Thermus thermophilus Membrane-associated ATPase INDICATION OF A EUBACTERIAL V-TYPE ATPase*

An ATPase with M, of 360,000 was purified from plasma membranes of a thermophilic eubacterium Thermus thermophilus, and was characterized. ATP hydrolytic activity of the purified enzyme was extremely low, 0.07 pmol of Pi released mg-’ min-‘, and it was stimulated up to 30-fold by bisulfite. The follow- ing properties of the enzyme indicate that it is not a usual F1-ATPase but that it belongs to the V-type ATP- ase family, another class of ATPases found in mem- branes of archaebacteria and eukaryotic endomem-branes. its

((Y), 55,000 (b), 30,000 (y), and 12,000 (a), the a subunit had a similar molecular size to the catalytic subunits of the V-type ATPases but was significantly larger than the (Y subunit of F1-AT-Pases. ATP hydrolytic activity was not affected by azide, an inhibitor of F1-ATPases, but was inhibited by nitrate, an inhibitor of the V-type ATPase. N-terminal amino acid sequences determined for the purified (Y and fl subunits showed much higher similarity to those of the V-type ATPases than those of F1-ATPases. Thus the distribution of the V-type ATPase in the prokaryotic kingdom may not be restricted to archaebacteria. FOF1-ATPases are responsible for ATP synthesis coupled with H' translocation across membranes. They are purified from a variety of sources such as mitochondria, chloroplasts, and bacteria and have remarkably common characteristics (l-8). A water-soluble moiety with ATP hydrolytic activity is easily detached from F,,F1-ATPase. This moiety, F,-ATPase, is composed of three copies of two major subunits, a (55-58 kDa) and p (51-55 kDa), and a single copy of three minor subunits (1,3,4,9). Amino acid sequences of the CY and /3 subunits have been highly conserved during evolution (1). FoF1-ATPase was once considered to be the only enzyme responsible for ATP synthesis coupled with H' flow in any type of cell. This notion has been challenged by the recent finding that archaebacteria, such as Sulfolobus acidocaldarius ( 12,14). Comparison of amino acid sequences deduced from DNA sequences finally established that archaebacterial ATPases and eukaryotic vacuolar H'-ATPases belong to a single genetically related ATPase family, which is now called V-type ATPase (15)(16)(17)(18) and that V-type ATPase and FOF1-ATPase are two subclasses of a superfamily of ATPases. It has also been suggested that the archaebacterial cx subunit and eukaryotic vacuolar H'-ATPase 66-70 kDa subunits are catalytic subunits and, therefore, correspond to the p subunit of FI-ATPase. All archaebacterial ATPases reported up to now belong to the V-type ATPase, while this ATPase has not been found in any eubacteria, that is prokaryote, other than archaebacteria. On the basis of these observations, the evolution of H+-ATPases and the evolutionary relationship between archaebacteria and eukaryotes have been postulated (15)(16)(17)(18)(19). However, the possibility still remains that eubacteria from which isolation of FoFI-and F1-ATPases has as yet been unsuccessful may contain V-type ATPase.
Here we report the isolation and characterization of an ATPase from a thermophilic eubacterium Thermus thermophilus which most probably belongs to the V-type ATPase family.

MATERIALS AND METHODS
Preparation of Membranes-T. thermophilus strain HB8 (ATCC 27634)' was grown at 75 "C under strong aeration in a medium containing 10 g of yeast extract, 10 g of polypeptone and 2 g of NaCl per liter (20). Cells (200 g) harvested at the log phase were suspended in 400 ml of 50 mM Tris-SO, buffer, pH 8.0, containing 5 mM MgCl, and disrupted by sonication. Membranes were precipitated by centrifugation at 100,000 X g for 15 min and washed twice by centrifugation with the same buffer.
At an early stage of this study, a low speed centrifugation (5,000 x g, 20 min) was carried out to remove cell debris. However, since we learned later that omission of this step improved the yield of the enzyme in the next chloroform treatment by about SO%, the disrupted cell suspension was directly applied for high speed centrifugation.

Purification of T. thermophilus
ATPose-The ATP-hydrolyzing activity of the membrane fraction from T. thermophilus was 0.09 unit/mg protein or 1,200 total units for membranes prepared from 200 g of cells. Since this activity was not stimulated by Na2S03 and about 10% of the total activity was extracted by chloroform treatment, only a small fraction of this activity corresponded to the specific ATPase that we purified. Similar to mitochondrial F1, vigorous stirring of the membrane suspension with chloroform proved to be effective in releasing ATPase from T. thermophilus membranes (23). Treatment with a low ionic strength buffer containing 5 mM EDTA was also effective in detaching ATPase from the membranes, although the yield was approximately half that of the chloroform treatment.
Butyl-Toyopearl column chromatography was the most effective step for purification (Table  I). Total activity increased slightly after this purification step, but the reason is not known. The content of ATPase in the cell, 30-40 mg/lOO g of wet cells, was as abundant as F1-ATPase in usual bacteria (3,26). The purity of the ATPase was confirmed by a gel permeation HPLC and by polyacryl-  amide gel electrophoresis in the absence of SDS. The protein was eluted from a column as a single peak which had ATPase activity (Fig. 1) and migrated as a single protein band in a gel (Fig. 2, left panel).

Molecular
Weight and Subunit Composition-In order to estimate molecular weight, the purified ATPase was analyzed with a gel permeation HPLC column (G-3000 SWXL, Tosoh) (Fig. 1). The T. thermophilus ATPase was eluted slightly later than F1-ATPase from a thermophilic bacterium PS3 (TF,), and its molecular weight was estimated to be 360,000. Polyacrylamide gel electrophoresis in the presence of SDS revealed that the purified ATPase consisted of at least four kinds of polypeptides, designated as the (Y, @, y, and 6 subunits in order of the decreasing molecular weight, and their apparent molecular weights were estimated to be 66,000, 55,000, 30,000, and 12,000, respectively, from the mobility in the gel (Fig. 2,  Left panel, polyacrylamide gel electrophoresis in the absence of SDS. Thirty micrograms of the purified T. Enzymatic Properties-The optimum pH for ATP hydrolytic activity of the T. thermophilus ATPase was pH 7.5, and the activities at pH 6.0, 6.5, 7.0, 8.0, 8.5, and 9.0 were 35, 35, 53, 67, 44, and 19%, respectively. The enzyme hydrolyzed ATP, GTP, ITP, UTP, and CTP at relative rates of 100, 55, 37,37, and 46%, respectively. Hydrolysis of UTP and/or CTP was reported for TF, and archaebacterial ATPases (3,10,12). It did not hydrolyze ADP and p-nitrophenyl phosphate. Divalent cations were required for the activity, since ATP was not hydrolyzed in the presence of 1 mM EDTA. Specificity for the requirements of divalent cations was rather broad; Mn*', Mg*+, Cd2+, Zn2+, Ca2', and Ni*' were effective as cofactors, and the relative ATP hydrolytic activities at 1 mM of the above cations were 180, 100, 72, 43, 40, and 21%, respectively.
Lineweaver-Burk plots for ATP hydrolysis by the T. thermophilus ATPase appeared to be linear in the ATP concentration range from 200 pM to 5 mM (data not shown).3 The K,,, and V,,, values were estimated to be 1.1 mM and 0.07 unit mg-', respectively, at pH 7.5 at 55 "C. When 200 mM Na2S03 was included in the assay mixture, the V,,, " ATPase activity at ATP concentrations lower than 200 pM was not measurable by this method. Enzyme-coupled assay with an ATP regenerating system using pyruvate kinase and lactate dehydrogenase was not applicable for this enzyme, since auxiliary enzymes were not stable at 55 "C in the presence of 200 mM Na2S03. It is not known if the linearity of the Lineweaver-Burk plots can be extended down to micromolar range concentrations of ATP.
increased to about 2.0 unit mg-', but the K, value was not changed. This kind of kinetic behavior, low specific activity in the absence of the activator anions, linear Lineweaver-Burk plots, and a large K, value have been reported for ATPases from archaebacteria (10)(11)(12)14). As expected, this enzyme was extremely heat-stable, and the maximum activity, 2.8 unit mg-', was observed at around 85 "C for a 5-min assay at 10 pg ml-' of the enzyme in the reaction mixture containing 200 mM Na2S03. However, it was inactivated completely after a lo-min incubation at 95 "C. Unlike the usual F1-ATPase, the purified enzyme (1 mg ml-') in 50 mM Tris-SO4 (pH 8.0) and 1 mM EDTA was not inactivated by overnight exposure to low temperature at 4 "C.

Effects of Ions and Inhibitors-Similar
to ATPases from archaebacteria (lo-12), but to a more pronounced extent, the inclusion of Na2S03 or NaHC03 in the assay mixtures caused remarkable activation of the ATP hydrolytic activity of the T. thermophilus ATPase (Fig. 3). High concentrations of these salts were required to gain maximum activation.
Since NaCl did not show such an activating effect, this activation was not due to Na'. Potassium, chloride, and sulfate ions were ineffective as activating ions. Since the extent of activation by Na2S03 decreased with the increase of pH (more than a 50fold activation at pH 6.5, 30-fold at pH 7.5, and B-fold at pH 8.5, for example) a truly effective ion species should be a bisulfite, but not a sulfite. Similarly, bicarbonate, but not carbonate, might be an effective anion. Sensitivity of the ATPase activity of the T. thermophilus ATPase to some of the specific inhibitors for other ATPases was examined (Table II). The ATPase activity of the T. thermophilus ATPase was neither inhibited by azide, an inhibitor of Fi-ATPases, nor by vanadate, an inhibitor of the ATPases that are characterized by the formation of a phosphorylated intermediate (27). Nitrate, known as an inhibitor of vacuolar H+-ATPases and archaebacterial ATPases (11, 28), showed a significant inhibitory effect on the activity of the T. thermophilus ATPase. It is known that covalent modification of a single glutamic acid residue of the /3 subunit of Fi-ATPases by dicyclohexylcarbodiimide (DCCD) results in inactivation of their ATPase activities (9). As shown in Table  II, half of ATP hydrolytic activity of the T. thermophilus ATPase was inactivated by DCCD. Since TF1 lost about 98% of its ATPase activity under the same conditions, the T.
thermophilus ATPase was relatively more resistant to DCCD inactivation than TFi. Another chemical modification reagent, 7-chloro-4-nitrobenzofrazan (NBf-Cl), which inactivates F,-ATPase and V-type ATPase, was a potent inhibitor of the T. thermophilus ATPase. The inhibition by DCCD and NBf-Cl did not reach completion under the conditions described. However, further addition of these reagents to the The activity was assayed in the presence of the indicated concentrations of sodium salts at 55 "C at pH 7.5. reaction nixture reinitiated inactivation. These results show that the sensitivity of the T. thermophilus ATPase to various inhibitors is more similar to that of V-type ATPases than that of F1-ATPases.

N-terminal
Amino Acid Sequence of a and p Subunits of T.
thermophilus ATPase-Forty-three and eighteen amino acid residues from the N termini of the (Y and p subunits, respectively, were determined by Edman degradation and compared with the N-terminal sequences of the catalytic and noncatalytic major subunits of several V-type ATPases and Fl-AT-Pases from bacteria, fungi, plants, and animals (Fig. 4). The 01 subunit of the T. thermophilus ATPase was compared with the catalytic subunits: eukaryotic vacuolar H+-ATPase 67-'IO-kDa subunits, S. acidocaldarius LY subunit, and F1-ATPase /3 subunit. The p subunit of the T. thermophilus ATPase was compared with the noncatalytic subunits; eukaryotic vacuolar H+-ATPase 57-58-kDa subunits, S. acidocaldarius /3 subunit, and F,-ATPase 01 subunit. The N-terminal sequence of the LY subunit of the T. thermophilus ATPase contained 10 amino acid residues that were identical to the catalytic subunits of all of the listed V-type ATPases. When the sequence of the (Y subunit of the T. thermophilus ATPase was compared with each of the V-type ATPases from Arabidopsis thaliana, Neurospora crassa, and S. acidocaldarius, as many as 44, 40, and 35%, respectively, of the amino acid residues were identical. In contrast, when the sequence was compared with the catalytic p subunits of F,-ATPases from Escherichia coli, tobacco chloroplasts, and bovine mitochondria, only 14, 14, and 19%, respectively, of the residues were identical. Similarly, 4 out of 18 amino acid residues in the N-terminal sequence of the p subunit of the T. thermophilus ATPase were identical to the noncatalytic subunits of all of the listed V-type ATPases. Again, the sequence similarity in the region of the N termini between the fi subunit of the T. thermophilus ATPase and the noncatalytic (Y subunits of F1-ATPases was so poor that a meaningful alignment appeared impossible. When a comparison was made between different combinations, between the T. thermophilus LY subunit and the F1-ATPase LY subunits or between the T. thermophilus /3 subunit and the F1-ATPase /3 subunits, no meaningful sequence similarity was found. These results strongly indicate that the T. thermophilus ATPase is not an F1-ATPase but belongs to the V-type ATPase family. DISCUSSION V-type ATPases have been found in a variety of eukaryotic endomembrane vacuolar vesicles, but, in the prokaryotic kingdom, they have so far been detected only in archaebacterial plasma membranes (17)(18)(19)25).
From this unique distribution of the V-type ATPases, it has been postulated that the vacuolar H+-ATPase of eukaryotes arose by the internalization of the plasma membrane H+-ATPase of an archaebacterial-like ancestral cell (15). We solubilized and purified an ATPase from the membranes of T. thermophilus. This ATPase was expected to be an F1-ATPase since the bacterium is a eubacterium. However, the properties of the T. thermophilus ATPase described in this report, such as resistance to azide inhibition, sensitivity to nitrate inhibition, a large (Y subunit by guest on March 24, 2020 http://www.jbc.org/ Downloaded from molecular weight, and a high degree of similarity in the Nterminal amino acid sequences of the two major subunits to those of the V-ATPases, strongly indicate that it belongs to the V-type ATPase family, not to the F,-ATPases. From an analysis of its [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31] and from the fact that its membranes do not contain ether lipids (unique components of archaebacterial membranes (32,33)) it has been well established already that T. thermophilus is a eubacterium, not an archaebacterium. Comparison of the full sequences of the T. thermophilus ATPase subunits with other classes of ATPases and knowledge of the distribution of V-type ATPase in various eubacteria will provide clues to discovering the evolutional reason why a eubacterium, T. thermophilus, has a Vtype ATPase. 17.