THE JOURNAL OF BIOLOGICAL CHEMISTRY Vul. 256, No 21, Iesue of November 10, pp. 11112-11116, 1981 Printedrn U.S.A.

Inhibition of Coupling Factor B Activity by Cadmium Ion, Arsenite-2,3- dimercaptopropanol, and Phenylarsine Oxide, and Preferential Reactivation by Dithiols"

(Received for publication, March 5, 1981)

Saroj Joshi and James B. Hughes From the Boston Biomedical Research Institute, Boston, Massachusetts 02114

Coupling factor B activity was measured by the stim- ulation of the ATP-driven NAD' reduction by succinate or the 32Pi-ATP exchange activity of Factor B-depleted submitochondrial particles. Half-maximal coupling ac- tivity was inhibited by 30 p~ cadmium, 5 PM phenylar- sine oxide, or 0.3 n m arsenite-2,3-dimercaptopropanol. The inhibition was relieved by slight excess of dithiol but not by a 10-fold molar excess of 2-mercaptoethanol. Inhibition of coupling activity by phenylarsine oxide or cadmium was not due to interference in binding of Factor B to depleted particles. Isolated Factor B binds phenylarsine oxide resulting in loss of ability to stim- ulate depleted submitochondrial particles. The inhibi- tion was largely overcome by dithiol but not by mono- thiols. The residual coupling activity of depleted sub- mitochondrial particles was highly resistant to cad- mium or arsenical. Moreover, binding of arsenical to the depleted particlesper se, did not result in inhibition of Factor B-stimulated activity. Furthermore, the ad- dition of phenylarsine oxide to H+-ATPase resulted in loss of Pi-ATP exchange and stimulation of oligomycin- sensitive ATPase activities. Both effects were further potentiated by 2-mercaptoethanol and reversed by di- thiols. These effects parallel uncoupling of oxidative phosphorylation in mitochondria by these inhibitors and point to Factor B as the probable component sen- sitive to these inhibitors.

Earlier work from this laboratory (1-6) had provided evi- dence for the involvement of a dithiol component in the coupling uf mitochondrial electron transport to oxidative phosphorylation. I t was shown that 5 p~ cadmium, 50 p~ arsenite-BAL,' and 33 p~ y-(p-arsenopheny1)-n-butyrate (5) uncouple ATP synthesis, stimulate latent ATPase, and inhibit Pi-ATP exchange activity in rat liver mitochondria. These levels had no significant effect on oxidation reactions. The increased ATPase activity as a result of the arsenical addition was still inhibited by oligomycin (5). Cd2' and BAL-arsenite also relieved the inhibition of State 3 respiration caused by oligomycin in tightly coupled mitochondria. Similar uncou-

* The work was supported in part by National Institutes of Health Grants GM 13641, GM-26420, and ES 02167. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' The abbreviations used are: BAL, 2,3-dimercaptopropanol; AE-P, ammonia-EDTA-extracted submitochondrial particles; BSA, bovine serum albumin; Cd2, cadmium; CM, carboxymethylcellulose; EDTA, ethylenediamine tetraacetate; ETPH, electron transport particles from heavy layer of mitochondria; FB, coupling factor B; KCN, potassium cyanide; PMS, phenazine methyl sulfate; PhAsO, phenyl- arsine oxide.

pling effects were observed in heart mitochondria and mito- chondrial fragments, although 5 times higher levels of Cd2' or BAL-arsenite were required ( 6 ) . When used separately, arsen- ite or BAL was ineffective as uncoupler. It was proposed that BAL was required for transporting arsenite to the dithiol site involved in the phosphorylation sequence. The requirement for BAL for potentiation of arsenite effects has since been observed with three other enzyme systems, namely, P-hydrox- ybutyric dehydrogenase (7), myosin ATPase (8), and acetyl coenzyme A carboxylase (9).

The uncoupling by Cd2' or BAL-arsenite was readily re- versed by stoichiometric amounts of dithiols but not by mon- othiols. It was postulated that the dithiol reagents act at a site located between the electron transport event and the oligomycin-sensitive step (5). Attempts to identify a specific Cd2+-binding component were unsuccessful, since the binding of Cd2' by mitochondria increased linearly well beyond the amount necessary for complete uncoupling (1).

As part of the program to elucidate the mechanism of uncoupling by Cd", we undertook the fractionation of the components of the oxidative phosphorylation system. This led to the isolation of a low molecular weight protein, namely coupling factor B from beef heart mitochondria (10). FB has no intrinsic activity but stimulates ATP-driven NAD+ reduc- tion by succinate, ATP-dependent NADP+ reduction by NADH, Pi-ATP exchange, and net phosphorylation in am- monia EDTA-extracted submitochondrial particles (10). Re- cently, FB has also been shown to be an essential component of the Fo segment of the Hf-ATPase2 (11) where it is needed for Pi-ATP exchange activity, but not for F1 binding, or oligomycin sensitivity. Amino acid analysis of FB indicates 1 mol each of cysteine and cystine/14,600-dalton monomer (13). The functional role of Factor B in the coupling mechanism, together with the presence of a thiol and a dithiol in the molecule, strongly implicate Fg as a possible Cd"+/arsenite- BAL-sensitive site.

In the present communication, evidence is presented to demonstrate that purified FB binds Cd" or phenylarsine oxide, and that the binding of these reagents inhibits FB- stimulated energy-linked activities. The inhibition is relieved by stoichiometric levels of dithiols but not by similar levels of monothiols.

MATERIALS AND METHODS

Beef heart mitochondria (14), ETPH (15), H+-ATPase (ll), and FB (11, 14) were isolated by published procedures. ETPH partially de-

' We have observed by direct pH measurements that the ATPase or Fo-FI complex made by the lysolecithin extraction procedure (11) is capable of utilizing ATP energy for proton pumping in the same manner as the H+-translocating ATPase preparation of Serrano et al. (12).

11112

Cadmium Inhibition of Mitochondrial Coupling Factor B 11113

pleted of FB (AE-P), were prepared as follows: ETPH were suspended at 2 mg ml" in 2 mM Tris-C1 (pH 7.2) containing 0.25 M sucrose and 5.0 m EDTA and the pH adjusted to 8.8 with 1 N NH40H. The suspension was centrifuged at 100,OOO X g for 30 min and the washing step repeated a second time. Factor B used in these experiments was in general purified to the CM-cellulose step (11, 14). Similar results were obtained using Fg purified through the Sephadex G-75 filtration step. The phenylarsine oxide-binding experiment described in Fig. 3 was carried out using purified FB, which was homogeneous by poly- acrylamide gel electrophoresis in sodium dodecyl sulfate (11, 14). One unit of Factor B produces a stimulation of AE-P activity in ATP- driven NAD' reduction by succinate to the extent of 1 pmol rnin". ATP-driven NAD' reduction by succinate was measured as described by Lam et al. (10). Pi-ATP exchange and ATPase activities were measured as described by Joshi et al. (11).

RESULTS

The general procedure for determining the effect of Cd2+, PhAsO, or BAL-arsenite on either Pi-ATP exchange or NAD' reduction by succinate involved incubation of protein (ETPH, AE-P, Fe, or H+-ATPase) in a volume of 50 p1 at 0 "C for 5 min with the reagent being tested. Each aliquot was then diluted with the appropriate assay medium containing no dithiothreitol and the assay completed as described. The concentrations of Cd2+ and PhAsO mentioned in the tables and figures refer to the final value in the total assay volume. Similar results were obtained when the inhibitors were added last to the reaction mixture at their final concentration, show- ing that preincubation in a small volume at higher inhibitor concentration is not necessary for these effects. The inhibition of FB-stimulated NAD+ reduction by succinate or Pi-ATP exchange in AE-P was interpreted as indicative of an effect on the coupling factor. Similar effects were observed with all three inhibitors, and the data in the text are representative of the general pattern.

Effect of PhAsO on ATP-driven NAD' Reduction by Suc- cinate in ETPH-As seen in Fig. 1, as little as 5 p~ PhAsO inhibited ATP-driven NAD' reduction by succinate in ETPH. The K, for the inhibition was in the range of 20 p ~ . At PhAsO concentrations up to 200 ELM, neither succinate-PMS reductase activity nor succinate oxidase activity was inhibited. The inhibition by the arsenical was relieved by a 2- or 3-fold excess of dithiothreitol or BAL (data not shown) but not by up to a 10-fold excess of 2-mercaptoethanol, cysteine, or glutathione (Table I). The effectiveness of a small excess of dithiothreitol

2"E, pM - 200

40 80 120 160 I 1 I

" L E, 20 40 60 80

PhAsO, pM FIG. 1. Effect of phenylarsine oxide on the ATP-driven

NAD' reduction by succinate in ETPH. Aliquots containing 0.5 mg of ETPH and indicated concentrations of phenylarsine oxide and/ or 2-mercaptoethanol (2-ME) were preincubated in 50 pl volume on ice for 5 min. The ATP-driven NAD' reduction by succinate was next measured by incubating samples at 38 "C for 2 min in a total volume of 2.9 ml containing 150 pmol of Tris-SO4 (pH 7.8), 6 -01 of ATP, 10 pmol of MgCL, 20 pmol of succinate, 2 mg of bovine serum albumin, and 1.5 pmol of NAD'. This was followed by the addition of 3 pmol of potassium cyanide in 0.1 ml. NAD' reduction was monitored by measuring absorbance at 340 nm.

TABLE I Reversibility ofphenylarsine oxide inhibition by mono- or dithiols

Aliquots containing 0.5 mg of ETPH in 50 pl volume were incubated at 0 "C for 5 min in the presence of phenylarsine oxide, and mono- or &thiols as indicated. (The above concentration of thiols had no effect on the activity of ETPH in the absence of phenylarsine oxide.) ATP- dependent NAD' reduction by succinate was assayed as described in Fig. 1.

Additions Amount NAD' reduced

nmol m?" mg-

- 160 + 50 p~ PhAsO 44 + 50 ~ L M PhAsO + 100 p~ 2-mercaptoethanol 5 + 50 p~ PhAsO + 500 p~ 2-mercaptoethanol 7 + 50 @.% PhAsO + 100 p~ cysteine 46 + 50 p~ PhAsO + 500 p~ cysteine 49 + 50 p~ PhAsO + 100 p~ glutathione 45 + 50 PM PhAsO + 500 p~ glutathione 46 + 50 p~ PhAsO + 100 p~ BAL 142 + 50 p~ PhAsO + 500 p~ BAL 149

and the ineffectiveness of large excesses of monothiol reagents in reversing PhAsO inhibition were both independent of the PhAsO concentration.

2-Mercaptoethanol by itself does not affect the NAD+ re- duction by succinate, but strongly potentiates the inhibitory effect of PhAsO (Fig. 1). It is interesting to note that more polar, hydrophilic thiols such as cysteine or glutathione are inert in this respect, as has been reported for the oxidative phosphorylation system ( 5 ) .

Inhibition of Pi-ATP exchange in the AE-P versus AE-P Reconstituted with Fe-In our efforts to localize the dithiol inhibitor binding site, AE-P were tested prior to, and subse- quent to reconstitution with FB. The reconstituted P,-ATP exchange activity is strongly inhibited by PhAsO as is indi- cated by the decreased level of stimulation following reconsti- tution (Fig. 2 A ) . The concentration for half-maximal inhibi- tion was 5 PM. The magnitude of the inhibition was independ- ent of whether PhAsO was added to AE-P, before (Fig. 2 A ) or after (Fig. 2B) reconstitution with Fg. The residual exchange activity in AE-P was resistant to PhAsO inhibition at concen- trations up to 50 ~ L M (Fig. 2 A ) . In contrast to results obtained with ETPH, the PhAsO sensitivity of Pi-ATP exchange in reconstituted AE-P was not potentiated by 2-mercaptoetha- nol. Higher sensitivity of reconstituted AE-P to PhAsO, as well as inability of 2-mercaptoethanol to potentiate inhibition by PhAsO as compared to ETPH, may be related to easier accessibility of the inhibitor to the dithiol site after depletion- reconstitution. The environment of the binding site may be more nonpolar (or membrane shielded) in the intact ETPH. This is consistent with OUT observations that Fe-stimulated activity of AE-P is more easily inhibited by anti-Fg serum than the activity of endogenous FB in the ETPH (16).

Cadmium and BAL-arsenite inhibited in a manner similar to that of PhAsO (data not presented), the K, for Cd2' being 30 p ~ . Uncoupling by arsenite was dependent upon the pres- ence of equimolar BAL as has been observed earlier (3). Of the three types of dithiol reagents tested, PhAsO was the most potent inhibitor of Fg activity. This may be related to the higher specificity of PhAsO for dithiol sites as well as easier penetration to the dithiol site on Fg.

The inhibition by PhAsO was largely overcome by equi- molar BAL but minimally affected by a 10-fold excess of 2- mercaptoethanol (Fig. 2, A and B ) .

Effect of PhAsO on F B versus Oligomycin-stimulated Ac- tivity of AE-P-Low levels of oligomycin stimulate phospho- rylation and energy-linked reactions in AE-P (10). The effects

11114 Cadmium Inhibition of Mitochondrial Coupling Factor B

160 A

0- t50pM BAL x

Tc ._ 120 R E

a 2i a .- - 0

E

(F6tPhAsO)tAE-P

4- +I50 pM 2-ME

+50pM BAL

0

E 25 50

0 I , I

25 50

pM PhAsO

0- t50pM BAL

25 50 500 c pM PhAsO

FIG. 2. Effect of phenylarsine oxide on the Pi-ATP exchange activity of AE-particles. Aliquots containing 0.5 mg of AE-particles and/or FB were preincubated with indicated concentrations of phen- ylarsine oxide in a total volume of 50 pl for 5 min on ice. The samples were assayed for Pi-ATP exchange activity by incubating at 23 "C for 10 min in a total volume of 0.5 ml containing 100 pmol of Tris-SO, buffer (pH 8.0), 0.5 pmol of MgC12, 2.5 mg of bovine serum albumin, and 2-mercaptoethanol(2-ME) or BAL where indicated. The reaction was initiated by adding 0.5-ml aliquot containing 10 pmol of 32Pi (2 to 3 X IO5 cpm), 15 -01 of ATP, 5 pmol of ADP, and 20 pmol of MgC12. After 15 min at 38 "C, the reaction was terminated by the addition of 0.5 ml of 20% trichloroacetic acid. In A, FB was preincubated with phenylarsine oxide prior to its reconstitution with AE-particles. In B,

phenylarsine oxide. (M, AE-P + FB + PhAsO; M, AE-P Factor B was mixed with AE-particles prior to its preincubation with

+ PhAsO, A, AE-P + FB + 15 PM PhAsO + 150 phf 2-mercaptoethanol; 0, AE-P + FB + 15 p~ PhAsO + 50 p~ BAL; ., AE-P + 15 PM PhAsO + 50 p~ BAL).

of F B and low and high oligomycin concentrations on the energy-linked binding of the voltage-sensitive dye, oxonol VI, to AE-P membranes were compared. The rate and extent of binding as well as the rate of discharge of the bound dye were measured3 (17). Low oligomycin increased the rate of binding with both NADH and ATP as substrates, and also partially inhibited the discharge. On the other hand, FB stimulated the binding with only ATP as the substrate, but had no significant effect on the rate of discharge with either substrate. The results are consistent with the suggestion of Lee and Emster (18) that low oligomycin increased the activity of AE-P by blocking energy leak while the data with FB are inconsistent with a similar explanation.

The basal and the oligomycin-stimulated activities of the AE-P are presumably dependent upon residual FB in the membrane. Consistent with this, the PhAsO concentration required to inhibit their activities was similar to that seen with ETPH and significantly higher than the concentrations for inhibition of the activity of Fe-reconstituted AE-P. (Com- pare Figs. 1 and 2.) Negligible effect was obtained with 17 p~ PhAsO, and 85 p~ PhAsO gave barely 50% inhibition.

Effect of PhAsO on FB Binding to AE-P"To test the possibility that the dithiol inhibitors interfere with F5 binding to the particles, aliquots of particles were preincubated f FB in the presence of PhAsO. Aliquots were then centrifuged to remove unbound FB and inhibitor, and the washed particles were assayed f dithiol (Table 11). During preincubation, PhAsO does not bind to any component of AE-P that is essential for FB-stimulated energy-dependent activities, since control AE-P and PhAsO-treated AE-P, when tested after centrifugation step with additional Fg, have the same activity (59 versus 61 nmol min" mg"). Furthermore, the addition of dithiol produced a marginal increase in their activity. Second, in the presence of FB and AE-P, PhAsO does bind to some

Unpublished.

site that is required for FB-stimulated activity in AE-P. The activity of the AE-P + FB reconstituted in the presence of PhAsO (12.1 nmol min" mg") is considerably less than the activity of the respective control (53 nmol min" mg"). Fi- nally, PhAsO does not interfere in the binding of F5 to AE-P since the activity of AE-P + FB + PhAsO particle (53.2) (line 4) approaches that of AE-P + FB (57.3) (line 3), when tested in the presence of dithiol, after the centrifugation step. Al- though there is a slight decrease in overall reconstituted activities of AE-P + F B after the centrifugation step, probably due to loss of some F5, the general pattern of PhAsO effect on Fg binding remains unaffected.

Does FB Bind PhAsO?-To test this possibility, FB was allowed to incubate with PhAsO and then was passed through a Sephadex G-25 column to remove unbound inhibitor. FB used in this experiment showed a single band in sodium dodecyl sulfate polyacrylamide gel electrophoresis (11). Inhib- itor-pretreated FB lost the coupling activity as compared to its respective control which caused a 3- to 5-fold increase in AE-P activity. Moreover, the coupling activity was fully re- stored by 30 p~ BAL but not by similar levels of P-mercap- toethanol (Fig. 3). This demonstrates that FB binds PhAsO. Similar results were obtained when Cd" was used instead of PhAsO (data not presented).

Effect of PhAsO on the Pi-ATP Exchange Activity of H + - ATPase-FB is a component of the H+-translocating ATPase and is required for the Pi-ATP exchange activity (1 1). Con-

TABLE XI Effect ofphenylarsine oxide on binding of Factor B to AE-P

Ten mg of AE-particles f Factor B were preincubated with 1 pmol of phenylarsine oxide in 2 ml for 5 min on ice, and then diluted with 8 ml of 2 mM Tris-C1, pH 7.0,0.25 M sucrose. Samples were centrifuged at 100,OOO X g for 40 min. The sediments were resuspended in 0.25 M sucrose and assayed for PI-ATP exchange activity as described in Fig. 2. Dithiothrietol (DTT) (5 pmol) and/or Factor B (0.5 unit) were added during the assav where indicated. ~~~~

P,-ATP exchange Additions during assay

None Fe DTT :& nmol rnin-l mg"

Additions during preincu- bation

AE-P 8.6 59 8.9 71 AE-P + PhAsO 8.9 61 10.9 68 AE-P + FB 56.2 57.3 AE-P + FB + PhAsO 12.2 53.2

- BAL

2 X 6o t 7-

pM Thiol FIG. 3. Reactivation of Factor B by 2-mercaptoethanol uer-

su8 BAL. Factor B (0.05 mg) purified through Sephadex G-75 step, was preincubated on ice for 5 min in a final volume of 100 111 containing 100 PM phenylarsine oxide. To remove unbound inhibitor, the FB aliquot was passed through a Sephadex G-25 column (1 X 10 cm) equilibrated with 20 mM Tris-SO4 pH 8.0 buffer. Fractions containing protein were pooled and 1-4 aliquots were assayed for coupling factor activity by addition to AE-particles and measuring stimulation of NAD+ reduction activity as described in Fig. 1. BAL or 2-mercapto- ethanol was added where indicated, during the 2-min incubation at 38 "C.

Cadmium Inhibition of Mitochondrial Coupling Factor B 11115

TABLE I11 Inhibition of Pi ATP exchange actiuity of H'-ATPase complex by

phenylarsine oxide T W O hundred-pg aliquots of H+-ATPase were preincubated on ice

for 5 min with indicated concentrations of phenylarsine oxide and/or 2-mercaptoethanol in a total volume of 50 pl. Samples were next assayed for P,-ATP exchange as described in Fig. 2. ATPase activity was assayed by incubating 100 pg of H+-ATPase (preincubated with phenylarsine oxide as described above) in a volume of 0.5 ml contain- ing 10 pmol of Tris-SO4, pH 7.5, 2.5 pmol of ATP, 2.5 pmol of MgClz. After 10 min at 30 "C, the reaction was terminated by the addition of 0.25 ml of 20% trichloroacetic acid.

Pi-ATP ex- ATPase

nnwl pnol P, change

Experiment 1. Additions to 200 pg H'-ATPase - 555 1.5 + 10 jtM PhAsO 186 2.4 + 50 pM PhAso 87 3.7 + 250 pM PhASo 24 4.2 + 1 m~ PhAsO 6 3.9

- 555 + 50 PM PhAsO + 50 p~ 2-mercaptoethanol 21 + 50 CM PhAsO + 150 PM 2-mercaptoethanol 12 + 50 p~ PhAsO + 500 p~ 2-mercaptoethanol 13

Experiment 2

sistent with this is the observation that Pi-ATP exchange activity of undepleted H+-ATPase is sensitive to PhAsO, although somewhat higher inhibitor levels are needed for complete inhibition as compared to ETPH (Table 111). The H+-ATPase used for these experiments purified by the lyso- lecithin extraction procedure and sucrose density gradient sedimentation, catalyzes Pi-ATP exchange activity of 1,000 to 1,400 nmol min" mg" (19). The preparation has essentially 13 bands in sodium dodecyl sulfate-polyacrylamide gel elec- trophore~is.~

The inhibitory effects of PhAsO on H+-ATPase were poten- tiated by 2-mercaptoethanol, just as in the case of ETPH. Also, the decrease in Pi-ATP exchange activity was accom- panied by an increase in ATPase activity (Table 111). Both the decrease in Pi-ATP exchange and increase in ATPase activity were reversed by the addition of low levels of BAL but not 2-mercaptoethanol (data not presented).

These experiments demonstrate that the dithiol site in- volved in Cd2+ and arsenite uncoupling of mitochondria is also present in the purified H+-ATPase. The other known com- ponents of H+-ATPase (F1-ATPase, oligomycin sensitivity- conferring protein, coupling factor 6, proteolipid) have not so far been reported to contain a dithiol that is essential for ATP-driven NAD+ reduction or Pi-ATP exchange activities.

DISCUSSION

Earlier work on dihydrolipoyl dehydrogenase provided the first evidence for the direct participation of a dithiol group in catalytic activity (20). It was shown that the dehydrogenase activity was strongly inhibited by low levels of cadmium chloride and arsenite, provided the enzyme was previously reduced by NADH. Reduction of the dehydrogenase was accompanied by exposure of two "SH groups and appearance of a characteristic absorption peak at 530 nm, both of which disappeared upon addition of Cd2+ or arsenite. Later, a cad- mium derivative of this enzyme was isolated and characterized (21). Based on these findings, two criteria were established for involvement of a dithiol group in enzyme activity, namely, inhibition by low levels of Cd2+ or arsenite and reversal of the

Unpublished data.

inhibition by a small excess of dithiols but not by even a large excess of monothiols. The first criterion alone is not sufficient since the apparent sensitivity of a monothiol is known to vary with its accessibility (22). The second criterion, testing whether the functional group in question can compete with dithiols for the inhibitor, provides a better rationale for decid- ing whether the site is a mono- or a dithiol. Based on these criteria, the earlier observations on uncoupling of oxidative phosphorylation by cadmium, and preferential release of un- coupling by di- but not monothiols, it was tentatively proposed that the mitochondrial coupling process may involve a dithiol group (1-6). Further, the dithiol group was assumed to be located in a hydrophobic environment since it was affected by arsenite only when equimolar BAL was present. BAL was presumed to function as a carrier for arsenite.

Our recent work on FB demonstrating its involvement in converting oligomycin-sensitive ATPase to an energy-trans- ducing complex, prompted consideration of FB as the possible site of the essential dithiol. Preliminary observations along these lines indicated (19) that FB stimulation of Pi-ATP ex- change and NAD+ reduction by succinate in AE-P was sen- sitive to BAL-arsenite and Cd2+. The inhibition was relieved by BAL but not by monothiols. Stigall et al. (23) also reported that FB-mediated ATP-driven NAD' reduction in AE-P was inhibited by Cd2+, phenylarsine oxide, p-chloromercuribenzo- ate, and diamide but did not test for preferential reversal by dithiols.

Present work using intact submitochondrial particles estab- lishes that the dithiol site involved in the coupling process is not located in the electron transport chain but probably in the H+-ATPase. Following extraction of FB, the AE-P lose most of their coupling activity along with the sensitivity to dithiol inhibitors, both of which are restored upon addition of FB. Addition of oligomycin partially restores coupling activity to AE-P but not sensitivity to Cd" and PhAsO, again sug- gesting involvement of FB in this inhibition of coupling activ- ity. Experiments on Cd', or PhAsO binding by FB demonstrate that FB has a dithiol site essential for its coupling activity. Furthermore, the requirement for equimolar BAL in addition to arsenite, to bring about the inhibition, as well as potentia- tion of phenylarsine oxide inhibition by 2-mercaptoethanol point to the tentative conclusion that the dithiol site on FB may be shielded in a hydrophobic environment. Implication of the FB dithiol in the coupling process is further strengthened by the observation that both ATP-Pi exchange and coupling of oligomycin-sensitive ATPase activities, known to involve FB (ll), are also sensitive to PhAsO in the purified H+-ATPase. This, together with the observation that FB has 1 mole each of cysteine and cystine/monomer, allow us to tentatively identify FB as the dithiol component involved in mitochondrial uncoupling by Cd" and arsenicals. How FB brings about the coupling and whether the dithiol has a specific chemical role or merely contributes to the structural requirements for function, remain to be studied.

Acknowledgments-We are grateful to Dr. D. R. Sanadi for helpful suggestions and to Ms. Louise Reed for assistance in the preparation of this manuscript. We also gratefully acknowledge the expert tech- nical assistance of Ms. Diane Ascoli.

REFERENCES 1. Jacobs, E. E., Jacob, M., Sanadi, D. R., and Bradley, L. B. (1956)

2. Fluharty, A., and Sanadi, D. R. (1960) Proc. Natl. Acad. Sei. U.

3. Fluharty, A. L., and Sanadi, D. R. (1961) J. Biol. Chem. 236,

4. Fletcher, M. J., and Sanadi, D. R. (1962) Arch. Biochem. Biophys.

J. Biol. Chem. 223, 147-156

S. A. 46,608-616

2772-2778

96, 139-142

11116 Cadmium Inhibition of Mitochondrial Coupling Factor 3

5. Fluharty, A. L., and Sanadi, D. R. (1963) Biochemistry 2.519-522 6. Fletcher, M. J., Fluharty, A. L., and Sanadi, D. R. (1962) Biochim.

7. Sekuzu, I., Jurtshuk, P., Jr., and Green, D. E. (1963) J. Biol.

8. Fluharty, A. L., and Sanadi, D. R. (1962) Arch. Biochem. Biophys.

9. Hatch, M. D., and Stumpf, P. K. (1961) J. Biol. Chem. 236,2879-

10. Lam, K. W., Warshaw, J. B., and Sanadi, D. R. (1967) Arch.

11. Joshi, S., Hughes, J. B., Shaikh, F., and Sanadi, D. R. (1979) J.

12. Serrano, R., Kanner, B. I. and Racker, E. (1976) J. Biol. Chem.

13. Lam, K. W., Swann, D., and Elzinga, M. (1969) Arch. Biochem.

14. Joshi, S., and Sanadi, D. R. (1979) Methods Enzymol. 55F, 384-

Biophys. Acta 60,425-427

Chem. 238,975-982

97, 164-167

2885

Biochem. Biophys. 119,477-484

Biol. Chem. 254, 10145-10152

251,2453-2461

Biophys. 130, 175-182

39 1 15. Beyer, R. E. (1967) Methods Enzymol. 10, 186-194 16. Lam, K. W., and Yang, S. S . (1969) Arch. Biochem. Biophys. 133,

17. Hughes, J. B., Joshi, S., and Sanadi, D. R. (1980) European Bioenergetics Conference, Urbino, Italy, pp. 193-194

18. Lee, C. P., and Ernster, L. (1965) Biochem. Biophys. Res. Com- mun. 18,523-529

19. Hughes, J. B., Joshi, S., Murf~tt, R. R., and Sanadi, D. R. (1979) in Membrane Bioenergetics (Lee, C. P., Schatz, G., and Ernster, L., eds) pp. 81-95, Addison-Wesley, Reading, MA

20. Searls, R. L., Peters, J. M., and Sanadi, D. R. (1961) J. Biol. Chem. 236,2317-2322

21. Stein, A. M., and Stein, J . H. (1971) J. Biol. Chem. 246,670-676 22. Marshall, M., and Cohen, P. P. (1980) J. Biol. Chem. 255, 7296-

23. Stiggall, D. L., Galante, Y. M., Kiehl, R., and Hatefi, Y. (1979)

366-372

7300

Arch. Biochem. Biophys. 196,638-644