A Q-cycle mechanism for the cyclic electron ... - Semantic Scholar

5 downloads 0 Views 316KB Size Report
(Crofts et al., 198 1). The experimental results and conclusions which have led us to reevaluate the Q-cycle are discussed below. A ntimycin-inhibited chain.
599th MEETING, BIRMINGHAM

20 1

A Q-cycle mechanism for the cyclic electron-transferchain of Rhodopseudomonas

sphaeroides ANTONY R. CROFTS and STEVE W. MEINHARDT Department of Physiology and Biophysics, University of Illinois, Urbana,I1 61801, U S A Over the past few years, work from a number of laboratories has shown that the ubiquinol :cytochrome c, (QH, :cyt. c,) oxidoreductase of the cyclic photosynthetic electron-transport chain of R hodopseudomonas sphaeroides is essentially similar to the mitochondria1 complex 111. Scheme 1 summarizes our present knowledge of the thermodynamics, kinetics and stoichiometry of the components of the chain, shown operating as a Q-cycle (Mitchell, 1976). We have previously pointed out that several features of the kinetics following flash illumination appeared to be inconsistent with Q-cycle mechanisms proposed to operate in this system (Crofts et al., 1975, 1981). In reviewing a number of recent results from our own work and that of other laboratories, we have now been able to show that the data can be accounted for by a Q-cycle, but only if this is formulated within narrow constraints. Recent information pertinent to our new formulation is summarized below. (i) The Rieske-type FeS centre has been characterized, with the midpoint redox potential, stoichiometry and kinetics shown in Scheme 1 (Bowyer et al., 1979, 1980). (ii) A bound cyt. c analogous to mitochondria1 cyt. c, has been shown to be present in the chain (Wood, 1980; Bowyer et al., 1981). (iii) The kinetics, stoichiometry and Em (midpoint potential)

Qz

I

values of cyt. c , and cyt. c, in chromatophores have been determined (Crofts et al., 1981), as shown in Scheme 1. (iv) The low potential b-type cytochrome (cyt. b,,,) has been shown to be a kinetically competent member of the chain (Bowyer & Crofts, 1981; Crofts et al., 1981). (v) The stoichiometries of the other components of the QH,:cyt. c, oxidoreductase complex in relation to the reaction centre (one complex per two reaction centres) have been assigned, assuming that the stoichiometries with respect to cyt. c, and cyt. b,,, are the same as those in mitochondria (Crofts et al., 198 1). The experimental results and conclusions which have led us to reevaluate the Q-cycle are discussed below. A ntimycin-inhibited chain (1) Because of the relative midpoint potentials of the cyt. c, and FeS couples, and the sequence of their reaction on oxidation of reaction-centre bacteriochlorophyll (P), the FeS centre will go oxidized only after a lag introduced by the time taken to partially oxidize cyt. cI' This lag (about 200ps) is seen experimentally when one measures the difference between the kinetics of (c,+c,+P) in the presence of antimycin, and of antimycin and 5-n-undecyl-6-hydroxy-4,7-dioxobenzothiazole (Bowyer & Crofts, 1981)). Because of the relative E m values of FeS and cyt. c,, and the stoichiometry with respect to reaction centre, FeS can be reduced by QH, while cyt. c, remains oxidized after a single turnover flash. This last conclusion invalidates many of our previous objections to the Q-cycle. (2) At values of E , (redox potential) when QH, is reduced

4

-loo 1300)

13001

50

-90

Scheme 1. Components of the cyclic electron-transport chain arranged as a Q-cycle Numbers in parentheses are half-times in ps for the reactions indicated. Numbers in square brackets are the stoichiometries of components (oxidized + reduced forms) above them. Unbracketed numbers are values for Em at p H 7 (in mv) for components above them, except that for the semiquinone couples the Em values are shown between the components of the half-cell. Continuous lines with continuous arrows show electron transfers. Continuous lines with broken arrows show chemical conversions. Broken lines show equilibration with the quinone pool. QH, is ubiquinol. Q-' and QH' are the unprotonated and protonated forms of ubisemiquinone and Q is ubiquinone. Q, and Q, are the first and second ubiquinone electron acceptors in the photosynthetic reaction centre. Qc-' is the stable semiquinone described by Ohnishi & Trumpower (1980). QzH, is the ubiquinol donor to FeS, the Rieske iron-sulphur protein. Qp is the ubiquinone pool. VOl. 10

202

BIOCHEMICAL SOCIETY TRANSACTIONS

before a flash, but cyt. b561is oxidized, the kinetics of reduction of cyt. b561closely match the kinetics of electron transfer from FeS to cyt. c , (Bowyer & Crofts, 198 I), showing a lag of about 200ps, and half time of reduction of about 3 0 0 , ~ sfor the change occurring after the lag. In view of this close match, the reactions by which FeS and cyt. b5,, are reduced by QH, must occur without any delay, so that the whole reaction is concerted and is determined essentially by the rate at which FeS is oxidized. (3) The kinetics and stoichiometries we have measured put some useful constraints on the system. Thus our measurements are compatible with a Q-cycle only if ( a ) cyt. c , has a lower Em than FeS, and (b) the stoichiometry of the complex is about 1:2 with respect to P, and (c)the quinone reactions are concerted. (4) Reduction of cyt. b5,, is seen only if cyt. b,,, is reduced before the flash (Crofts et al., 198 1). Under the redox conditions stated above, complete reduction of cyt. b,,, is observed on the first flash, and cyt. b566becomes largely reduced on a second flash (given 8.33ms later). This could occur in a Q-cycle only if the QH, used up in reducing FeS and cyt. b,,, had been regenerated. If we assume that the antimycin block is essentially complete, and that the Q, site holds only one equivalent, then the QH, can only be replaced if the quinone has vacated the site, and been replaced by a quinol. It seems likely that the site must be able rapidly to equilibrate with the quinol pool. (5) At redox potentials where Q is oxidized before the flash, the reduction of cyt. b,,, occurs much more slowly, and speeds up on reductive titration with the same apparent Em and n value as Q,/QzH, (Evans & Crofts, 1974; Bowyer & Crofts, 1981). The Q-cycle provides an obvious explanation for this phenomenon. ( 6 ) If the Q-cycle is the only pathway for cyt. bJ6,reduction, as seems likely if the reaction is concerted, then the slow reduction seen at higher E , values must reflect the rate (ti 7ms) at which Q, H, reduced in the photoreactions diffuses to the Q,

4

QB

QC

+ +

+-

--

t

4

+ +

+-

--

++

5

+ +

+-

--

0

++

+-

++

State

Qa

Qa

b561

site, giving us a measure of this rate. This hypothesis also accounts for the pattern of two we had seen in the reduction of cyt. b,,, as a function of flash number at E , > 300mV and which we had interpreted as showing that full reduction of the quinone in the two-electron gate was required before electrons could get to cyt. b,,, (Bowyer et al., 1979). (7) The various phenomena associated with the concerted reaction-oxidant-induced reduction of cytochrome b. 5-nundecyl-6-hydroxy-4,7-dioxobenzothiazoleinhibition of cyt. b reduction, loss of cyt. b reduction in spheroplasts and in various mutants blocked in the high-potential chain (Bowyer, 1979; Bowyer & Crofts, 1 9 8 1 t a r e now all simply explained by the Q-cycle.

Uninhibited chain It has been shown previously that the oxidation of cyt. bJ6,, the antimycin-sensitive reduction of the c-type cytochromes, and the slow phase of the carotenoid change, shared a number of common characteristics-antimycin-sensitivity, the kinetics of the antimycin-sensitive process, the dependence on the state of reduction of Qz (Prince & Dutton, 1975; Crofts et al., 1975: Bowyer & Crofts, 1981). It was clear from these observations that the three phenomena were part of a common kinetic process. We had concluded that the oxidation of cyt. bJ6,was kinetically linked to the reduction of cyt. ( c ,+ c,), and that the oxidation was part of the electrogenic process. We had previously believed that these results were incompatible with a Q-cycle type of mechanism, and that they indicated instead a linear chain. However, the observations can now be readily interpreted in the light of the Q-cycle shown above. (8) The kinetics of the antimycin-sensitive processes show a lag after the flash of about 500,~s.and a half-time for the subsequent change of about 700ps (Bowyer & Crofts, 1981). It will be clear from the discussion above that the extra electron

b566

9

FeS

C1

c2

P

P

0 100ns 1

200ps 2

300p

3

700 p s

1 ms?

-

--

+++?

-

--

+& 5 - -+ -

L-

-

-

--

Scheme 2. Idealized distribution of electrons (-) and oxidizing equivalents (electron holes, + ) f o r successive states of the Q-cycle The starting state is the chain poised at E , about 100mV. The arrows show the fate of electrons on going from one state to the next. The State-0-State-l transition is the photochemical reaction. The transitions from State 3 to State 4, and from State 5 to State 0 are electrogenic. The transition from State 3 to State 4 is inhibited by antimycin. In the inhibited chain, the quinone left at the Q, site after formation of State 3 can be replaced by quinol from the pool or from Q B H YUnder these circumstances, a second flash can pull the concerted reaction over so as to reduce cyt. b,,,.

1982

203

599th MEETING, BIRMINGHAM arriving at ( c , + c , + P ) in the absence of antimycin must represent the oxidation of a second equivalent of QH, by the cyt. cI which remained oxidized after the first cycle of the complex. This second electron transfer presumably also occurs via the FeS centre, so that effectively the ‘invisible’ electron on the FeS centre gets replaced and pushed over on to cyt. c, in a second turnover of the complex. It is important to realize that this second turnover of the complex is a necessary consequence of the relative stoichiometries of the complex and P. Every time the reaction centre turns over, the complex has to turn over twice to return things to the starting state. (9) Because of the constraints discussed above, it is clear that the electrogenic process is the oxidation of cyt. b,,]. This suggests that the redox drop in this reaction must be substantial (>200mV). (10) The antimycin-sensitive reduction of the c-cytochromes must reflect the regeneration of Q,H, in the uninhibited cycle from an oxidized entity (presumably Q’H or Q-*) which receives electrons from cyt. b,,, and which must be available to the cycle in a time less than, or equal to, the 500ps lag observed for these processes. The lag could reasonably be viewed as arising from the time taken to reduce cyt. b,,,. One area in which we have insufficient information to define the cycle is the pathway by which the entity above is generated. It has been suggested by Slater and colleagues (Berden et al., 1981 ;de Vries et al., 1981), that the stable semiquinone, Qc-* (Ohnishi & Trumpower, 1980) is formed at a binding site in equilibrium with the pool, by simple disproportionation, and that this is the acceptor of electrons from cyt. b,,, in mitochondria. ( 1 1) To complete the cycle, it would be necessary to regenerate the QC-* used in the first turn and to reoxidize cyt. bS6,. The simplest assumption would be that the oxidant is Q. either from the pool, or from the cycle. This would mean that cyt. b,,, could donate electrons to Q or Q-’ at the site on the cytoplasmic side of the bacteria. Effectively this site would then have the same mechanism as the ‘two-electron gate’ of the reaction-centre secondary quinone. It would also be necessary to replace the quinone at the Q, site with a quinol. For ‘book-keeping’ purposes, this would have to be the quinol generated in the photoreactions, though in practice it might well come from the pool.

We have attempted to summarize in Scheme 2 the electron distribution, and the fates of the electrons, as the cycle goes through a succession of states. Obviously, these are highly idealized, but they provide a compact and comprehensible description of the Q-cycle in terms of reactions which can be written out in conventional chemical form. Space does not permit a more detailed consideration of mechanistic aspects which follow from the parameters we have measured, and the constraints these put on a Q-cycle. The cycle explains with beautiful economy many of the phenomena which we have measured over the past few years-an economy which complements the set of constraints provided by the measurements themselves. Berden, J. A., de Vries, S. & Slater, E. C. (1981) in Function of Quinones in Energy Conserving S.vstems (Trumpower. B.. ed.). Academic Press, New York. in the press Bowyer, J. R. (1979) Ph.D. Dissertation, University of Bristol Bowyer, J. R. & Crofts, A. R. (1981) Biochim. Biophvs. Acta 636, 218-233 Bowyer, J. R., Tierney, G. V. & Crofts, A. R. (1979) FEBS Lett. 101, 20 1-206 Bowyer, J. R., Dutton, P. L., Prince, R. C. & Crofts, A. R. (1980) Biochim. Biophys. Acta 592,445-460 Bowyer, J. R., Meinhardt, S.W., Tierney. G. V. & Crofts, A. R. (1981) Biochim. Biophys. Acta 635, 167-186 Crofts, A. R., Crowther, D. & Tierney, G. V. (1975) in EIectron Transfer Chains and Oxidative Phosphorviation (Quagliariello.E.. Papa, A.. Palmieri, E., Slater. E. C. & Siliprandi. N.. eds.). pp. 233-24 1, North-Holland. Amsterdam Crofts, A. R., Meinhardt, S. W. & Bowyer. J. R. (1981) in Function of Quinones in Energy Conserving S.vstems (Trumpower, B.. ed.). Academic Press, New York. in the press de Vries, S., Berden, J. A. & Slater. E. C. (1981) in Function of Quinones in Energy Conserving Svstems (Trumpower, B.. ed.). Academic Press, New York, in the press Evans, E. H. & Crofts, A. R. (1974) Biochim. Bi0ph.w. Acta 357. 89-102 Mitchell, P. (1976)J. Theor. Bid. 62. 327-367 Ohnishi, T. & Trumpower. B. L. (1980)J. Bid. Chem. 255. 3278-3284 Prince, R. C. and Dutton, P. L. (1975) Biochim. Biophw. Acta 387. 6 0 9 4 I3 Wood, P. M. (1980) Biochem. J . 189,385-391

Conservation of structure in proton-translocating ATPases of Escherichia coli and mitochondria JOHN E. WALKER, ALEX EBERLE.’ NICHOLAS J. GAY, MICHAEL J. RUNSWICK and MATTI SARASTE Laboratory of Molecular Biology, the M.R.C. Centre, Hills Road, Cambridge C B 2 ZQH, U.K. The mitochondria1 and bacterial (and also chloroplast) H+translocating ATPase complexes have remarkably similar structures [for reviews, see Senior (1979). Fillingame (1981) and Racker (1981): see also Figs. 1 & 21. They are membranebound enzymes composed of a membrane sector. F,. with an attached extra-membrane portion. F ,. F,, contains a transmembrane proton channel. Under physiological conditions it couples the proton gradient across the membrane to drive the formation of ATP from ADP and Pi (Mitchell. 1981). F , contains catalytic and regulatory sites that bind ATP and ADP (Harris. 1978). In all species examined. F , contains five different polypeptides a, P, y, 6 and E (Fig. 2), with three copies

* Present address: Research Department (ZLF). Kantonsspital. Hebelstrasse 20. CH-4031 Basel. Switzerland. Vol. 10

of both a and Pand one each of y. 6 and E (Catterall & Pedersen, 1971: Bragg & Hou, 1975). F,, from Escherichia coli is made from three polypeptides, a, b and c (Fillingame, 1981) (Fig. 1) and counterparts are also present in bovine F,, (Fig. 2). These similarities notwithstanding. striking differences are also to be found. The mitochondrial enzyme is more complex than the bacterial one: for example. extra subunits are associated with F , (see Fig. 2). The best studied are the inhibitor protein (I) (Pullman & Monroy, 1963) and the oligomycinsensitivity-conferral protein (oscp) (Senior. 197 I). F, is less well characterized in mitochondria, but may also have additional subunits (Galante et al.. 1979: Glaser et al., 1981). Recently DNA sequences, and thereby amino acid sequences. of the entire E. coli complex (Gay & Walker, 1981a.b: Saraste e f al., 1981; Walker et al., 1 9 8 2 ~and ) of constituent proteins of the bovine complex have been determined (Anderson et al., 1982: M. Runswick & J. E. Walker. unpublished work: E. Wachter, personal communication), thereby allowing us to analyse these similarities in closer detail. These studies show that different parts of the complex have evolved at different rates. The a- and P-subunits containing regulatory and catalytic sites