- Email: [email protected]
F1-ATPase with cysteine instead of serine at residue 373 of the α subunit
ARCHIVES
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 297, No. 2, September, pp. 334-339,1992
F,-ATPase with Cysteine Instead of Serine at Residue 373 of the O!Subunit Rita S.-F. Lee, Susan Wilke-Mounts, Department
of Biochemistry,
University
and Alan E. Senior’
of Rochester Medical Center, Rochester, New York 14642
Received March 26, 1992, and in revised form May 12, 1992
Escherichia coli strain AN718 contains the aS373F mutation in F,F,-ATP synthase which blocks ATP synthesis (oxidative phosphorylation) and steady-state F1ATPase activity. The revertant strain AN718SS2 containing the mutation cyC373 was isolated and shown to confer a phenotype of higher growth yield than that of the wild type in liquid medium containing limiting glucose, succinate, or LB. Purified F1 from strain AN718SS2 was found to have 30% of wild-type steady-state ATPase activity and 60% of wild-type oxidative phosphorylation activity. Azide sensitivity of ATPase activity and ADPinduced enhancement of bound aurovertin fluorescence, both of which are lost in aS373F mutant F1, were regained in aC373 F1. N-Ethylmaleimide (NEM) inactivated (~C373 F1 steady-state ATPase potently but had no effect on unisite ATPase. Complete inactivation of c&373 F1 steady-state ATPase corresponded to incorporation of one NEM per F1 (mol/mol), in just one of the three (Y subunits. NEM-inactivated enzyme showed azide-insensitive residual ATPase activity and loss of ADP-induced enhancement of bound aurovertin fluorescence. The data confirm the view that placement at residue (~373 of a bulky amino acid side-chain (phenylalanyl or NEM-derivatized cysteinyl) blocks positive catalytic cooperativity in F1. The fact that NEM inhibits steady-state ATPase when only one (Y subunit of three is reacted suggests a cyclical catalytic mechanism. 0 1992 Academic PWS, IW.
F,F,-ATP synthase of Escherichia coli catalyzes ATP synthesis during oxidative phosphorylation and ATPdriven proton pumping. It consists of two sectors: (1) the membrane-embedded FO, made up of three subunits albc10, which constitutes a specific transmembrane proton-conducting pathway, and (2) the membrane-extrinsic F1, made up of five subunits c@~Y&, which carries catalytic sites for ATP synthesis and hydrolysis. The F1 sector can be purified in soluble form for detailed studies. It is i To whom correspondence 334
should be addressed.
currently thought that the p subunits contain the catalytic sites [reviewed in (1, 2)]. Several mutations in the uncA gene (encoding the (Y subunit) lead to loss of function. The most debilitating is &373F (uncA401) which, while having no discernible effects on enzyme structure or oligomeric stability, causes essentially total loss of steady-state ATPase and ATP synthesis activity, although “unisite” ATPase activity is not affected (3). The (rS373 residue is very strongly conserved across species lines, and the mutation to F appears to interfere with positive cooperativity between catalytic sites by interrupting intersubunit a-0 conformational interactions (3-5). Here, we isolated spontaneous revertants of the aS373F mutation, we subjected the revertants to PCR’ and DNA sequencing, and identified a new mutation in the (Y subunit, namely aC373. We describe the properties of the cuC373 F1 and its inactivation by N-ethylmaleimide. METHODS Strains of E. coli, growth of cells, preparation of cell membrane vesicles, purification of F,, and growth yields. The strains of E. coli used are described in Table I. Cells were grown in l- or 13-liter batches (the latter for Fi purification) and passed through a French press, and membrane vesicles were prepared as described (6). Soluble F, was released from membranes and purified as described (7,8). Growth yields in limiting (3 mM) glucose medium were measured as described by Senior et al. (9), and in LB medium as described by Kumamoto and Simoni (10). Growth yields in 15 mM succinate medium were measured at 37°C in a 15-ml volume. Cells were first grown in minimal medium plus 30 mM glucose to OD = 0.5, 1 ml of cell suspension was centrifuged, washed twice, resuspended in 0.5 ml minimal medium, and then inoculated into the succinate medium. Biochemical techniques. Steady-state ATPase activity was assayed in 0.5-1.0 ml assay medium (50 mM Tris-SO,, 10 mM ATP, 4 mM MgS04, pH 8.5) at 30°C. Pi was estimated by the method of Taussky and Shorr (11) or Van Veldhoven and Mannaerts (12). Membrane vesicles (25100 ag protein) or F, (a26 nM) were assayed for l-5 min, and linear proportionality between Pi release and time was seen. ATP-dependent
2 Abbreviations used: PCR, polymerase dodecyl sulfate; dsDNA, double-stranded imide.
chain reaction; SDS, sodium DNA; NEM, N-ethylmale-
0003-9s61/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Escherichia TABLE Strains
of Escherichia
Coli aS373C MUTANT
I coli
Used
Strain
Genotype
Reference
AN1339 AN718 AN718SS2
WC+, argH, pyrE, e&A uncA401 (aS373F). argH, pyrE, e&A uncA (c&373), argH, pyrE, entA
JPl JP2 TGlrA
~IuncA, argH, pyrE, entA AuncA, argH, pyrE, e&A, recA recA derivative of TGl
(25) (5) Rev&ant of AN718 isolated in this work (13) (13)
(13)
pH gradient formation in membrane vesicles was assayed by acridine orange fluorescence quenching (13). NADH-driven ATP synthesis was measured as described by Wise and Senior (14) using 125 pg of F1depleted membranes saturated with Fi (normal or mutant) in a 1.25-ml final volume of assay medium. ADP-induced enhancement of F,-bound aurovertin B fluorescence was as described by Wise et al. (4) in a SPEX Fluorolog 2 spectrofluorometer. Gel electrophoresis in SDS buffer or nondenaturing buffer was as described (15). Protein estimation was as described by Lowry et al. (16) and Bradford (17) for membranes and soluble protein, respectively. Unisite ATP hydrolysis was measured as recently described by Al-Shawi and Senior (8), using 2 pM F1 and 0.2 FM [Y-~~P]ATP. Zsolation ojreuertants. Strain AN718 (uncA401, (xS373F) was grown overnight at 37°C in 20 ml of minimal salts-glucose (30 mM) medium with the required nutritional supplements arginine, uracil, and 2,3-dihydroxybenzoate (18). Two milliliters of suspension was centrifuged; the cells were washed once, then resuspended in 2 ml minimal medium. One-tenth milliliter of lo’, lo-‘, and 10m2dilutions were spread on LB plates. After l-3 days at 37°C a background of smaller colonies (AN718) was seen to be punctuated by larger colonies. The latter were restreaked sequentially on LB-glucose plates to obtain single colonies. Growth tests were then performed as described above on plates containing 30 mM glucose f required supplements, or 30 mM succinate + casamino acids with required supplements, or LB plates; growth yield tests were performed as described above. Strain AN1339 was used as the normal (uric’) control for growth tests. Sequencing of DNA from revertants. Chromosomal DNA was prepared from 2 ml of cell suspension (19) and the whole of the uncA gene was amplified by PCR. The 20-mer primers used were as follows: forward primer, 3’ end - bp -15 upstream of codon 1 of uncA; reverse primer, 3’ end = bp f44 downstream of uncA termination codon. Briefly, conditions used for PCR were 1 pg chromosomal DNA, 0.5 mM primers, 30 cycles each of 1 min at 94”C, 1 min at 55”C, 1 min at 72°C. Amplified dsDNA was digested with SphI and SaZI and cloned into these sites of pUC118. After transformation of strain TGlrA (13), recombinant plasmid DNA was isolated and dsDNA was sequenced by the chain termination method using a series of six sequencing primers covering the entire uncA gene (20). Three clones of each PCR-amplified DNA segment were sequenced to eliminate any polymerase-induced artifacts. Other molecular biology procedures followed standard protocols (21). Mater&. Aurovertin B was from Sigma Chemical Company. [3H]ADP and N-[3H]ethylmaleimide were from New England Nuclear. N-Ethylmaleimide was from Kodak. [y-32P]ATP was from Amersham. RESULTS
Isolation of Revertant Strain AN718SS2 Strain AN718 carries the mutant allele uncA401 encoding the (uS373F mutation (5). This Ser-373 + Phe substitution in the (Y subunit blocks ATP synthesis and
F,-ATPase
335
ATPase activities of F,F,-ATP synthase, so that AN718 cannot grow on succinate and produces small colonies on LB plates. Several revertant colonies were obtained (see Methods) which grew well on succinate plates, showed large colonies on LB plates, and consistently gave significantly higher growth yields in limiting (3 mM) glucose liquid media than the wild-type strain AN1339. They also retained the nutritional supplement requirements of AN718 Two of these revertants were subjected to DNA sequencing (see Methods) and in each case it was found that a new mutation was present at the original site; namely, codon 373 was now TGC whereas it is TCC in the wild type and TTC in AN718 Therefore, cysteine now occupied the position (residue 373) occupied by serine in the wild-type Fi a subunit and by phenylalanine in AN718. One revertant strain was designated AN718SS2 and subjected to further studies. Demonstration That the aS373C Mutation Is Responsible for Enhanced Growth Yields The SphI to Sal1 segment of PCR-amplified DNA from strain AN718SS2 which was cloned into pUC118 for DNA sequencing contained entirely wild-type sequence except for the mutation in codon 373 of the 01 subunit noted above. This segment of DNA encodes residues l-474 of the CYsubunit. Therefore, other mutations might exist elsewhere on the chromosome of strain AN718SS2. An SfiI to AsuII restriction fragment (equivalent to residues 104 to 466 of the a subunit) from the original PCR-amplified DNA was therefore cloned into plasmid pDP34 (20), yielding the new plasmid pRL2. Plasmid pDP34 encodes and expresses all the wild-type F,F,-ATP synthase (uric) structural genes. It was confirmed that the sequence of the DNA in plasmid pRL2 between the SfiI and AsuII sites was wild type except for the single mutation noted above in codon 373 of the a subunit. Plasmid pRL2 was then transformed into strain JP2, which carries a deletion in the chromosomal uncA gene (equivalent to residues l424 of the (Y subunit). The resultant strain pRL2/JP2 showed higher-than-normal growth yields (Table II). The aC373 revertant allele was then moved from plasmid pRL2 into the chromosome of strain JPl (ret’ equivalent of JP2) by recombination (see Table II legend). Growth yields of the resultant recombinants were also higher than those of the wild type (Table II). Amplification of the uncA gene from one JPl recombinant by PCR and subsequent DNA sequencing showed that the aC373 mutation was now present in the chromosome. Therefore, we have established that the aC373 mutation is responsible for the high-growth-yield phenotype. In other work (data not shown) we found that the growth yield in LB medium of strain JM105 (22), which is commonly used in molecular biology studies, was increased by 11% when the aC373 mutation was transduced by Pl phage into the chromosome (using a derivative of JM105 containing a deletion of the uric genes as intermediary).
336
LEE,
WILKE-MOUNTS, TABLE Growth
3 mM glucose
AND
SENIOR
II Yields
Strain
(OD 590 nm)
LB (OD 600 nm)
15 mM succinate” (OD 590 nm)
AN1339 (uric+, oS373) AN718 (oS373F) AN718SS2 (c&373) pRL2/JP2 (oC373) JPl recombinants ((uC373)’
0.88 0.45 1.03 1.04* 1.01
5.60 2.07 6.91 6.03* 6.22
2.2 0 3.1 3.0* 3.1
’ The pH of all cultures at stationary phase was -7.4. * Ampicillin (50 pg/ml) was added. ’ Strain pRL2/JPl was grown in minimal medium with 30 mM succinate and casamino acids but without ampicillin, at 43°C overnight. 0.1 ml was inoculated into fresh medium and the growth repeated twice, after which an aliquot of cell suspension was spread on succinate-casamino acid plates. Sue+, Amps colonies were isolated, and supplemental nutrient requirements checked and shown to be the same as for JPl.
Properties of Membranes from Strain AN718SS2 In strain AN718SS2, membrane ATPase activity was lower than normal and NADH-driven pH gradient formation was somewhat decreased, suggesting the membranes were somewhat proton leaky (Table III). These effects were more pronounced in strain pRL2/JP2, where all the FIFo genes are expressed from the plasmid. ATPase activities and ATP-driven pH gradient formation were in all fully sensitive to 50 PM dicyclohexylcarbodiimide cases. Addition of purified soluble wild-type Fi to the AN718SS2 or pRL2/JP2 membranes restored the NADHdriven acridine orange quenching up to wild-type levels. Therefore, it appeared the Fi had dissociated from the membrane during membrane preparation, leaving unoccupied F. sites in AN718SS2 and pRL2/JP2. It was expected that there would be more unoccupied F. sites in the latter case, consistent with the enhanced proton leakiness. Properties of Purified Soluble FI from Strain AN718SS2 Purified Fi, of normal molecular size as judged by Sephacryl S-300 chromatography, was prepared from TABLE Properties
III
of Membrane
Vesicles pH gradient formation (W quench acridine orange fluorescence)
Strain
Membrane ATPase activity (%)
NADH
ATP
AN1339 (c&373) AN718 (aF373) AN718SS2 (oC373) pRL2/JP2 ((uC373)
100
97 93 88 63”
98 0 44 18”
“Expressed as percentage of membrane ATPase activity of strain pDP34/JP2, the appropriate “wild-type” control for pRL2/JP2. Values for NADH- and ATP-induced pH gradient formation in pDP34/JP2 were similar to AN1339 values.
strain AN718SS2, and it appeared to be partially (-50%) deficient in the 6 subunit as judged by SDS-gel electrophoresis and Coomassie blue staining. The V,,, for ATP hydrolysis was 8.9 units/mg at 30°C, pH 8.5, i.e., 30% of the normal control. The K,ATP was the same as for the wild type (0.25 mM). Stability at 4°C and pH dependence of ATPase activity at pH 7.0-9.5 [measured as in Ref. (13)] were the same as for normal Fi. The relative specificity of AN718SS2 Fi for ATP, GTP, ITP, xanthosine triphosphate, 8-bromo-ATP, and etheno-ATP hydrolysis was similar to that for wild-type Fi. Sodium azide inhibits wild-type steady-state Fi-ATPase activity potently, but does not inhibit the very low, residual ATPase activity in Fi from strain AN718 (3), nor does it inhibit unisite ATP hydrolysis in wild-type Fi (23). The revertant AN718SS2 Fi had largely, although not completely, regained sensitivity to sodium azide; with wildtype Fi, 6 PM azide gave 50% inhibition of steady-state ATPase activity, whereas with AN718SS2 Fi, 31 PM azide gave 50% inhibition. One feature of AN718 Fi noted previously was that it shows no ADP-induced enhancement of bound aurovertin fluorescence (4). This was interpreted as being due to loss of c+l intersubunit conformational interaction related to positive catalytic cooperativity (3). As Fig. 1 shows, AN718SS2 F1 had recovered this property; indeed the fluorescence response was slightly larger than in wild-type F,. The apparent K,ADP for this effect was 1.8 WM for AN718SS2 Fi and 1.9 PM for wild-type F,. Oxidative Phosphorylation
Catalyzed by AN718SS2
FI
ATP synthesis activity was measured in a reconstituted membrane vesicle system, consisting of purified F, from strain AN718SS2 or from wild type, rebound to Fi-depleted membranes from strain AN718. Sufficient Fi was added to obtain maximal NADH-driven acridine orange fluorescence quenching and it was noted that to saturate the F0 sites in the membranes, approximately twice as much AN718SS2 Fi was required as compared to wildtype Fi, per milligram of F,-depleted membranes. Rates
Esckichiu 2.0
I
I
Coli &373C
MUTANT
F,-ATPase
337
AN718SS2 enzyme, which was inhibited by NEM such that only 0.05% of steady-state ATP hydrolysis activity remained, nevertheless showed a unisite ATP hydrolysis rate similar to that of the wild type.
I
NEM Inactivation Is Essentially Complete When Only One of the Three a Subunits per F1 Is Reacted
0.0 0
10
20 ADP
30
40
50
(PM)
FIG. 1. ADP-induced enhancement of bound aurovertin fluorescence in purified Fi. Fi was equilibrated in 0.25 M sucrose, 10 mM Tris-Cl, pH 7.4,0.5 mM EDTA by centrifuge column equilibration, then mixed with aurovertin B (F, = 157 nM, aurovertin B = 10 pM) in a total volume of 0.5 ml at 23°C. Enhancement of fluorescence consequent upon stepwise addition of ADP was measured (X,, = 365 nm, X,, = 475 nm). 0, normal Fi; n , AN718 F,; A, AN718SS2 Fr; A, NEM-treated AN718SS2 F,. In the last case, AN718SS2 F1 (2.6 fiM in buffer consisting of 50 mM TrisHCl, pH 7.0, 0.5 mM EDTA, 10% v/v glycerol) was treated for 30 min at 23°C with 1 mM NEM (final concentration) to inactivate. Then 30 ~1 of inactivated enzyme was added to 0.47 ml of 0.25 M sucrose, 10 mM Tris-Cl, pH 7.4, 0.5 mM EDTA, aurovertin was added to 10 PM, and the fluorescence experiment was performed as above. It was confirmed in control experiments that an equivalent amount of free NEM did not affect the fluorescence of 10 pM free aurovertin.
of NADH-driven ATP synthesis were 26 and 71 nmol/ min/mg membrane protein at 23 and 37”C, respectively, with AN718SS2 Fi, as compared to corresponding rates of 45 and 116 with wild-type Fi. The rate of oxidative phosphorylation with AN718SS2 was therefore 60% that of the wild type at both temperatures. NEM Inactivates Steady-State ATPase but not Unisite ATPase Activity of AN718SS2 F1 Preliminary experiments showed that wild-type Fi showed no loss of steady-state (multisite) ATPase activity when incubated at 2-8 PM Fi with 1 mM N-ethylmaleimide (NEM) in 50 mM Tris-HCl, pH 7.0,0.5 mM EDTA, 10% (v/v) glycerol at 23°C for 24 h, whereas AN718SS2 Fi was rapidly inactivated, with only 2% of ATPase activity remaining after 5 min incubation and ~0.05% remaining after 30 min incubation. The latter residual activity was not inhibited by 1 mM sodium azide. The pH dependence for NEM inactivation of AN718SS2 F1 steady-state ATPase showed a sigmoidal curve, with slow inactivation below pH 5, half-maximal rate of inactivation at pH 6.3, and very rapid inactivation above pH 7.5, consistent with the view that inactivation was due to reaction with the new cysteine at residue (r373.
Inhibition by NEM was studied in more detail, and as seen in Fig. 2A, the AN718SS2 F1 was found to be extremely sensitive, with 2.6 pM F1 being rapidly inactivated by 3.5 or 10 pM NEM. ATP (2 mM) gave partial slowing of inactivation. Labeling studies with [3H]NEM (Fig. 2B) showed a correlation between degree of labeling and degree of inactivation. Extrapolation to 100% inactivation corresponded to labeling to the extent of 1.3 mol [3H]NEM/ mol Fi. However, wild-type Fi also showed a small degree of incorporation of radioactivity under reaction conditions similar to those of Fig. 2B, although there was zero loss of activity. Correction for this gave a value of 1.1 mol [3H]NEM incorporated into AN718SS2 Fi (mol/mol) at 100% inactivation. Fi labeled with [3H]NEM was subjected to SDS-gel electrophoresis and fluorography. Figure 3 demonstrates that in [3H]NEM-inactivated AN718SS2 F,, essentially all of the radioactivity was incorporated into the (Y subunit, whereas wild-type F1 showed no incorporation of radioactivity into the a subunit; rather only a small labeling of the 6 subunit was seen, consistent with previous work (24). From these observations we may conclude that reaction with NEM at the &373 residue in just one of the three copies of the a subunit per F, is sufficient to fully inactivate AN718SS2 Fi steady-state ATPase. NEM-inactivated AN718SS2 F1 was able to bind aurovertin normally as judged by the fluorescence response. However, ADP did not induce any enhancement of bound aurovertin fluorescence in the NEM-inactivated enzyme (Fig. 1). Nondenaturing gel electrophoresis showed that NEM-inactivated AN718SS2 Fr migrated as an Fi oligomer like wild-type Fi, and [3H]ADP binding experiments showed that the loss of ADP-induced enhancement of bound aurovertin fluorescence was not due to loss of ADP binding capability (data not shown). DISCUSSION Mutation of serine at residue 373 of F, (Y subunits to phenylalanine in strain AN718 blocks steady-state ATPase and ATP synthesis activity, abolishes ADP-induced enhancement of bound aurovertin fluorescence, but does not greatly affect the unisite ATP hydrolysis rate (3,4). This, together with other evidence described in the references cited, indicates that an (r-0 intersubunit conformational interaction necessary for positive catalytic cooperativity is blocked by the aS373F mutation. In this work we looked for suppressors of this mutation, and found revertant strains where growth yields were actually restored to levels above that of the wild type. Two of these
338
LEE,
0
50
100
150
Time
200
250
WILKE-MOUNTS.
300
350
(s)
0.6
0.0
0.4
0.6
[‘H]NEM/F,
1.2
AND
SENIOR
is more efficient in oxidative phosphorylation than the wild type. No explanation for this physiological effect is evident from the biochemical data obtained so far. The phenomenon may prove useful in laboratory applications, e.g., overexpression of proteins in E. coli. The (uC373 mutation may be unstable in natural environments due to the fact that it renders F,-ATPase extremely sensitive to SH-reactive reagents. As well as finding the cuC373 pseudorevertants we also encountered several instances of true reversion with growth yields the same as that of the wild type. In all of these cases, DNA sequencing showed that serine was restored at residue ~~373.Possible amino acid substitutions that could occur as a result of single-base changes in codon 373 of the o subunit of strain AN718 would be L, I, V, Y, C, and S. We presume that the first four, being hydrophobic and large, like phenylalanine, cannot give rise to substantial restoration of enzyme function and were not found in the revertant search for this reason. Purified Fi from strain AN718SSZ (aC373) showed, in the main, a large restoration of function, but it differed from wild-type Fi in some characteristics. First, the partial deficiency of 6 subunit in the purified Fi implies that interaction between 6 and the rest of Fi is weaker than normal. This would explain why the AN718SS2 and pRL2/JP2 membrane vesicles appeared to have unoccupied F0 sites (Table II), and why an excess of AN718SS2 F, was required to saturate F0 sites in Fi-depleted membranes in the oxidative phosphorylation experiments. Second, the purified AN718SS2 Fi had only -30% of wild-type F,-ATPase activity and 60% of wild-type oxidative phosphorylation activity. Given that the enzyme is extremely sensitive to NEM inhibition, we considered that our purified enzyme preparation might contain a
1.6
(moi/mol)
FIG. 2. Inactivation and labeling of AN718SS2 F, by N-ethylmaleimide. (A) AN718SS2 (aC373) F1 was equilibrated in 50 mM Tris-HCl, 0.5 mM EDTA, 10% (v/v) glycerol, pH 7.0, by centrifuge column elution, and adjusted to 5.25 pM protein concentration. At time zero an equal volume of buffer containing [3H]NEM was added, and at time intervals 25-~1 aliquots were removed and quenched with 25 ~1 buffer containing 5 mM dithiothreitol. All manipulations were at 23°C. ATPase activity was measured in the quenched samples. Control samples, incubated for 5 min in buffer alone (no [3H]NEM) and then quenched with dithiothreitol, showed no change in ATPase activity. (B) [3H]NEM labeling was measured by adding 20 ~1 of each quenched sample to 80 ~1 of buffer containing bovine serum albumin (1 mg/ml), and passing the solutions through l-ml centrifuge columns equilibrated in buffer also containing 1 mg/ml bovine serum albumin. The entire eluate was then counted. Control experiments showed that the recovery of F,-bound radioactivity from these columns was 97%. l ,3.5 j.tM [aH]NEM (final concentration); A, 10 pM [sHINEM, A, 10 pM NEM plus 2 mM ATP.
revertant strains were subjected to DNA sequencing and both were found to contain a novel residue at residue 373 of the a subunit, namely, cysteine. The results of the growth yield tests imply that the c&373 mutant enzyme
1
234
5676910
FIG. 3. SDS-polyacrylamide gel electrophoresis of [3H]NEM-labeled AN718SS2 F1. F, was labeled with 13H]NEM as described in Fig. 2. Five micrograms of protein from quenched samples was electrophoresed in a 10% SDS-polyacrylamide gel. After staining with Coomassie blue and destaining, the gel was soaked in Amplify (Amersham), dried, and fluorographed at -70°C using preflashed Kodak X-OMAT film. Expt lF1 labeled with 10 pM [3H]NEM: lanes 1 and 2, wild-type F, labeled for 2 and 5 min; lanes 3 and 4, AN718SS2 F, labeled for 2 and 5 min. Expt 2-Fi labeled with 13 FM [aH]NEM: lanes 5, 6, and 7, wild-type F, labeled for 5, 10, and 20 min; lanes 8, 9, and 10, AN718SS2 F, labeled for 5,10, and 20 min. In Expt 1 the degree of labeling of F, was as shown in Fig. 2. In Expt 2 the maximal labeling by [sHINEM (20-min samples, lanes 7 and 10) was 0.65 mol/mol (wild-type F,) and 2.78 mol/mol (AN718SS2 F,).
Escherichia
Coli aS373C MUTANT
population of molecules that had been inactivated during cell breakage and enzyme purification by extraneous SHreactive reagents. Incubation of purified Fi for extended periods with 50 mM dithiothreitol failed to increase ATPase activity; however, this consideration remains a possibility. The restoration of function in AN718SS2 F, was derived from the amino acid substitution aS373F + C and was manifested in recovery of such activities as steadystate ATPase, ATP synthesis, sensitivity of ATPase to sodium azide, and ADP-induced enhancement of bound aurovertin fluorescence. This restoration of function could then be again reversed by chemical modification with NEM, i.e., olS373F + C + C-NEM, where “C-NEM” represents NEM-derivatized cysteine ([N-ethylsuccinimido]-S cysteine), which is not very dissimilar from a phenylalanyl side-chain. Indeed, NEM-derivatized cysteine provides a more bulky side-chain than phenylalanine. It is clear that the properties of the NEM-derivatized ~uC373 F, were similar to those of the original aS373F mutant F,; e.g., there was strong impairment of steadystate (multisite) ATPase with the residual activity being azide insensitive, no effect on unisite ATPase activity, but abolition of ADP-induced enhancement of bound aurovertin fluorescence. The effects of the chemical modification reagent (NEM) and the genetic modification (phenylalanyl) are therefore most likely steric in nature, and not due to generalized disruption of folding of the LY subunit. Overall, these data strongly support our earlier conclusions that placement of a bulky side-chain at residue a373 interferes significantly with intersubunit a-6 conformational interactions and positive catalytic cooperativity in F,-ATPase (3). The major finding of this work, however, is that reaction of residue aC373 with NEM in just one copy of the (Y subunit out of the three copies of (Y subunit per F1 yields complete inactivation of steady-state ATPase activity. This is the first report of “one-third of sites inhibition” in F1 due to covalent reaction of the LYsubunit. Such an effect has previously been seen with reagents that react with the p subunit (e.g., dicyclohexylcarbodiimide and 4chloro-7-nitrobenzofurazan). Our results show that in the aC373 mutant F, reacted with NEM on just one (Ysubunit, unisite catalysis occurred normally, and therefore the intrinsic activity of the high-affinity catalytic site is not impaired; rather it is cooperative catalysis that is very significantly impaired. One possible interpretation of our results is that steady-state ATPase activity involves a cyclical mechanism, and that reaction of just one aC373 residue with NEM can block the catalytic cycle.
339
F,-ATPase
ACKNOWLEDGMENTS This work was supported by Grant GM-25349 to A.E.S. We gratefully acknowledge the helpful comments of Professor R. A. Capaldi, Dr. J. Weber, and Dr. J. Pagan.
REFERENCES 1. Futai, M., Noumi, T., and Maeda, M. (1989) Annu. Reu. Biochem. 58,111-136. 2. Senior, A. E. (1990) Annu. Reu. Biophys. Biophys. Chem. 19, 7-41. 3. Wise, J. G., Latchney, L. R., Ferguson, (1984) Biochemistry 23,1426-1432. 4. Wise, J. G., Latchney, 256.10383-10389.
A. M., and Senior, A. E.
L. R., and Senior, A. E. (1981) J. Biol. Chem.
5. Maggio, M. B., Pagan, J., Parsonage, D., Hatch, A. E. (1987) J. Biol. Chem. 262,8981-8984.
L., and Senior,
6. Senior, A. E., Fayle, D. R. H., Downie, J. A., Gibson, F., and Cox, G. B. (1979) Biochem. J. 180, 111-118. 7. Duncan, T. M., and Senior, A. E. (1985) J. Biol. Chem. 260,49014907. 8. Al-Shawi, 885.
M. K., and Senior, A. E. (1992) Biochemistry
31, 878-
9. Senior, A. E., Latchney, L. R., Ferguson, A. M., and Wise, J. G. (1984) Arch. Biochem. Biophys. 228, 49-53. 10. Kumamoto, C. A., and Simoni, 10037-10042.
R. D. (1986) J. Biol. Chem. 261,
11. Taussky, H. H., and Shorr, E. (1953) J. Biol. Chem. 202,675-685. 12. Van Veldhoven,
P. P., and Mannaerts,
G. P. (1987) Anal. Biochem.
161,45-4&k 13. Rao, R., Pagan, J., and Senior, A. E. (1988) J. Biol. Chem. 263, 15957-15963. 14. Wise, J. G., and Senior, A. E. (1985) Biochemistry 24, 6949-6954. 15. Rao, R., Perlin, D. S., and Senior, A. E. (1987) Arch. Biochem. Biophys. 255, 309-315. 16. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, (1951)J. Biol. Chem. 193,265-275. 17. Bradford,
R. J.
M. M. (1976) Anal. Biochem. 72, 248-254.
18. Gibson, F., Cox, G. B., Downie, J. A., and Radik, J. (1977) Biochem.
J. 162,665-670. 19. Lee, R. S. F., Pagan, J., Wilke-Mounts, Biochemistry 30, 6842-6847. 20. Maggio, M. B., Parsonage, Chem. 263,4619-4623.
S., and Senior, A. E. (1991)
D., and Senior, A. E. (1988) J. Biol.
21. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 22. Yanisch-Perron, 103-119.
C., Vieira,
J., and Messing,
J. (1985) Gene 33,
23. Noumi, T., Maeda, M., and Futai, M. (1987) FEBS Lett. 213,381384. J., and Capaldi, 24. Mendel-Hartvig, Acta 1060,115-124.
R. A. (1991) Biochim.
Biophys.
25. Perlin, D. S., Latchney, L. R., and Senior, A. E. (1985) Biochim. Biophys. Acta 807, 238-244.
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 297, No. 2, September, pp. 334-339,1992
F,-ATPase with Cysteine Instead of Serine at Residue 373 of the O!Subunit Rita S.-F. Lee, Susan Wilke-Mounts, Department
of Biochemistry,
University
and Alan E. Senior’
of Rochester Medical Center, Rochester, New York 14642
Received March 26, 1992, and in revised form May 12, 1992
Escherichia coli strain AN718 contains the aS373F mutation in F,F,-ATP synthase which blocks ATP synthesis (oxidative phosphorylation) and steady-state F1ATPase activity. The revertant strain AN718SS2 containing the mutation cyC373 was isolated and shown to confer a phenotype of higher growth yield than that of the wild type in liquid medium containing limiting glucose, succinate, or LB. Purified F1 from strain AN718SS2 was found to have 30% of wild-type steady-state ATPase activity and 60% of wild-type oxidative phosphorylation activity. Azide sensitivity of ATPase activity and ADPinduced enhancement of bound aurovertin fluorescence, both of which are lost in aS373F mutant F1, were regained in aC373 F1. N-Ethylmaleimide (NEM) inactivated (~C373 F1 steady-state ATPase potently but had no effect on unisite ATPase. Complete inactivation of c&373 F1 steady-state ATPase corresponded to incorporation of one NEM per F1 (mol/mol), in just one of the three (Y subunits. NEM-inactivated enzyme showed azide-insensitive residual ATPase activity and loss of ADP-induced enhancement of bound aurovertin fluorescence. The data confirm the view that placement at residue (~373 of a bulky amino acid side-chain (phenylalanyl or NEM-derivatized cysteinyl) blocks positive catalytic cooperativity in F1. The fact that NEM inhibits steady-state ATPase when only one (Y subunit of three is reacted suggests a cyclical catalytic mechanism. 0 1992 Academic PWS, IW.
F,F,-ATP synthase of Escherichia coli catalyzes ATP synthesis during oxidative phosphorylation and ATPdriven proton pumping. It consists of two sectors: (1) the membrane-embedded FO, made up of three subunits albc10, which constitutes a specific transmembrane proton-conducting pathway, and (2) the membrane-extrinsic F1, made up of five subunits c@~Y&, which carries catalytic sites for ATP synthesis and hydrolysis. The F1 sector can be purified in soluble form for detailed studies. It is i To whom correspondence 334
should be addressed.
currently thought that the p subunits contain the catalytic sites [reviewed in (1, 2)]. Several mutations in the uncA gene (encoding the (Y subunit) lead to loss of function. The most debilitating is &373F (uncA401) which, while having no discernible effects on enzyme structure or oligomeric stability, causes essentially total loss of steady-state ATPase and ATP synthesis activity, although “unisite” ATPase activity is not affected (3). The (rS373 residue is very strongly conserved across species lines, and the mutation to F appears to interfere with positive cooperativity between catalytic sites by interrupting intersubunit a-0 conformational interactions (3-5). Here, we isolated spontaneous revertants of the aS373F mutation, we subjected the revertants to PCR’ and DNA sequencing, and identified a new mutation in the (Y subunit, namely aC373. We describe the properties of the cuC373 F1 and its inactivation by N-ethylmaleimide. METHODS Strains of E. coli, growth of cells, preparation of cell membrane vesicles, purification of F,, and growth yields. The strains of E. coli used are described in Table I. Cells were grown in l- or 13-liter batches (the latter for Fi purification) and passed through a French press, and membrane vesicles were prepared as described (6). Soluble F, was released from membranes and purified as described (7,8). Growth yields in limiting (3 mM) glucose medium were measured as described by Senior et al. (9), and in LB medium as described by Kumamoto and Simoni (10). Growth yields in 15 mM succinate medium were measured at 37°C in a 15-ml volume. Cells were first grown in minimal medium plus 30 mM glucose to OD = 0.5, 1 ml of cell suspension was centrifuged, washed twice, resuspended in 0.5 ml minimal medium, and then inoculated into the succinate medium. Biochemical techniques. Steady-state ATPase activity was assayed in 0.5-1.0 ml assay medium (50 mM Tris-SO,, 10 mM ATP, 4 mM MgS04, pH 8.5) at 30°C. Pi was estimated by the method of Taussky and Shorr (11) or Van Veldhoven and Mannaerts (12). Membrane vesicles (25100 ag protein) or F, (a26 nM) were assayed for l-5 min, and linear proportionality between Pi release and time was seen. ATP-dependent
2 Abbreviations used: PCR, polymerase dodecyl sulfate; dsDNA, double-stranded imide.
chain reaction; SDS, sodium DNA; NEM, N-ethylmale-
0003-9s61/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Escherichia TABLE Strains
of Escherichia
Coli aS373C MUTANT
I coli
Used
Strain
Genotype
Reference
AN1339 AN718 AN718SS2
WC+, argH, pyrE, e&A uncA401 (aS373F). argH, pyrE, e&A uncA (c&373), argH, pyrE, entA
JPl JP2 TGlrA
~IuncA, argH, pyrE, entA AuncA, argH, pyrE, e&A, recA recA derivative of TGl
(25) (5) Rev&ant of AN718 isolated in this work (13) (13)
(13)
pH gradient formation in membrane vesicles was assayed by acridine orange fluorescence quenching (13). NADH-driven ATP synthesis was measured as described by Wise and Senior (14) using 125 pg of F1depleted membranes saturated with Fi (normal or mutant) in a 1.25-ml final volume of assay medium. ADP-induced enhancement of F,-bound aurovertin B fluorescence was as described by Wise et al. (4) in a SPEX Fluorolog 2 spectrofluorometer. Gel electrophoresis in SDS buffer or nondenaturing buffer was as described (15). Protein estimation was as described by Lowry et al. (16) and Bradford (17) for membranes and soluble protein, respectively. Unisite ATP hydrolysis was measured as recently described by Al-Shawi and Senior (8), using 2 pM F1 and 0.2 FM [Y-~~P]ATP. Zsolation ojreuertants. Strain AN718 (uncA401, (xS373F) was grown overnight at 37°C in 20 ml of minimal salts-glucose (30 mM) medium with the required nutritional supplements arginine, uracil, and 2,3-dihydroxybenzoate (18). Two milliliters of suspension was centrifuged; the cells were washed once, then resuspended in 2 ml minimal medium. One-tenth milliliter of lo’, lo-‘, and 10m2dilutions were spread on LB plates. After l-3 days at 37°C a background of smaller colonies (AN718) was seen to be punctuated by larger colonies. The latter were restreaked sequentially on LB-glucose plates to obtain single colonies. Growth tests were then performed as described above on plates containing 30 mM glucose f required supplements, or 30 mM succinate + casamino acids with required supplements, or LB plates; growth yield tests were performed as described above. Strain AN1339 was used as the normal (uric’) control for growth tests. Sequencing of DNA from revertants. Chromosomal DNA was prepared from 2 ml of cell suspension (19) and the whole of the uncA gene was amplified by PCR. The 20-mer primers used were as follows: forward primer, 3’ end - bp -15 upstream of codon 1 of uncA; reverse primer, 3’ end = bp f44 downstream of uncA termination codon. Briefly, conditions used for PCR were 1 pg chromosomal DNA, 0.5 mM primers, 30 cycles each of 1 min at 94”C, 1 min at 55”C, 1 min at 72°C. Amplified dsDNA was digested with SphI and SaZI and cloned into these sites of pUC118. After transformation of strain TGlrA (13), recombinant plasmid DNA was isolated and dsDNA was sequenced by the chain termination method using a series of six sequencing primers covering the entire uncA gene (20). Three clones of each PCR-amplified DNA segment were sequenced to eliminate any polymerase-induced artifacts. Other molecular biology procedures followed standard protocols (21). Mater&. Aurovertin B was from Sigma Chemical Company. [3H]ADP and N-[3H]ethylmaleimide were from New England Nuclear. N-Ethylmaleimide was from Kodak. [y-32P]ATP was from Amersham. RESULTS
Isolation of Revertant Strain AN718SS2 Strain AN718 carries the mutant allele uncA401 encoding the (uS373F mutation (5). This Ser-373 + Phe substitution in the (Y subunit blocks ATP synthesis and
F,-ATPase
335
ATPase activities of F,F,-ATP synthase, so that AN718 cannot grow on succinate and produces small colonies on LB plates. Several revertant colonies were obtained (see Methods) which grew well on succinate plates, showed large colonies on LB plates, and consistently gave significantly higher growth yields in limiting (3 mM) glucose liquid media than the wild-type strain AN1339. They also retained the nutritional supplement requirements of AN718 Two of these revertants were subjected to DNA sequencing (see Methods) and in each case it was found that a new mutation was present at the original site; namely, codon 373 was now TGC whereas it is TCC in the wild type and TTC in AN718 Therefore, cysteine now occupied the position (residue 373) occupied by serine in the wild-type Fi a subunit and by phenylalanine in AN718. One revertant strain was designated AN718SS2 and subjected to further studies. Demonstration That the aS373C Mutation Is Responsible for Enhanced Growth Yields The SphI to Sal1 segment of PCR-amplified DNA from strain AN718SS2 which was cloned into pUC118 for DNA sequencing contained entirely wild-type sequence except for the mutation in codon 373 of the 01 subunit noted above. This segment of DNA encodes residues l-474 of the CYsubunit. Therefore, other mutations might exist elsewhere on the chromosome of strain AN718SS2. An SfiI to AsuII restriction fragment (equivalent to residues 104 to 466 of the a subunit) from the original PCR-amplified DNA was therefore cloned into plasmid pDP34 (20), yielding the new plasmid pRL2. Plasmid pDP34 encodes and expresses all the wild-type F,F,-ATP synthase (uric) structural genes. It was confirmed that the sequence of the DNA in plasmid pRL2 between the SfiI and AsuII sites was wild type except for the single mutation noted above in codon 373 of the a subunit. Plasmid pRL2 was then transformed into strain JP2, which carries a deletion in the chromosomal uncA gene (equivalent to residues l424 of the (Y subunit). The resultant strain pRL2/JP2 showed higher-than-normal growth yields (Table II). The aC373 revertant allele was then moved from plasmid pRL2 into the chromosome of strain JPl (ret’ equivalent of JP2) by recombination (see Table II legend). Growth yields of the resultant recombinants were also higher than those of the wild type (Table II). Amplification of the uncA gene from one JPl recombinant by PCR and subsequent DNA sequencing showed that the aC373 mutation was now present in the chromosome. Therefore, we have established that the aC373 mutation is responsible for the high-growth-yield phenotype. In other work (data not shown) we found that the growth yield in LB medium of strain JM105 (22), which is commonly used in molecular biology studies, was increased by 11% when the aC373 mutation was transduced by Pl phage into the chromosome (using a derivative of JM105 containing a deletion of the uric genes as intermediary).
336
LEE,
WILKE-MOUNTS, TABLE Growth
3 mM glucose
AND
SENIOR
II Yields
Strain
(OD 590 nm)
LB (OD 600 nm)
15 mM succinate” (OD 590 nm)
AN1339 (uric+, oS373) AN718 (oS373F) AN718SS2 (c&373) pRL2/JP2 (oC373) JPl recombinants ((uC373)’
0.88 0.45 1.03 1.04* 1.01
5.60 2.07 6.91 6.03* 6.22
2.2 0 3.1 3.0* 3.1
’ The pH of all cultures at stationary phase was -7.4. * Ampicillin (50 pg/ml) was added. ’ Strain pRL2/JPl was grown in minimal medium with 30 mM succinate and casamino acids but without ampicillin, at 43°C overnight. 0.1 ml was inoculated into fresh medium and the growth repeated twice, after which an aliquot of cell suspension was spread on succinate-casamino acid plates. Sue+, Amps colonies were isolated, and supplemental nutrient requirements checked and shown to be the same as for JPl.
Properties of Membranes from Strain AN718SS2 In strain AN718SS2, membrane ATPase activity was lower than normal and NADH-driven pH gradient formation was somewhat decreased, suggesting the membranes were somewhat proton leaky (Table III). These effects were more pronounced in strain pRL2/JP2, where all the FIFo genes are expressed from the plasmid. ATPase activities and ATP-driven pH gradient formation were in all fully sensitive to 50 PM dicyclohexylcarbodiimide cases. Addition of purified soluble wild-type Fi to the AN718SS2 or pRL2/JP2 membranes restored the NADHdriven acridine orange quenching up to wild-type levels. Therefore, it appeared the Fi had dissociated from the membrane during membrane preparation, leaving unoccupied F. sites in AN718SS2 and pRL2/JP2. It was expected that there would be more unoccupied F. sites in the latter case, consistent with the enhanced proton leakiness. Properties of Purified Soluble FI from Strain AN718SS2 Purified Fi, of normal molecular size as judged by Sephacryl S-300 chromatography, was prepared from TABLE Properties
III
of Membrane
Vesicles pH gradient formation (W quench acridine orange fluorescence)
Strain
Membrane ATPase activity (%)
NADH
ATP
AN1339 (c&373) AN718 (aF373) AN718SS2 (oC373) pRL2/JP2 ((uC373)
100
97 93 88 63”
98 0 44 18”
“Expressed as percentage of membrane ATPase activity of strain pDP34/JP2, the appropriate “wild-type” control for pRL2/JP2. Values for NADH- and ATP-induced pH gradient formation in pDP34/JP2 were similar to AN1339 values.
strain AN718SS2, and it appeared to be partially (-50%) deficient in the 6 subunit as judged by SDS-gel electrophoresis and Coomassie blue staining. The V,,, for ATP hydrolysis was 8.9 units/mg at 30°C, pH 8.5, i.e., 30% of the normal control. The K,ATP was the same as for the wild type (0.25 mM). Stability at 4°C and pH dependence of ATPase activity at pH 7.0-9.5 [measured as in Ref. (13)] were the same as for normal Fi. The relative specificity of AN718SS2 Fi for ATP, GTP, ITP, xanthosine triphosphate, 8-bromo-ATP, and etheno-ATP hydrolysis was similar to that for wild-type Fi. Sodium azide inhibits wild-type steady-state Fi-ATPase activity potently, but does not inhibit the very low, residual ATPase activity in Fi from strain AN718 (3), nor does it inhibit unisite ATP hydrolysis in wild-type Fi (23). The revertant AN718SS2 Fi had largely, although not completely, regained sensitivity to sodium azide; with wildtype Fi, 6 PM azide gave 50% inhibition of steady-state ATPase activity, whereas with AN718SS2 Fi, 31 PM azide gave 50% inhibition. One feature of AN718 Fi noted previously was that it shows no ADP-induced enhancement of bound aurovertin fluorescence (4). This was interpreted as being due to loss of c+l intersubunit conformational interaction related to positive catalytic cooperativity (3). As Fig. 1 shows, AN718SS2 F1 had recovered this property; indeed the fluorescence response was slightly larger than in wild-type F,. The apparent K,ADP for this effect was 1.8 WM for AN718SS2 Fi and 1.9 PM for wild-type F,. Oxidative Phosphorylation
Catalyzed by AN718SS2
FI
ATP synthesis activity was measured in a reconstituted membrane vesicle system, consisting of purified F, from strain AN718SS2 or from wild type, rebound to Fi-depleted membranes from strain AN718. Sufficient Fi was added to obtain maximal NADH-driven acridine orange fluorescence quenching and it was noted that to saturate the F0 sites in the membranes, approximately twice as much AN718SS2 Fi was required as compared to wildtype Fi, per milligram of F,-depleted membranes. Rates
Esckichiu 2.0
I
I
Coli &373C
MUTANT
F,-ATPase
337
AN718SS2 enzyme, which was inhibited by NEM such that only 0.05% of steady-state ATP hydrolysis activity remained, nevertheless showed a unisite ATP hydrolysis rate similar to that of the wild type.
I
NEM Inactivation Is Essentially Complete When Only One of the Three a Subunits per F1 Is Reacted
0.0 0
10
20 ADP
30
40
50
(PM)
FIG. 1. ADP-induced enhancement of bound aurovertin fluorescence in purified Fi. Fi was equilibrated in 0.25 M sucrose, 10 mM Tris-Cl, pH 7.4,0.5 mM EDTA by centrifuge column equilibration, then mixed with aurovertin B (F, = 157 nM, aurovertin B = 10 pM) in a total volume of 0.5 ml at 23°C. Enhancement of fluorescence consequent upon stepwise addition of ADP was measured (X,, = 365 nm, X,, = 475 nm). 0, normal Fi; n , AN718 F,; A, AN718SS2 Fr; A, NEM-treated AN718SS2 F,. In the last case, AN718SS2 F1 (2.6 fiM in buffer consisting of 50 mM TrisHCl, pH 7.0, 0.5 mM EDTA, 10% v/v glycerol) was treated for 30 min at 23°C with 1 mM NEM (final concentration) to inactivate. Then 30 ~1 of inactivated enzyme was added to 0.47 ml of 0.25 M sucrose, 10 mM Tris-Cl, pH 7.4, 0.5 mM EDTA, aurovertin was added to 10 PM, and the fluorescence experiment was performed as above. It was confirmed in control experiments that an equivalent amount of free NEM did not affect the fluorescence of 10 pM free aurovertin.
of NADH-driven ATP synthesis were 26 and 71 nmol/ min/mg membrane protein at 23 and 37”C, respectively, with AN718SS2 Fi, as compared to corresponding rates of 45 and 116 with wild-type Fi. The rate of oxidative phosphorylation with AN718SS2 was therefore 60% that of the wild type at both temperatures. NEM Inactivates Steady-State ATPase but not Unisite ATPase Activity of AN718SS2 F1 Preliminary experiments showed that wild-type Fi showed no loss of steady-state (multisite) ATPase activity when incubated at 2-8 PM Fi with 1 mM N-ethylmaleimide (NEM) in 50 mM Tris-HCl, pH 7.0,0.5 mM EDTA, 10% (v/v) glycerol at 23°C for 24 h, whereas AN718SS2 Fi was rapidly inactivated, with only 2% of ATPase activity remaining after 5 min incubation and ~0.05% remaining after 30 min incubation. The latter residual activity was not inhibited by 1 mM sodium azide. The pH dependence for NEM inactivation of AN718SS2 F1 steady-state ATPase showed a sigmoidal curve, with slow inactivation below pH 5, half-maximal rate of inactivation at pH 6.3, and very rapid inactivation above pH 7.5, consistent with the view that inactivation was due to reaction with the new cysteine at residue (r373.
Inhibition by NEM was studied in more detail, and as seen in Fig. 2A, the AN718SS2 F1 was found to be extremely sensitive, with 2.6 pM F1 being rapidly inactivated by 3.5 or 10 pM NEM. ATP (2 mM) gave partial slowing of inactivation. Labeling studies with [3H]NEM (Fig. 2B) showed a correlation between degree of labeling and degree of inactivation. Extrapolation to 100% inactivation corresponded to labeling to the extent of 1.3 mol [3H]NEM/ mol Fi. However, wild-type Fi also showed a small degree of incorporation of radioactivity under reaction conditions similar to those of Fig. 2B, although there was zero loss of activity. Correction for this gave a value of 1.1 mol [3H]NEM incorporated into AN718SS2 Fi (mol/mol) at 100% inactivation. Fi labeled with [3H]NEM was subjected to SDS-gel electrophoresis and fluorography. Figure 3 demonstrates that in [3H]NEM-inactivated AN718SS2 F,, essentially all of the radioactivity was incorporated into the (Y subunit, whereas wild-type F1 showed no incorporation of radioactivity into the a subunit; rather only a small labeling of the 6 subunit was seen, consistent with previous work (24). From these observations we may conclude that reaction with NEM at the &373 residue in just one of the three copies of the a subunit per F, is sufficient to fully inactivate AN718SS2 Fi steady-state ATPase. NEM-inactivated AN718SS2 F1 was able to bind aurovertin normally as judged by the fluorescence response. However, ADP did not induce any enhancement of bound aurovertin fluorescence in the NEM-inactivated enzyme (Fig. 1). Nondenaturing gel electrophoresis showed that NEM-inactivated AN718SS2 Fr migrated as an Fi oligomer like wild-type Fi, and [3H]ADP binding experiments showed that the loss of ADP-induced enhancement of bound aurovertin fluorescence was not due to loss of ADP binding capability (data not shown). DISCUSSION Mutation of serine at residue 373 of F, (Y subunits to phenylalanine in strain AN718 blocks steady-state ATPase and ATP synthesis activity, abolishes ADP-induced enhancement of bound aurovertin fluorescence, but does not greatly affect the unisite ATP hydrolysis rate (3,4). This, together with other evidence described in the references cited, indicates that an (r-0 intersubunit conformational interaction necessary for positive catalytic cooperativity is blocked by the aS373F mutation. In this work we looked for suppressors of this mutation, and found revertant strains where growth yields were actually restored to levels above that of the wild type. Two of these
338
LEE,
0
50
100
150
Time
200
250
WILKE-MOUNTS.
300
350
(s)
0.6
0.0
0.4
0.6
[‘H]NEM/F,
1.2
AND
SENIOR
is more efficient in oxidative phosphorylation than the wild type. No explanation for this physiological effect is evident from the biochemical data obtained so far. The phenomenon may prove useful in laboratory applications, e.g., overexpression of proteins in E. coli. The (uC373 mutation may be unstable in natural environments due to the fact that it renders F,-ATPase extremely sensitive to SH-reactive reagents. As well as finding the cuC373 pseudorevertants we also encountered several instances of true reversion with growth yields the same as that of the wild type. In all of these cases, DNA sequencing showed that serine was restored at residue ~~373.Possible amino acid substitutions that could occur as a result of single-base changes in codon 373 of the o subunit of strain AN718 would be L, I, V, Y, C, and S. We presume that the first four, being hydrophobic and large, like phenylalanine, cannot give rise to substantial restoration of enzyme function and were not found in the revertant search for this reason. Purified Fi from strain AN718SSZ (aC373) showed, in the main, a large restoration of function, but it differed from wild-type Fi in some characteristics. First, the partial deficiency of 6 subunit in the purified Fi implies that interaction between 6 and the rest of Fi is weaker than normal. This would explain why the AN718SS2 and pRL2/JP2 membrane vesicles appeared to have unoccupied F0 sites (Table II), and why an excess of AN718SS2 F, was required to saturate F0 sites in Fi-depleted membranes in the oxidative phosphorylation experiments. Second, the purified AN718SS2 Fi had only -30% of wild-type F,-ATPase activity and 60% of wild-type oxidative phosphorylation activity. Given that the enzyme is extremely sensitive to NEM inhibition, we considered that our purified enzyme preparation might contain a
1.6
(moi/mol)
FIG. 2. Inactivation and labeling of AN718SS2 F, by N-ethylmaleimide. (A) AN718SS2 (aC373) F1 was equilibrated in 50 mM Tris-HCl, 0.5 mM EDTA, 10% (v/v) glycerol, pH 7.0, by centrifuge column elution, and adjusted to 5.25 pM protein concentration. At time zero an equal volume of buffer containing [3H]NEM was added, and at time intervals 25-~1 aliquots were removed and quenched with 25 ~1 buffer containing 5 mM dithiothreitol. All manipulations were at 23°C. ATPase activity was measured in the quenched samples. Control samples, incubated for 5 min in buffer alone (no [3H]NEM) and then quenched with dithiothreitol, showed no change in ATPase activity. (B) [3H]NEM labeling was measured by adding 20 ~1 of each quenched sample to 80 ~1 of buffer containing bovine serum albumin (1 mg/ml), and passing the solutions through l-ml centrifuge columns equilibrated in buffer also containing 1 mg/ml bovine serum albumin. The entire eluate was then counted. Control experiments showed that the recovery of F,-bound radioactivity from these columns was 97%. l ,3.5 j.tM [aH]NEM (final concentration); A, 10 pM [sHINEM, A, 10 pM NEM plus 2 mM ATP.
revertant strains were subjected to DNA sequencing and both were found to contain a novel residue at residue 373 of the a subunit, namely, cysteine. The results of the growth yield tests imply that the c&373 mutant enzyme
1
234
5676910
FIG. 3. SDS-polyacrylamide gel electrophoresis of [3H]NEM-labeled AN718SS2 F1. F, was labeled with 13H]NEM as described in Fig. 2. Five micrograms of protein from quenched samples was electrophoresed in a 10% SDS-polyacrylamide gel. After staining with Coomassie blue and destaining, the gel was soaked in Amplify (Amersham), dried, and fluorographed at -70°C using preflashed Kodak X-OMAT film. Expt lF1 labeled with 10 pM [3H]NEM: lanes 1 and 2, wild-type F, labeled for 2 and 5 min; lanes 3 and 4, AN718SS2 F, labeled for 2 and 5 min. Expt 2-Fi labeled with 13 FM [aH]NEM: lanes 5, 6, and 7, wild-type F, labeled for 5, 10, and 20 min; lanes 8, 9, and 10, AN718SS2 F, labeled for 5,10, and 20 min. In Expt 1 the degree of labeling of F, was as shown in Fig. 2. In Expt 2 the maximal labeling by [sHINEM (20-min samples, lanes 7 and 10) was 0.65 mol/mol (wild-type F,) and 2.78 mol/mol (AN718SS2 F,).
Escherichia
Coli aS373C MUTANT
population of molecules that had been inactivated during cell breakage and enzyme purification by extraneous SHreactive reagents. Incubation of purified Fi for extended periods with 50 mM dithiothreitol failed to increase ATPase activity; however, this consideration remains a possibility. The restoration of function in AN718SS2 F, was derived from the amino acid substitution aS373F + C and was manifested in recovery of such activities as steadystate ATPase, ATP synthesis, sensitivity of ATPase to sodium azide, and ADP-induced enhancement of bound aurovertin fluorescence. This restoration of function could then be again reversed by chemical modification with NEM, i.e., olS373F + C + C-NEM, where “C-NEM” represents NEM-derivatized cysteine ([N-ethylsuccinimido]-S cysteine), which is not very dissimilar from a phenylalanyl side-chain. Indeed, NEM-derivatized cysteine provides a more bulky side-chain than phenylalanine. It is clear that the properties of the NEM-derivatized ~uC373 F, were similar to those of the original aS373F mutant F,; e.g., there was strong impairment of steadystate (multisite) ATPase with the residual activity being azide insensitive, no effect on unisite ATPase activity, but abolition of ADP-induced enhancement of bound aurovertin fluorescence. The effects of the chemical modification reagent (NEM) and the genetic modification (phenylalanyl) are therefore most likely steric in nature, and not due to generalized disruption of folding of the LY subunit. Overall, these data strongly support our earlier conclusions that placement of a bulky side-chain at residue a373 interferes significantly with intersubunit a-6 conformational interactions and positive catalytic cooperativity in F,-ATPase (3). The major finding of this work, however, is that reaction of residue aC373 with NEM in just one copy of the (Y subunit out of the three copies of (Y subunit per F1 yields complete inactivation of steady-state ATPase activity. This is the first report of “one-third of sites inhibition” in F1 due to covalent reaction of the LYsubunit. Such an effect has previously been seen with reagents that react with the p subunit (e.g., dicyclohexylcarbodiimide and 4chloro-7-nitrobenzofurazan). Our results show that in the aC373 mutant F, reacted with NEM on just one (Ysubunit, unisite catalysis occurred normally, and therefore the intrinsic activity of the high-affinity catalytic site is not impaired; rather it is cooperative catalysis that is very significantly impaired. One possible interpretation of our results is that steady-state ATPase activity involves a cyclical mechanism, and that reaction of just one aC373 residue with NEM can block the catalytic cycle.
339
F,-ATPase
ACKNOWLEDGMENTS This work was supported by Grant GM-25349 to A.E.S. We gratefully acknowledge the helpful comments of Professor R. A. Capaldi, Dr. J. Weber, and Dr. J. Pagan.
REFERENCES 1. Futai, M., Noumi, T., and Maeda, M. (1989) Annu. Reu. Biochem. 58,111-136. 2. Senior, A. E. (1990) Annu. Reu. Biophys. Biophys. Chem. 19, 7-41. 3. Wise, J. G., Latchney, L. R., Ferguson, (1984) Biochemistry 23,1426-1432. 4. Wise, J. G., Latchney, 256.10383-10389.
A. M., and Senior, A. E.
L. R., and Senior, A. E. (1981) J. Biol. Chem.
5. Maggio, M. B., Pagan, J., Parsonage, D., Hatch, A. E. (1987) J. Biol. Chem. 262,8981-8984.
L., and Senior,
6. Senior, A. E., Fayle, D. R. H., Downie, J. A., Gibson, F., and Cox, G. B. (1979) Biochem. J. 180, 111-118. 7. Duncan, T. M., and Senior, A. E. (1985) J. Biol. Chem. 260,49014907. 8. Al-Shawi, 885.
M. K., and Senior, A. E. (1992) Biochemistry
31, 878-
9. Senior, A. E., Latchney, L. R., Ferguson, A. M., and Wise, J. G. (1984) Arch. Biochem. Biophys. 228, 49-53. 10. Kumamoto, C. A., and Simoni, 10037-10042.
R. D. (1986) J. Biol. Chem. 261,
11. Taussky, H. H., and Shorr, E. (1953) J. Biol. Chem. 202,675-685. 12. Van Veldhoven,
P. P., and Mannaerts,
G. P. (1987) Anal. Biochem.
161,45-4&k 13. Rao, R., Pagan, J., and Senior, A. E. (1988) J. Biol. Chem. 263, 15957-15963. 14. Wise, J. G., and Senior, A. E. (1985) Biochemistry 24, 6949-6954. 15. Rao, R., Perlin, D. S., and Senior, A. E. (1987) Arch. Biochem. Biophys. 255, 309-315. 16. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, (1951)J. Biol. Chem. 193,265-275. 17. Bradford,
R. J.
M. M. (1976) Anal. Biochem. 72, 248-254.
18. Gibson, F., Cox, G. B., Downie, J. A., and Radik, J. (1977) Biochem.
J. 162,665-670. 19. Lee, R. S. F., Pagan, J., Wilke-Mounts, Biochemistry 30, 6842-6847. 20. Maggio, M. B., Parsonage, Chem. 263,4619-4623.
S., and Senior, A. E. (1991)
D., and Senior, A. E. (1988) J. Biol.
21. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 22. Yanisch-Perron, 103-119.
C., Vieira,
J., and Messing,
J. (1985) Gene 33,
23. Noumi, T., Maeda, M., and Futai, M. (1987) FEBS Lett. 213,381384. J., and Capaldi, 24. Mendel-Hartvig, Acta 1060,115-124.
R. A. (1991) Biochim.
Biophys.
25. Perlin, D. S., Latchney, L. R., and Senior, A. E. (1985) Biochim. Biophys. Acta 807, 238-244.