoresis for membrane protein complexes, and eight kinds of polypeptides are found when this band is developed in a second dimensiondenaturing gel electrophoresis in the presence of sodium dodecyl sulfate. The 66-, 56-, 30-, and 12-kDa polypeptides are the subunits of Vl-ATPaseand the loo-, 38-,24-, and 13-kDa polypeptides are candidates for V, (or V,associated) subunits. The amino-terminalamino acidsequences of the 38-and 24-kDa subunits do not show obvious similarity to any subunits of eukaryotic V,VlATPases. The kinetic properties of the purified V,VlATPase are very similar to those of Vi-ATPase:a very low ATPase activity, stimulation by bisulfite, inhibition by nitrate, and resistance against inhibitors of eukaryotic V-type ATPases, BafilomycinA1 and N-ethylmaleimide. The V, vesicles prepared from reconstituted V,V,-ATPase vesicles by 6 M urea treatment show the H’ channel activity. This H’ channel activity is abolished either by treatment of vesicles with dicyclohexylcarbodiimide or by the back addition of VI-ATPase. These results indicate that thecoupling ionof this V,V,-ATPase is H+.

(9). The eukaryotic V-type ATPase has an apparentfunctional mass of 400 - 600 kDa and comprises at least 9 different subunits. For example, bovine clathrin-coated vesicle V-type ATPase is composed of subunits with apparent molecular sizes, 116,73,58,40,38,34,33,20, and kDa 17 (10). The 73-, 58-, and 17-kDa subunits are homologous to the p, a , and c subunits of F,F,-ATPase. Similar to the F,F,-ATPase, which can be separated into a water-soluble F, section and a membrane-integrated F, section, the V-type ATPase is composed of a watersoluble set of subunits (termed V,) and a membrane-integral set of subunits (termedVJ. The 73-, 58-, 40-, 34-, and 33-kDa subunits belong to V, section and the116-, 38-, 20-, and 17-kDa subunits arecontained in V, section (11,121. However, different from F,F,-ATPase, once V, is detached from V, section, VI dissociates into individual subunits with accompanying loss of ATPase activity, and V, does not show the activity as a H’ channel (13). In prokaryotic cells, although it hadlong been believed that bacterial plasma membrane H+-translocatingATPases are all F,F,-ATPases, recent studies have revealed the presence of V-type ATPases a t first in archaebacteria(14-19), and then in two eubacteria, Thermus thermophilus (20,211 andEnterococcus hirae (22,23). All the purified preparations reported above are V, sections of V-type ATPases and, different from eukaryotic equivalents, prokaryotic VI is a stable ATPase-active complex termed as V,-ATPase. The isolation of intact prokaryotic V-type ATPase has been attempted for three archaebacteria, Methanosarcina barkeri (18),Methanothrir thermophila (24), and Sulfolobus acidocaldarius(17). However, since V, section is detached from membrane very easily, homogeneous V,V,ATPase a t enough quantity and stabilityto allow biochemical characterization has not been available and this limits our knowledge on prokaryotic V,V,-ATPase. For example, subunit F,F,-ATPase and V-type ATPase are two subclasses of a su- composition of prokaryotic V-type ATPase has been still unperfamily of H+-translocatingATPases (1,2). F,F,-ATPases are clear. Only a single kind of polypeptide, of which amino acid responsible for ATP synthesis coupled with H’ translocation sequence is homologous to the c subunit of F,, has been so far one of prokaryotic V, subunits from across membranes, and, ineukaryotic cells, they are present in identified with certainty as inner membranesof mitochondria and thylakoid membranes of a deduced sequence of the V-type ATPase operon of S. acidochloroplasts. On the other hand,V-type ATPases pump H’ into caldarius (25) andfrom analysis of the [14C]dicyclohexylcarboeukaryoticvacuolar vesicles and acidify thelumen of the diimide-labeled polypeptide (17, 26). In a previous study (20), vesicles. They are found inmembranes of clathrin-coated we reported the purification of V,-ATPase from a n aerobic thervesicles (3), chromafin granules (41, kidney microsomes (5), mophilic eubacterium, I: thermophilus. The I: thermophilus plant vacuoles (6,7), andvacuoles of Neurospora ( 8 ) and yeast VI-ATPase comprises four subunits, a (64 kDa), p (53 kDa), $30 kDa), and 6 (12 kDa). The amino acid sequences of the a and p subunits were deduced from nucleotide sequence (21). * This work was supported by a grant-in-aidfor scientific research in the priority area of no. 04266103 (to M. Y.) from the Ministry of Edu- Here, we reportthe purification, subunit composition, and cation, Science and Culture of Japan. T h e costs of publication of this characteristics of the V-type ATPase from I: thermophilus article were defrayed in part by the payment of page charges. This plasma membrane. The demonstrationof H’ channel activity of article must therefore be hereby marked “advertisement”in accordance V, vesicles is also reported. The advantage of non-denaturing with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondence should be addressed: Research Laboratory polyacrylamide gel electrophoresis in the presence of an alkyl of Resources Utilization,Tokyo Institute of Technology, Nagatsuta 4259, ether sulfate and a tertiary alkylamine oxide, for analysis of membrane protein complexes is pointed out (27). Yokohama 227, Japan, Fax:81-45-922-5179. 12248

12249

Prokaryotic V,V,-ATPase EXPERIMENTALPROCEDURES Materials -T thernophilus strain HB8 was grownas described previously (20). Softes-12 was purchased from Wako Junyaku (Osaka). Nucleotide triphosphates were purchased from Kyowa Hakkow Corp. Other reagents were purchased from Sigma. F,-ATPase and F,F,ATPase from a thermophilic Bacillus PS3 were purified as described elsewhere (28). Purification of ATPase-Cells (200 g) harvested at log phase growth were suspended in 400 ml of 50 mM Tris-C1, pH 8.0, containing 5 m MgCl, and disrupted by sonication. Membranes were precipitated by centrifugation at 100,000x g for 15 min and washed twice by centrifugation with the same buffer. Sonication and centrifugation were performed a t 4 "C, and the washed membranes were stored at -85 "C until use. The washed membranes were thawed and suspended in 150 ml of 50 m Tris-C1, pH 8.0, containing 5 m~ MgCl,, 200 m N&SO,, and 2% Triton X-100.Batches of 50 ml of the suspension were sonicated for 2 min at 0 "C. Then they were centrifuged at 100,000 x g for 60 min, and the amber supernatant was dialyzed for3 h against10-fold volumesof 50 m Tris-SO,, pH 8.0,5 m~ MgSO,, 0.5% Triton X-100 (BufferA). The dialyzed solution was applied to a DEAEkellulose column (inside diameter 3x 30 cm)equilibrated with Buffer A. The column waswashed successively at a flow rate 400 mVh with 500 ml of Buffer A, 300 ml of Buffer A plus 50 m~ NqSO,, and 500 ml of Buffer A plus 200 m~ Na,SO,. The ATPase-active fractions eluted at the 50 m~ Na,SO, washing step were collected (50 ml) and concentrated to 5 ml by ultrafiltration with an Amicon X"300 membrane. The concentrated solution was applied to a Sepharose GB-CL column (inside diameter 2 x 90cm) equilibrated with 50 m~ Tris-SO,, pH 8.0, 5 m MgSO,, 0.25% Triton X-100, 200 m~ Na,SO, and was eluted with the same buffer (50 mVh). The active fractions were combined and concentrated to 1 ml with a Centricon 30 microconcentrator (Amicon). The purified V,V,-ATPase was stored at -85 "C. Gel Permeation HPLC-The purified VoV,-ATPasewas loaded on a TSK G-3000SWZ column equilibrated with 50 m Tris-SO,, pH 7.2, 5 m~ MgSO,, 200 m~ NqSO,, and 1% @xtylglucoside. The column was eluted at a flow rate of 0.5 mYml at room temperature. ATPase activities of the eluted fractions were assayed, and fractions were analyzed with polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE). ATPase and Protein Assays-ATPase activity was assayed by the release of inorganic phosphate from ATP at 55 "C. The reaction mixture contained 50 m~ Hepes-Na, pH 7.5, 5 m MnCl,, 5 m ATP, 200 m~ Na,SO,, and thereaction was initiated by the addition of ATP. One unit was defined as the activity that releases 1 pmol of P,/min. Protein concentrations were determined with the method of Bradford (29) with bovine serum albumin as a standard. Electrophoresis-Polyacrylamide gel electrophoresis in the presence of an alkyl ether sulfate and a tertiary alkylamine oxide (AES-PAGE)' was carried out as described previously (30) with some minor modifications. Briefly, concentrations of gels were 6% for separating gels and 4% for stacking gels. Gels contained 0.1% Softes-12 (Wako Chemicals, Tokyo). Softes-12 is a 10%solution of detergents, a mixture of 7% of an alkyl ether sulfate (sodium oligooxyethylenedodecyl ether sulfate, average number of oxyethylene unit = 7) and 3% of a tertiary alkylamine oxide (dodecyldimethylamine oxide), and 0.1%Softes-12 means one hundredth dilution of the original Softes-12 solution. The samples were mixed gently with equal volume of the sample buffer containing 1% Softes-12, 0.1% bromphenol blue, 100 m~ Tris-C1,pH 6.8, and 10% glycerol, and were applied to a slab gel (10 x 10 cm). The gelwas electrophoresed at a constant current of 20 mA for 2 hat 4 "C. Proteins in the gels were stained with Coomassie Brilliant Blue R-250. When necessary, gels were stained with ATPase activity (28). AFS-SDS twodimensional PAGE was carried out as follows. The firstAES-PAGE was performed as described above. After electrophoresis, each sample lane of the gel wascut into 10 x 100-mm pieces,and pieces were boiledin 1% SDS, 0.1% bromphenol blue, 100 m Tris-C1, pH 6.8, for 5 min. The boiled gel piecewas put on top of the SDS slab gel and fixed with the 4% acrylamide stacking gel. The second dimension of electrophoresis was performed according to the procedure of Laemmli (31) on 12% polyacrylamide gels with a 4% stacking gel containing 0.1% SDS.

Preparation of V,V,Vesicles-The vesicles were reconstituted with the dialysis method (32). The purified ATPase (0.3 mg) and phospholipids from soybeans (40mg) were dissolvedin 1ml of (final concentration) 1%cholate, 1%deoxycholate, 5 m dithiothreitol, 1m EDTA, and 10 m~ Tris-SO,, pH 8.0. The solution was sonicated in a bath sonicator for 5 min to make the solution translucent. The solution was dialyzed overnight a t room temperature against 4 liters of 10 m Tris-SO,, pH 8.0, 5 m dithiothreitol, and 1m EDTA (Buffer B). The dialyzed solution was centrifuged at 100,000 x g for 30 min a t 4 "C, and the pellet was resuspended at a protein concentration of 1 mg/ml in Buffer B. ElectronMicrograph of V,V, Vesicles-The suspension of VoV, vesicles prepared as above was applied on a copper grid covered with a carbon support film which was made hydrophilicby ion bombardment. After the negative staining with phosphotungstic acid (pH 7.251, the gridwas examined with transmission electron microscope (JEOL JEM2000EX) with an accelerating voltage of 160 kV. Images were recorded ontothe Kodak electron image film (50-163) at magnification of 41,800. H+Channel Activity of V, Vesicles-The suspension of V,V, vesicles was centrifuged again and thepellet was suspended in 1ml of Buffer B containing 6 M urea to induce denaturation and dissociation of V, subunits. After incubation at 4 "C for 4 h, thesuspension was centrifuged, and the V, vesicles were collected as a pellet. The V,, vesicles were suspended in 1 ml of 0.5 M KC1 containing 1 m dithiothreitol. The suspension was incubated for 30 min at 55 "C, then chilled in an ice bath, and mixed with 10 pl of 1 M MgSO,. The K+-loaded vesicles were collected bycentrifugation, and was suspended in 500 pl of 0.5 M sucrose and 2.5 m~ MgSO,. Then, 50 pl of K+-loadedV, vesicles weresuspended in 2.5 ml (a final volume) of 4 p~ 9-aminoacridine, 0.25 M sucrose, 2.5 m MgSO,, 10 m~ Tricine-Na,pH 7.5.Assays of H channel activity was started by the addition of 10 ng of valinomycin.The fluorescencechange was monitored with PTL 396s(Jasco) fluorometer using the wavelength of 365 nm for excitation and 451 nm for emission. The medium in the quartz cuvette was maintained at 30 "C by circulating water. Protein Sequencing-The purified V,V,-ATPase (20 pg) was applied to the 14% SDS-PAGE.After electrophoresis,the protein in the gel were electrophoretically transferred to polyvinylidene difluoridemembrane. The protein bands on the membrane were stained with Coomassie Brilliant Blue R-250 and cut out fmm the membrane, and the sequences were analyzed with an Applied Biosystems 470 Sequencer (33). RESULTS

Identification of V,V,-ATPase-Different from eukaryotic Vtype ATPases, the peripheral V, section of prokaryotic V-type ATPases is easily released into soluble fraction during the preparation of membranes (14,151.For the purpose of isolation of intact V-type ATPase (V,V,-ATPase),a method to differentiate V,V,-ATPase from soluble V,-ATPase was required since pursuit of the ATPase-activefraction could result with the isolation ofV,-ATPase. We found that AES-PAGE, non-denaturing polyacrylamide gelelectrophoresis for membrane protein complexes, met this requirement. When a Triton X-100-solubilized membrane fraction from I: thermophilus was analyzed with this method, no band appeared at the position of VI-ATPase and, instead, a discrete band with a slower mobility was observed (Fig. l A ,indicated by an arrow). The activity staining of the gel showedthat thisband had an ATPase activity (Fig. lB ). The AES-SDS two-dimensionalPAGE of the same preparation indicated that this band appeared to contain all subunits of V,-ATPase and additional several polypeptides (Fig. 1C).These results suggest that thelow mobility band is V,,V,-ATPase. We adopted AES-PAGE as a method to identify V,V,-ATPase during purification procedures. Purification of V,V,-ATPase-A number of detergents were tried for the solubilization of the V,Vl-ATPase from I: thermophilus membranes, andthe mostsuccessful was Triton X-100. Other detergents, such as deoxycholate, cholate, and The abbreviations used are; AES-PAGE; polyacrylamide gel electro- P-octylglucoside, solubilizedthe ATPase activity at high yield phoresis in the presence of an alkyl ether sulfate (sodium oligooxyeth- from membranes, but AES-PAGE analyses showed that the ylenedodecyl ether sulfate) and a tertiary alkylamine oxide(dodefractions solubilized by these detergents contained significant cyldimethylamine oxide); CCCP, carbonylcyanidem-chlorophenylhydrazone; DCCD,N,N"dicyclohexylcarbodiimide; PAGE, polyacrylam- amount of released V,-ATPase. The Triton X-100-solubilized ide gel electrophoresis; HPLC, high pressure liquid chromatography. fractions were applied onto a DEAE-cellulosecolumn, and

V,V,-ATPase

12250 1

2

1

2

vov,

‘I

Prokaryotic VI

‘I

PIES-

-. F

vov,+

1‘:

v)

- v )

‘1 A.CBR

E. ATPase

‘’

FIG.1. A n a l y s i s of Triton X-100-solubilized membrane f r a c t i o n of T. thermophilun w i t h ( A , R ) hES-PAGE and w i t h ( C )AES-SDS t w o - d i m e n s i o n a l PACE. Gels wrrr stnincd with (:onmassir Hrilllant Blur It-250 1.4, ( ’ ) or wlth ATPasr activity In). l a n e I of A and 13, purified 7: thrmmphilus VI-ATPasr (10 pgl; lnnr 2 of A and B , Triton X-100-soluhilizrdmrmhranrfraction I20 pg); C . TritonX-100-soluhilized mrmhranr fraction (50 pgl. Elcctrophorrgrams shown on top of and at right side of the AES-SDS two-dimensional PACE arr AESPAGE oftheTritonX-100-soluhilizrdmrmhranefractionand SI)% PAGE of thr purifird 7: Ihrrmophilus V,-ATTPasr (10 pgl, resprctivrly. Positions ofV,-ATPnsc and V,V,-ATPasr arc indicated hy arrows. Othrr drtailrd conditions arr drscrihrd undrr “Experimmtal Procrdurrs.”

ATPase activity was eluted with a stepwise increase of Na,SO,, (Fig. 2). ATPase activities were eluted both at 50 mM Na,SO,l J 24kand 200 mM Na,SO,. AES-PAGE analysis revealed that V,V,ATPase wasa major component in the fractions eluted a t 50 m v Na,SO,,, and the fractions eluted at 200mM Na,SO, contained VI-ATPase (Fig.,?A,lanes 1 and 2).Since Triton X-100-solubi1, lized membrane fractions did not contain VI-ATPase (Fig. lane 2),dissociation of some population of V,V,-ATPase into VI-ATPase andV, section should occur during column chromatography. Total activity eluted a t 50 m>f Na,SO, and 200 mM Nn,SO, was about doubleof the original activity applied to the column. The reason for this activation is not known. The fractions eluted at 50 mM Na,SO,l were further purified with Sepharose GB-CL column chromatography and the ATPase thus purified migrated as an apparrntly single hand in AES-PAGE (Fig. 3A, lane 3 ) . The spccific ATPase activity of the purified 1.9 units/mg of protein, a comparable V,,V,-ATPase is 1.7 value to the specific activity of V,-ATPase (2.0 unitslmg) (20). V,V,-ATPase is an abundant membrane protein (Fig. l A , lanr 2 ), and about 20 mg of purified enzyme were obtained from 200 g of cells. The purified V,,V,-ATPase is stable in ammonium sulfate precipitation. Purification procedures are summarized in Table I in which the figures are averages of the three preparations. Subunit Composition a n d Molecular Sizr-Analysis of V,V,Fraction ATPase with SDS-PAGE shows that it contains eight kinds of polypeptides (Fig. 3B3,lane I ). Four of them are subunits of rrn1ts I l n l f c ‘WIX ‘: w V,-ATPase, that is, polypeptides with molecular sizes of 65, 54, 4R n.nS.1 1on 1 xno Solubilizrd fraction 30, and 12 kDa correspond to n, p, y , and fi subunits, respec1 of; 0.7 I 1fIH 15n DEAR crllulosr :I4 1.7 :15 Scpharosr C L 6 B 2‘0 tively(Fig. 3B, lane 2 ) (20).Fouradditionalpolypeptides, whose apparent molecular sizes were 100, 38, 24, and 13 kDa, are candidates for V, (or V,-associated) subunits. The subunit (data not shown). The protein with thc snmr clrctrophorctic mobility as the 13-kDa polypeptide was cxtrnctcd from mrmcompositions of I: therrnophilus V,V,-ATPase and V,-ATPase branes of I: thrrrnophilus with chroloform-mcthnnol (dntn not are evidently different from those of usual F,F,-ATPase and F,-ATPase, respectively (Fig. 3B, lanes 4 and 5 ). The 20 amino- shown).Probably, it corrcsponds to t h r protrolipidsuhunit terminal amino acid sequences of the 24- and 38-kDa polypep- which is homologous to t h r F