Temperature dependence of charge recombination from the P+QA- and P+QB- states in photosynthetic...

Eur. J. Biochem. 202, 625-634 (1991) 0 FEBS 1991

Temperature dependence of charge recombination from the P+ QA and P+ QB states in photosynthetic reaction centers isolated from the thermophilic bacterium Chloroflexus aurantiacus Giovanni VENTUROLI ', Massimo TROTTA', Reiner FEICK ', B. Andrea MELANDRI' and Davide ZANNONI'

Dipartimento di Biologia, Universita' di Bologna, Italy Centro Studi Chimico-Fisici sulla Interazione Luce-Materia, CNR, Bari, Italy Max-Planck lnstitut fur Biochemie, Martinsried, Federal Republic of Germany

(Received May 13/July 26, 1991) - EJB 91 0612

The temperature dependence of charge recombination from the P'QA and from the P'QS states produced by a flash was studied in reaction centers isolated from the photosynthetic thermophilic bacterium Chlorojlexus aurantiacus. P designates the primary electron donor; QA and Qe the primary and secondary quinone electron acceptors respectively. In QB-depleted reaction centers the rate constant (kAp) for P'QA recombination was temperature independent between 0-50°C (17.6 & 0.7 s - l at pH 8 and pH 10). The same value was obtained in intact membranes in the presence of o-phenanthroline. Upon lowering the temperature from 250 K to 160 K, kAP increased by a factor of two and remained constant down to 80 K. The overall temperature dependence of kAP was consistent with an activationless process. Ubiquinone (UQ-3) and different types of menaquinone were used for QB reconstitution. In UQ-3-reconstituted reaction centers charge recombination was monoexponential (rate constant k = 0.18 0.03 s-') and temperature independent between 5-40°C. In contrast, in menaquinone-3- and menaquinone-4-reconstituted reaction centers P + rereduction following a flash was markedly biphasic and temperature dependent. In menaquinone-6-reconstituted reaction centers a minor contribution from a third kinetic phase corresponding to P'QA charge recombination was detected. Analysis of these kinetics and of the effects of the inhibitor o-phenanthroline at high temperature suggest that in detergent suspensions of menaquinone-reconstituted reaction centers a redox reaction removing electrons from the quinone acceptor complex competes with charge recombination. Instability of the semiquinone anions is more pronounced when QB is a short-chain menaquinone. From the temperature dependence of P + decay the activation parameters for the P'Q; recombination and for the competing side oxidation of the reduced menaquinone acceptor have been derived. For both reactions the activation enthalpies and entropies change markedly with menaquinone chain length but counterbalance each other, resulting in activation free energies at ambient temperature independent of the menaquinone tail. When reaction centers are incorporated into phospholipid vesicles containing menaquinone-8 a temperature-dependent, monophasic, o-phenanthroline-sensitive recombination from the P'Q; state is observed, which is consistent with the formation of stable semiquinone anions. This result seems to indicate a proper QB functioning in the two-subunit reaction center isolated from Chlorojlexus aurantiucus when the complex is inserted into a lipid bilayer.

Reaction centers of photosynthetic organisms are membrane-bound pigment-protein complexes that convert the photon energy absorbed by the antenna system into electro- chemical energy. In two species of purple photosynthetic bacteria (Rhodopseudomonas viridis and Rhodobacter sphaeroides) the three-dimensional structure of the reaction center (RC) has been determined to atomic resolution by X-ray diffraction. The RC is composed of two membrane-spanning polypeptides called L and M, containiiig five transmembrane a-helical re-

Correspondence to G. Venturoii, Dipartimento di Biologia, Univcrsita' di Bologna, via Irnerio 42, 1-40126 Bologna, Italy

Abbreviations. C'., Chloroflexus; LDAO, lauryldimethylamine N-oxide; P, primary electron donor of the reaction center; QA, primary quinme acceptor; QR, secondary quinone acceptor; Rb., Rhodohactei; RC, reaction center; Rps., Rhodopseudomonas; UQ, ubiquinone.

gions each. These two polypeptides bear the cofactors involved in charge separation. A third polypeptide chain, subunit H, consists of one membrane-spanning helice and several P-sheets organized into a globular domain which caps the LM complex [l - 51.

The first light-induced charge separation occurs between the primary electron donor, a bacteriochlorophyll dimer (P), and a bacteriopheophytin molecule bound to the L subunit. The electron is then transferred to a pair of quinone molecules, acting in sequence as primary (QA) and secondary (QB) quinone acceptors (for a review see [6]).

The photosynthetic RC isolated from the gliding thermophilic bacterium Chlorojlexus aurantiucus catalyzes a quite similar electron transfer sequence which involves a photoactive bacteriochlorophyll a dimer, bacteriopheoph ytin as an intermediate acceptor and a two-menaquinone acceptor

626

system [7 - 111. The protein component of this photochemical entity, however, possesses three unique features as compared to purple bacteria RCs: (a) the complex is thermally stable up to about 50°C [12, 131; (b) the RC is composed of only two protein subunits 112, 141 and (c) contains three bacteriochlorophyll and three bacteriopheophytin molecules as compared to four and two, respectively in purple bacteria RCs [14]. The two polypeptide chains of Cfl. auruntiacus are very similar to the L and M subunits of Rb. sphaeroides and Rps. i1iridi.s RC [ 1 5 - 1 71.

In RCs from Rb. sphaeroides R-26, removal of the H subunit is known to affect drammatically electron transfer within the quinone acceptor complex [18]. Dissociation of H results in several effects which include: (a) decrease in the rate of electron transfer from QI\ to QR by a factor of 102-103; (b) reduction by a comparable factor in the stability of the semiquinone anions QA and QK produced by light activation; (c) reduced sensitivity to inhibitors of QA to QB electron transfer; (d) a tenfold decrease in the binding affinity of ubiquinone (UQ) 10 at the QB site and (e) functional loss of Q,$ as a two-electron 'gate'. Reduction by a factor of two in the rate of charge recombination between P + and QA is also observed. Subunit H is therefore mainly and critically involved in determining binding and reactivity at QB [18].

In view of these observations it is of interest to test the functioning of the quinone acceptor complex in the two-subunit RC isolated from Cjl. aurantiacus, which is thought to operate in vivo without a third polypeptide chain. In the present paper this problem has been addressed by undertaking a kinetic and thermodynamic characterization of the charge recombination reactions between the photo-reduced quinone acceptor and the oxidized primary donor.

The decay by recombination of the light-induced charge separation states P'QAQB and P'QAQ; has been largely studied in RCs from purple bacteria. A detailed knowledge of the reaction pathways involved and of the relevant thermodynamic parameters has been reached in both Rb. sphaeroides [19-241 and Rps. viridis RCs [25-291. In a previous study from our laboratory [30] the redox properties of the primary acceptor QA have been determined as they appear in cytoplas- mic membrane fragments from Cfl. aurantiacus. The free energy difference between the P'QAQ; and P+QAQB states has been also evaluated in vivo on the basis of charge recombination kinetics. We report here the temperature dependence of charge recombination from the P'QA and from the P'Q; states in the LM complex isolated from CjZ. aurantiacus. The temperature range investigated has been extended up to 50°C as a result of thermal stability of C'. aurantiucus RCs.

MATERIALS AND METHODS

Purification of the reaction center and preparation ojproteoliposomes

Cells of Cfl. aurantiacus, strain J-10-fl, were grown photoheterotrophically as described in [30]. The method for the preparation of whole-membrane fragments is detailed in [31]. The C'Z. aurantiacus RC was isolated and purified as previously described by Shiozawa et al. [32].

The secondary quinone (QR) activity was reconstituted by adding to the suspension of isolated RCs ubiquinone-3 (UQ- 3) (Sigma) or menaquinone with different hydrocarbon tail length (kindly provided by Hoffmann-La Roche) dissolved in 10% lauryldimethylamine N-oxide (LDAO).

Reconstitution of proteoliposomes was carried out by the following procedure (slightly modified from [33]). Azolectin (60 mg; type IV-S Sigma), supplemented with menaquinone- 8, was dispersed in 300 pl 20 mM sodium phosphate pH 7.5 and 3% sodium cholate and sonicated for 30 s. The suspension was mixed with 300 pl of a solution containing 20 mM Tris pH 8.0,0.2% LDAO and 40 pM isolated RCs. The final mix- ture was sonicated twice for 3 s and applied onto a column (1 x 15 cm) of Sephadex G-50 equilibrated in 20 mM sodium phosphate pH 7.5.

Absorbance measurements and data analysis

The kinetics of the absorbance changes associated with the redox state of the primary electron donor (P) of the RC were monitored with a single-beam spectrophotometer of local design. Measuring light from a Cary 15 monochromator was passed through the sample and onto the photomultiplier (Philips XP2013B), protected by a Corning glass 4/96 filter, by means of optical fibers.

For measurements in the temperature range between 0 C and 70"C, the sample cuvette was held in a water-thermo- statted jacket. The temperature was monitored by a Ptl 00 ceramic resistance (Degussa GR2105) with a tolerance of 0.3 "C, immersed directly into the sample. A liquid nitrogen cryostat (Oxford Instruments Co. Ltd, DN 70) was used for absorbance measurements at low temperatures. Temperature was monitored by a calibrated carbon resistance thermometer.

Actinic flashes were provided by a xenon lamp (3.25 J discharge energy), screened through two layers of Wratten 88A gelatin filter, giving a light pulse of 4 ps duration at half- maximal intensity; 10- 15% of the total RC population was photoactivated by a single flash excitation.

Rapid digitization and averaging of the amplifier output was done by a LeCroy TR8818 transient recorder equipped with a 128K memory module (MM8105) and controlled by an Olivetti M280 personal computer.

Light-induced redox changes of the primary electron do- nor of the RC were monitored at 542 nm or 605 nm [7, lo]. Deconvolution of charge recombinant kinetics into multiple exponential decays was performed by computer routines based on the Marquardt algorithm [34].

RESULTS

Kinetics of recombination of P+ QA

The decay kinetics of the photooxidized primary donor, P', in RCs isolated from Cfl. aurantiucus were studied in aqueous buffer over the temperature range between 280 - 340 K at pH 8.20 and pH 10.00. Over the whole temperature range investigated the decay of P + could be accurately fitted to a single exponential function. At both pH values the rate constant obtained, kAP, was temperature-independent and equal to 17.6 f 0.7 s - l (Fig. 1 a, closed symbols). The same kinetics was observed when the charge recombination reaction following a single turnover flash was monitored at 605 nm and at 542 nm. The monoexponential decay of P' was unaffected by o-phenanthroline (Fig. 1 a, open triangles), in- dicating that RCs isolated from CjZ. aurantiacu,s following the procedure described under Materials and Methods were deprived of the secondary quinone acceptor, QB; and that the observed P+ decay was reflecting the charge recombination between P' and the primary quinone acceptor QA.

627

k 251 a

4

"1 l o [ 5

** t

a * a *

, * a +

0 " ' ' ' ' ~ ' 0 10 20 30 40 50 60 70 80

temperature ( O C)

Fig. 1. Temperature dependence of the kinetics of charge recombination f rom the P'Q; state in RCs of'cfl. aurantiacus. (a) The ratc constant (kAp), determined from monoexponential fi t to the kinetic traces of P + decay, as a function of temperature. (+, V ) Measuremcnts pcrforrned at 542 nrn on isolated RCs at pH 8.0 in the absence and in the presence of 4 mM o-phcnanthroline, respectively. Conditions: 2 pM RCs 0.07% LDAO, 30 mM Tris. (m, A) Data obtained at 542 nm and 605 nm respectively in isolated RCs at pH 10.0. Con- ditions: 2 pM RC, 0.07% LDAO, 40 mM 3-[Cyclohexylamino]-2- propane-sulfonic acid. (0) Membrane fragments in 50 mM Mops, 100 rnM KCI. pH 7.0, in the presence of 4 pM o-phenanthroline. (b) The cxtent of P+ induced in isolated RCs by a single turnover flash vs temperature at pH 8.0 (+ )and pH 10.0 (W). Conditions as in (a). Between 10 - 50 'C the RC was stable for hours, enduring consecutive heating and cooling cycles. The data at > 50'C have been taken during an heating cycle. The sample has been exposed to the indicated temperature for approximately 15 min

As shown in Fig. 1 b, heating of the RC during measurements up to approximately 50°C had no effect on the extent of the primary donor P oxidized by the flash. Above this temperature the photochemical activity of the preparation decreased monotonically. At pH I0 an almost complete thermal inactivation of primary photochemistry appeared to take place at 70"C, while at pH 8 a somewhat higher temperature was required in order to inactivate the RC fully. This pattern of thermal inactivation agrees with heat denaturation curves of the RC obtained from the absorbance decrease of the 81 5- nm and 865-nm bands [12, 131.

The kinetics of charge recombination of the P+Q, state in isolated RCs was in good agreement with the results of analogous measurements performed in whole membranes of Cfl. uurantiacus. In fact, when the QA to QR electron transfer was inhibited by o-phenanthroline, a monoexponential decay of P + was observed in membrane fragments and essentially thc same, temperature-independent, rate constant value measured in isolated RCs was obtained (Fig. l a, open diamonds). In these measurements a less accurate determination of kAP could be achieved due to the strong background optical ab- sorption of membrane suspensions.

The temperature independence of khP between 280 K and 340 K indicates that the recombination of P'QA in RCs from cf!. a ~ r ~ n t i u c u s proceeds via an activationless electron-trans- f'er mechanism. In order to substantiate this conclusion

F

1 2 - 40 50r

rn a a rn

a a

I

1 1 T i 2 4 6 8 10 1 2 I4

10OO/T (K-1)

Fig. 2. Extended temperature dependence of the rate constant k,,, 7.5 pM isolated RCs in 50 mM Tris pH 8.0 and 0.04% LDAO were diluted to 60% glycerol

further, the kinetics of P'QA recombination were studied at lower temperatures in isolated RCs suspended in 60% glycerol, to allow freezing of the sample down to 80 K (Fig. 2). At pH 8.0 the recombination rate constant was 16.4 1.5 s-' at 293 K. Between 250 K and 160 K, kAP increased to approximately 33 s-'. The rate constant remained at this value as the temperature was lowered further from 160 K to 80 K . The extended temperature dependence observed in Cfl. aurantiacus RCs appears to be qualitatively similar to that reported for the kinetics of P'QA recombination in RCs from Rh. sphaeroides [19-21, 24, 25, 351 and is consistent with the notion that the back-reaction occurs via tunnelling through a potential barrier.

P ' QF charge recomhinution kinetics

As indicated by the results outlined in the previous paragraph, RCs isolated from Cfl. aurantiacus retained the primary menaquinone acceptor QA but were devoid of the secondary quinone acceptor QR. Ubiquinone (UQ-3) and menaquinones of different hydrocarbon tail length were used in restoring the QB function. In these reconstituted RCs the kinetics of P'QB recombination has been studied over a wide temperature range.

u) Reaction centers reconstituted with uhiquinone

The addition of UQ-3 in LDAO solution to a suspension of RCs almost completely restored the QH function as judged from the kinetic analysis of charge recombination (Fig. 3). Over the temperature range examined (from 5 ('C to 60 'C) the decay could be reasonably described by a monoexponential function. As a consequence of ubiquinoiie addition the half- time of Pf decay appeared to increase dramatically from approximately 40 ms (the half-time measured for the P+Q/\ recombination) to 3-4 s. In order to resolve possible kinetic components in the millisecond time domain, the recovery of P+ was routinely recorded also at a higher sampling frequency and deconvoluted into a slow and a fast exponential decay. The amplitude of the fast component (characterized by a half-time of approximately 40 ms, indicative of the charge recombination between P+ and QA) was never found to exceed 5% of the total amplitude, demonstrating an essentially full reconstitution of the secondary acceptor activity. This rcsult is in line with the observation that in ubiquinone-reconstituted RCs isolated from Cfr. auruntiucus the formation of ubisemiquinone undergoes binary oscillations following a

628

8

r) 6 - 0

x 4 - Q

7

a 2 -

0

1 r-'

c * *

* * ..

0

1 1 I

0

7 -2 I? 0

X

8 -4 (D

," -6

-8

1 0 ' 1 0 2 4 6 8 1 0 1 2

time ( s )

Fig. 3. The kinetics of charge recombination following a single turnover .flash in ubiquinone-3-reconstituted RCs isolated from Cfl. aurantiacus, recorded at 12.6"C. Conditions: 2 pM RCs, 0.07% LDAO, 20 mM Tris pH 8.0. Ubiquinone-3 dissolved in LDAO was added to the sample to a final concentration of 40 pM. The smooth curve is a fit to a single exponential decay with an half-time equal to 3.78 s. The upper plot shows the difference between the fitted function and the experimental data

0.1 3.0 3.2 3.4 3 .6

1000/T (K-1)

Fig. 4. The temperature dependence of the rate constant of charge recombinationfrom the Pt QB state in ubiquinone-3-reconstituted RCs. Conditions as in Fig. 3. Kinetic traces have been recorded at 605 nm ( W ) and at 542 nm (+)

series of flashes [36]. The rate constant for the P'Qi charge recombination, kgp, was found to be temperature independent between 5°C and 40°C and equal to 0.18 0.03 s - l (Fig. 4). The charge recombination between P' and Q< appears therefore to occur directly without involving a thermally activated intermediate state. The alternative possibility of an apparent temperature independence due to an opposing temperature dependence of QB binding and of the PfQA/PtQe equilibrium, discussed by McComb et al. [37], seems unlikely here due to the large temperature range examined. When measurements were performed at temperatures higher than 50°C the decay of P+ was progressively accelerated; the apparent rate constant for PtQ" charge recombination increased by approximately one order-of-magnitude over a 10 "C interval, reaching a value of 2 sC1 at 60°C (Fig. 4). The interpretation of this behaviour is hindered by the fact that exposure of the RC to temperatures higher than 50 "C can cause a partial thermal denaturation of the QB site, being this temperature is close to the temperature range (6O-7O0C, Fig. 1 b) over which photoactivity is thermally inactivated.

M O 0 c

x -1 LD g -2

a -3

-4

Q

I

0 1 2 3 4

time ( s )

b ) Reaction centers recomtituted with menaquinone

The main physiologically active menaquinones in membranes of Cfz. aurantiacus are menaquinone-8 and menaquinone-10 [I 13. In our reconstitution studies with detergent- solubilized reaction centers shorter-chain menaquinones (menaquinone-3, -4 and -6) were used, since their solubility properties allowed a successful reconstitution.

Fig. 5 shows two examples of the recombination decay of Pt recorded at 5°C (Fig. 5a) and at 43°C (Fig. 5b) in RCs reconstituted with menaquinone-4. At both temperatures the recovery kinetics is markedly biphasic. Traces are easily deconvoluted in two exponential components, the half-times of which differ by more than one order-of-magnitude. At temperatures close to 0°C the slow phase (t1,2 > 9 s) contributes approximately 15% ofthe total signal. The relative amplitude of the slow component increases at increasing temperatures (see Fig. 5b) and exceeds 80% of the total at 50°C (see

629

A A

A A

b .

1

500 c

+.

'c 0

x v 40 6o t

A A "1 A A A

0 0 20 40 60

temperature ("C)

Fig. 6. Temperature dependence o f the kinetics of charge recombination in RCs reconstituted with menaquinone-3 (A) and with menaquinone- 4 (A). Conditions as in Fig. 5. The recovery kinetics of P+ following a single turnover flash was deconvoluted into a fast and a slow exponential decay (see Fig. 5, a and b). (a) Half-time of the fast component. (b) Amplitude of the slow phase

Fig. 6b, closed symbols). The half-time of the fast phase, approximately 350 ms at 5 "C, decreases monotonically when the temperature of the sample increases, being close to 100 ms at 50°C (Fig. 6a). This effect of temperature was fully rever- sible. The reconstitution of the function of the secondary acceptor is essentially complete since no kinetic phase fast enough to be attributable to recombination of the P 'QA state could be detected.

The markedly biphasic decay observed at 605 nm reflects a genuine kinetic feature of P f recovery, as demonstrated by the coincidence in shape of difference spectra taken immedi- ately after the flash and more than 1 s after the photoactivation (when the fast phase has already decayed) (Fig. 5c). The same temperature dependence of P+ recovery was observed at 542 nm (data not shown).

The results of the kinetic analysis illustrated in Fig. 5a, b are summarized in Fig. 6 for RCs reconstituted with menaquinone-4 (closed symbols) and menaquinone-3 (open symbols). In both cases the half-time of the fast phase shows quite a good correlation with temperature. Its value at 20°C is comparable to the half-time measured for the P'QB recombination in intact membranes [30]. The fast phase of P+ recovery is therefore consistent with the expected charge recombination of P+ with the secondary acceptor Q e , occurring through a thermally activated state.

The simplest explanation for the presence of a much slower phase in the recovery of Pf is that a reaction which removes reducing equivalents from the quinone acceptor complex can compete with charge recombination between P i and Q i . In fact the decay of P by a slow process can only occur when the radical pair P'QR (or P'QA) has somehow lost its reactivity. This is most likely to arise if Q; or QA are rapidly oxidized by the excess quinone present in the sample, leaving the primary donor in its oxidized form and unable to back- react.

0 I . , , I I , ' I , -1 '

I

- I 2 -1 4 t.1 0 200 400 600 800 1000

time (ms)

0

0 -2- x - 4 - LO -6 -

-8

7

0 (0

Q a -10-

-12 - - 1 4 ' I

0 200 400 600 800 1000

time (ms)

Fig. 7. Menaquinone-6 reconstitution: kinetic traces of charge recomhi- nation f rom the P'QB state recorded at 8.8' C ( a ) and 44.3 C (b ) . Conditions: 2 FM RCs, 0.07% LDAO, 20 mM Tris pH 8.0. Menaquinone-6 dissolved in LDAO was added to the sample to a final concentration of 40 pM. The signals have been fitted to the sum of three exponential decays. The calculated half-times (tip) and amplitudes ( A ) are given in the following. (a) Phase 1, A = 10.2%, t1p

= 35 ms; phase 2, A = 86%, t l p = 334 ms; phase 3, A = 3.8%, t l p = 9 s. (b) Phase 1, A = 21.4%, t I l2 = 31 ms; phase 2, A = 54%, t l 12 = 140 ms; phase 3, A = 24.6%, = 5 s. Residuals are shown in the upper plot corresponding to each trace

Fig. 7 shows the time course of Pf decay in menaquinone- 6-reconstituted RCs at 8.8"C (Fig. 7a) and at 44.3"C (Fig. 7 b). Over this temperature range single-exponential fits to kinetic data were unacceptable as judged by the residuals. Similarly bi-exponential fits were not satisfactory, particularly at temperatures over 20 "C, and yielded time constants poorly correlated with temperature. An acceptable treatment was obtained by using as a fitting function the sum of three exponential decays (see plot of residuals in Fig. 7a and b). For the sake of clarity, the three kinetic phases will be designated here phases 1, 2 and 3, corresponding to increasing order-of- magnitude of their half-time.

As is apparent from Fig. 8a, the half-time of phase 1 is temperature-independent and fluctuates in a range of some tens of milliseconds around an average of 40 ms. This value is consistent with the temperature-independent rate constant measured for the charge recombination between P+ and QA under a variety of conditions (Fig. 1 a). The relative amplitude

630

500

400 E - 300 t

200

I 100

0

0 80

60

40

a 20 E Q n

._

'c U

c- 0 4

4 Lc

x v

3 r .- -

I.'

a

0 20 40 60

temperature ("C) Fig. 8. Tempwu ture dependence ofthe half- tinie ( a ) and of the relative umplitude ( h ) qf'the kinetic phases present in the recombination decuj of P+ in RCs reconstituted with menuquinone-6. Data have been obtained by deconvoluting kinetic traces into three exponential components as shown in Fig. 7. (A, 7 ) Phase 1; (M, +) phase 2; (1, 0) phase 3. Thc different symbols for each kinetic component correspond to data obtaincd by analyzing traces recorded at 542 nm and 605 nm, respectively

of phase 1 is close to 10% between 5°C and 25°C and tends to increase at higher temperatures (Fig. 8b). This result im- plies that the addition of menaquinone-6 in LDAO to isolated RCs reconstitutes the function of the secondary acceptor in approximately 90% of RCs. At 50°C this fraction decreases to approximately 75% of the total RC population.

Phase 2 dominates the recovery kinetics of P f (Fig. 8 b). Its half-time (Fig. 8 a) is temperature-dependent, decreasing from approximately 350 ms at 5°C to 100 ms at 50°C. Similar values and an analogous temperature dependence has been obtained for the half-time of the fast phase of Pf decay in RCs reconstituted with menaquinone-3 or menaquinone-4. I t is therefore proposed that phase 2 reflects mainly the recombination between P+ and QR. The decrease of the relative amplitude of phase 2 which takes place at high temperatures (Fig. 8 b) is paralleled by an increase of the contribution to the decay of phase 1 and phase 3. The half-times obtained for phase 3 were always larger than those of phase 2 by at least one order-of-magnitude. Therefore, over the time scale of Fig. 7, phase 3 can be considered essentially constant in time. The presence of a slow phase (decaying in tens of seconds) suggests that, also in menaquinone-6-reconstituted RCs, QG and/or QA can be oxidized by exogenous menaquinone in a reaction which occurs in parallel to the recombination of the P'Q; (or P'QA) couple. In the case of menaquinone-6, however, as compared to menaquinone-4 or menaquinone-3, this reaction is less effective in competing with charge recombination within the RC and the temperature dependence of its contribution to the overall decay of P + is much less pronounced.

The proposed interpretation of the slow kinetic phase occurring at high temperatures in menaquinone-reconstituted RCs is corroborated by the analysis of P' decay in the presence of o-phenanthroline. When this inhibitor was added at a

4 mM concentration to RCs reconstituted with menaquinone- 3 or menaquinone-4 the recovery of P+ following a single turnover flash was still bi-exponential. The slow kinetic component was unaffected by the inhibitor, except for a slight increase of its relative amplitude, expecially at high temperature. The fast phase observed in the presence of o- phenanthroline decayed in some tens of milliseconds, as expected for charge recombination of Pf with QA, whenelectron transfer to the secondary acceptor QB is inhibited (data not shown). The effect of o-phenanthroline in RCs reconstituted with menaquinone-6 was qualitatively similar. Under these conditions, however, the relative contribution of the slow phase (phase 3) increased dramatically upon addition of o- phenanthroline: at 20"C, in the presence of the inhibitor, the slow phase accounted for approximately 50% of the overall decay and, at 40' C, its relative amplitude exceeded 90% of the total signal (M. Trotta, R. Feick and G. Venturoli, unpublished results). These results are consistent with the idea that in RCs isolated from Cjl. aurantiacus the reduced quinone acceptor complex is exposed to a relatively rapid oxidation by menaquinone in solution. Since o-phenanthroline, which is supposed to bind at the QB site, considerably reduces the relative amount of P + undergoing a normal charge recombination, it seems likely that QA is able to react directly with the oxidized exogenous menaquinone.

Reconstitution in proteoliposomes

The peculiar behaviour observed in isolated RCs when menaquinone is present in solution poses the question of the integrity of the QR site and of the proper functioning of the quinone acceptor complex in an RC which lacks subunit H. These points have been further investigated by studying the kinetics of charge recombination in RCs reincorporated into phospholipid vesicles. In this case menaquinone-8, one of the menaquinones normally present in intact membranes, was used. The lipid environment almost completely restored the kinetics of charge recombination observed in intact membranes, preventing redox side reactions of the menaquinone acceptor complex (phase 3). In the presence of menaquinone- 8 dissolved in the lipid phase, in a ratio of 100 menaquinone molecules/RC, the Qs function appeared to be almost completely reconstituted: the decay of P + following a single flash did not differ markedly from a monoexponential recovery with an half-time consistent with charge recombination between P f and QR. In order to compare these results with those obtained in detergent suspensions the same kinetic analysis, i.e. deconvolution into three components, was applied. Since in proteoliposomes the amplitude of phase 1 and phase 3 (the faster and the slower one) was always very limited, the number of degrees of freedom in the fitting procedure was reduced: in a first approximation phase 3 was assumed to be constant on the time scale of the recorded decay of P' (1.6 s) and the half-time of phase 1 was kept fixed to 39 ms, i.e. to the characteristic time for charge recombination of the P'Q; state. The result of the analysis is presented in Fig. 9. At variance with the data obtainedin menaquinone-6-reconstituted RCs in detergent solution (Fig. 8), over the whole temperature range examined (between 5°C and S O T ) the relative contribution of phase 1 plus phase 3 never exceeded 15% of the total signal (Fig. 9b). The half-time of phase 2 (Fig. 9a), which accounts for most of the P+ decay, is consistent with a thermally activated charge recombination of the P+Q; state. In line with this kinetic behaviour and quite importantly, upon addition of o-phenanthroline, the decay of

63 1

m E

f 400

v

.-

0

0 -

0

a

a

0 10 20 30 40 50 temperature ("C)

Fig. 9. Temperature dependence of the kinetics of P+ decay in RCs reincorporated in phospholipid vesicles. QB was reconstituted with menaquinone-8. Proteoliposomes equivalent to 0.42 pM RC were suspended in 20 mM sodium phosphate pH 7.5. The recovery of P+ following a flash has been deconvoluted in two exponential decays (phase 1 and phase 2) plus a constant term (phase 3). (a) Half-time of phase 2. (b) Relative amplitudes of phase 1 (A), phase 2 (0 ) and phase 3 (+)

P + was monoexponential with a half-time of approximately 40 ms (data not shown). No unusual effect of the inhibitor was detected even a t 50°C. The results obtained in proteoliposomes indicate, therefore, that, when inserted into a proper lipid environment, the isolated RC undergoes a physiological charge recombination reaction and that the anomalous kinetic behaviour observed in detergent suspensions is not due to an irreversible damage of the QR site but reflects an intrinsic property of the RC isolated with LDAO from C'. aurantincus. Whether this behaviour is related to the lack of subunit H is unknown, since analogous studies of the LM complex from Rh. sphaeroides incorporated into phospholipid vesicles have not yet been performed.

DISCUSSION

Temperature dependence of the kinetics of charge recombination from the P'Q; state

The overall temperature dependence of the kinetics of recombination of P'Q; in Chloroflexus aurantiacus RCs is analogous to that observed for Rh. sphaeroides RCs, either with the native UQ-10 [I9 - 21, 24, 351 or with menaquinone- 8 acting as QA [25]. Taking advantage of the thermal stability of CJ. aurnntiucus RCs, measurements have been extended to temperatures as high as 3 10 K, without detecting any change in the rate constant as compared to 280 K. Between 280 K and 340 K charge recombination in Cfl. aurantiacus RCs appears to proceed directly and does not involve any thermally activated process. This behaviour is in contrast to the temperature-dependent P + QA recombination observed in native Rps. viridis RCs [25] and also in Rb. sphaeroides RCs when anthra- quinone serves as QA [24, 351. The back reaction in these two systems has been proposed to proceed through two parallel

mechanisms : an activationless route (dominant at low temperatures) and a thermally activated pathway through an energetically elevated intermediate state P t I - (which becomes relevant at high temperatures). According to this scheme the occurrence of a thermally activated charge recombination at high temperature will depend primarily on the energy gap between the P t I-QA and P ' IQA states [24, 25, 351.

The rate of P'QA recombination in Cfl. nurantiacus RCs increases by a factor of about two between 280 K and approximately 150 K, and then remains constant down to 80 K. This peculiar temperature dependence, characterized by a small apparent negative activation energy, is qualitatively similar to the behaviour of Rb. sphaeroides RCs either when ubiquinone [19-21, 241 or different quinones are present at the QA site [24, 251. Interestingly the absolute value of the rate constant at 300 K and its relative increase between 300 K and 150 K are very close in CfZ. aurantiacus and in Rh. sphaeroides RCs when the native UQ is replaced by menaquinone-8 [25]. The onset of the inverted temperature dependence has received different interpretations relying on nonadiabatic multiphonon theories of intramolecular electron transfer (for a review see [38]). Both the electronic (tunnelling) matrix element [20, 21, 24,39,40] and the thermally averaged Franck-Codon nuclear overlap factors [41-441 have been proposed and are likely to contribute to the observed inverse temperature dependence [451.

Kinetics of charge recombination f rom the Pt Qi state

The temperature dependence of the P 'Qs recombjnation kinetics observed in Cjl. aurantiacus RCs will be discussed within the context of the following kinetic scheme, Scheme 1 [46, 471:

kAP

P'QAQG -P+QQAQR-PQAQR I t

k w

Two possible competing pathways for recombination are considered : a direct, activationless route and a parallel, indirect route, which proceeds via a thermally activated intermediate state. The involvement of an energetically elevated intermediate state in the charge recombination process of P'QAQ; has been firmly established in Rb. sphaeroides RCs; at ambient temperatures this back reaction has indeed been shown to proceed via the PtQQAQH state, the direct pathway contribu- ting less than 5'/0 [23]. Evidence in favour of an indirect recombination pathway via P'Q;QB has been also obtained in Rps. viridis RCs [26, 291.

According to Scheme 1 , assuming quasi-equilibrium between PtQAQR and P'QiQg during charge recombination, the relative populations of these two states are given by the Boltzmann factor

(1)

where AG is the free energy difference between P'Q/\QR and P'QAQi, kJ3 Boltzmann's constant and T absolute temperature. The observed rate constant ( k ) for Pf decay is the sum of the contributions from the direct and from the indirect routes, i.e.

[P'QA QRJ/[P'QAQ~I= exp( - AG/kgT)

where the rate constants k,, and kgp are defined in Scheme 1.

632

In order to analyze the temperature dependence of k Eqn (2) can usefully be written as

k = (kAP-k)exp(-AG/kBT)+ kBp. (3)

a ) P+ Q i recombination in UQ-reconstituted reaction centers

Our results clearly show that UQ-3 can function as a secondary quinone acceptor in Cfl. aurantiucus RCs. Further independent evidence for stable ubisemiquinone formation at the QB site is provided by measuring binary semiquinone oscillations in a train of flashes [36].

The back-reaction observed in UQ-3-reconstituted RCs at ambient temperature is approximately one order-of-magnitude slower as compared to the physiological charge recombination between P + and the secondary (native) menaquinone acceptor [30]. Therateconstant (k = 0.18 & 0.03 s-') is essentially temperature-independent between 5 - 40°C. If the conclusion is correct that no thermally activated process is involved, at least up to 35"C, a lower limit for the free energy difference AG between PtQAQB and P'QAQB can be calculated from Eqn (2). In order that, at 35"C, the contribution of the indirect recombination pathway to the rate constant does not exceed lo%, AG must be > 180 meV. This value seems reasonable when compared to the much smaller P'QLQB-P'QAQi energy gap evaluated in vivo [30] and in menaquinone-reconstituted RCs (see below) and is consistent with the more positive redox midpoint potential of UQ-3 as compared to menaquinone [48].

At temperatures higher than 40°C, the rate constant increases rather steeply, suggesting that a thermally activated process is starting to compete with the direct recombination route. An Arrhenius plot of (k-kBp)/(kAp-k) (see Eqn 3) displays some deviation from linearity at the highest temperatures (not shown) and gives an apparent activation enthalpy larger than 800 meV. It should be noticed, however, that at temperatures above 60°C partial denaturation of the RC already takes place as juged from the photoactivity (see Fig. 1 b). This process could be preceded by a derangement of the QB site causing a perturbation of the recombination kinetics and therefore preventing a reliable estimation of the activation parameters for a putative indirect recombination pathway active at very high temperatures.

h ) P+ Qi recombination in menaquinone-reconstituted reaction centers

The recovery of P' in detergent suspensions of menaquinone-reconstituted RCs displays a complex multiphasic kinetics which depends markedly upon temperature between 280-320 K. A slow phase (phase 3), decaying in tens of seconds, contributes to P' disappearance to a different extent in RCs in which either a short- or long-chain menaquinone acts as secondary acceptor. In Rb. sphaeroides RCs a biphasic P+ relaxation analogous to the one detected in menaquinone- 3- and menaquinone-4-reconstituted RCs has been observed upon dissociation of the H subunit [18]. In this system, however, the faster component was attributed to recombination between P' and QA, whilst in Cfl. aurantiacus RCs the faster phase is consistent with back-reaction from the P'QAQ; state. The decay of a fraction of P' in the time scale of tens of seconds has been taken by us to indicate that a reaction removing reducing equivalents from the quinone acceptor complex is competing with charge recombination of P'Qn. It is suggested that Qe and/or QA can be rapidly oxidized by

the excess quinone in the suspension thus depriving P' of its recombination partner. This interpretation is consistent with : (a) the spectral homogeneity of the fast and slow components of P+ decay; (b) the strong temperature dependence of the amplitude of the slow phase which dominates at high temperatures; (c) the anomalous effect of o-phenanthroline, which affects the contribution of the slow phase to the decay of P'; (d) the disappearance of the slow component of Pf recovery when RCs are reincorporated into phospholipid vesicles (see below). Indeed, a relatively fast oxidation of the semiquinone anions QA and Q i has been directly observed in the LM complex isolated from Rb. sphaeroides [18]. Moreover, in Rps. viridis RCs, QA is oxidized by ferricyanide in a reaction which can compete with charge recombination of P'QA [49].

According to the proposed interpretation, the decay of P' after a flash of light can be described by the following scheme, Scheme 2:

k

PQAQA + P'QAQ~ PfQAQ6 + e- in which the intermediate P'QAQB state has not been considered explicitly, assuming quasi-equilibrium between the P'QAQe and the P'QiQB states during charge recombination. As a consequence, the competing 'side' reaction has been taken to oxidize the quinone acceptor complex in the QA Qe state, although oxidation of the QAQB state is equally probable. In the absence of more detailed information this process is represented as the irreversible removal of an electron from the quinone complex (pseudo-first-order rate constant kJ. The slow relaxation of the P'QAQB state is not included in Scheme 2, since its rate is negligible on the time scale of k and k,.

With the initial conditions [PQAQB] = [P+QAQB] =. 0, [PCQAQi] = [P'l0, Scheme 2 gives for the time ( t ) evolution [P'], of the oxidized primary acceptor:

By equating k,/(k + k,) to the relative amplitude of the slow component and (k + k,) to the apparent rate constant of the fast phase of P' decay, the rate constant for charge recombination (k) and for the 'side' oxidation reaction (k,) can be obtained from the observed kinetics of P' relaxation.

This procedure has been applied to the data of Fig. 6a, b (menaquinone-3- and menaquinone-4-reconstituted RCs) and to the kinetics of P' recovery in menaquinone-6-reconstituted RCs (Fig. 8a, b). In RCs supplemented with the long-chain menaquinone, a third phase (named phase 1) was present, due to incomplete reconstitution of the QB function. Phase 2 and phase 3 were therefore considered in applying Eqn (4) to this set of data.

The temperature dependence of k for menaquinone-6-reconstituted RCs is presented in Fig. 10. In view of the behaviour of Rh. sphaeroides [23] and Rps. viridis [26] RCs, the obvious candidate for the thermally activated intermediate state is P+QAQB. The enthalpy (AH) and entropy ( A S ) difference between the two states P'QAQB and P'QAQ; has been therefore estimated on the basis of Eqn (3) with kgp = 0, by fitting the data to

The activation parameters for RCs reconstituted with different menaquinones, obtained from such calculations, are

633

-0.5

-1.0

' -1.5

- - Y

5 -2.0 u c -

-2.5

-3.0 3.0 3.2 3.4 3.6

I / T ~ lo3 (K-') Fig. 10. Arrhenzusplot for the rate constant k of charge recombination from the P+ QAQ; state in menaquinone-6-reconstituted RCs. k values have been obtained from the kinetics of P+ decay (see Fig. 8) as described in the text. The derived thermodynamic parameters ( A H , A S ) are given in Table 1

Table 1. Enthalpy, entropy and free energy values .for the electron iran.sfer equilibrium reaction P'Q; QB* P+ QAQB at p H 8.0 in Cfl. aurantiacus RCs with dcferent menaquinone isoprene homolopes acting as QB

meV meV K - ' meV Menaquinone-6 - 147 -0.34 - 47 Menaquinone-4 - 59 -0.02 - 53 Menaquinone-3 - 52 + 0.01 - 54

summarized in Table 1. When menaquinone-6 acts as a secondary acceptor, the P+QAQB --f P'QAQS reaction is characterized by a negative enthalpy change, partially offset by a substantial negative entropy change. Qualitatively similar enthalpy and entropy contributions have been found for this reaction in Rb. sphaeroides [22] and in Rps. viridis [29]. The free energy change at 18°C associated with the transfer of an electron from QA to QB is close to -50 meV, in good agreement with the AG value obtained in intact membranes [30]. Upon shortening of the menaquinone tail, both the enthalpic and the entropic components decrease in absolute value (- TAS becomes negative in menaquinone-3-reconstituted RCs). These concomitant changes essentially compensate each other, resulting in an almost constant net affinity of the reaction at ambient temperature.

The analysis of the temperature dependence of the 'side' reaction which oxidizes the quinone acceptor competing with charge recombination has been carried out within the context of the activated complex theory for reaction rates [50].

The values for the activation enthalpy (AH*) and entropy (AS*) derived from the temperature dependence of k, for different menaquinone prenylogues are summarized in Tab- le 2. For menaquinone-6 the reaction has a positive activation enthalpy and a large negative activation entropy. When the size of the isoprenyl side-chain is reduced, the activation enthalpy increases and the entropic contributions ( - TdS*) decreases undergoing a substantial change and becomes negative in nienaquinone-3-reconstituted RCs. Again the free energy of activation at ambient temperature is remarkably insensitive to prenyl chain length. Analogous effects of strong entropy- enthalpy compensation have been recently reported for the binding of ubiquinones of different tail length at the QA site of Rb. spharroidrs [37]. The suggestion that the counterbalance

Table 2. Activation parameters for the side reaction oxidizing the quinone acceptor complex in detergent suspensions of Cfl. aurantiacus RCs with menaquinones ojdiferent tail length

Menaquinone-n AH* AS* AG?,,

kJmol- ' Jmol-I K - ' kJrno1-l

Menaquinone-6 45 - 100 74

Menaquinone-3 81 + 27 74 Menaquinone-4 48 - 82 72

of enthalpic and entropic contributions is due to an antago- nistic interaction between the headgroup and the side chain seems quite reasonable also in relation to our data.

c) Behaviour ojproteoliposomes As pointed out in the previous paragraph, the kinetic

analysis of P' decay in detergent suspensions of menaquinone-reconstituted Chloroflexus RCs indicates that the reduced quinone acceptor is somewhat unstable against oxidation on the time scale of charge recombination. Semiquinone instability is especially evident in the presence of short-chain menaquinones at high temperatures. This behaviour, which resembles that observed in Rb. sphaeroides RCs deprived of the H subunit [18], is not due to an irreversible damage of the quinone binding sites. In fact, reincorporation of the RCs into phospholipid vesicles completely restores the kinetics of charge recombination between Q; and P' observed in intact membranes. The temperature dependence of the rate constant, analyzed according to Eqn (9, yields AH = -72 meV and A S = -0.05 meV K- ' resulting in AGZg1 = - 57 meV, a value close to the free energy difference obtained in detergent suspensions of menaquinone-reconstituted RCs (cf. Table 1). Remarkably the essentially monophasic P+ decay detected in proteoliposomes is sensitive to o- phenanthroline at the concentrations normally effective in intact membranes. It appears therefore that reinsertion of the RC into a lipid bilayer prevents electrons from leaking off QA and QG, possibly reducing the accessibility of excess menaquinone dissolved in the lipids to the QA and QB sites. This result is quite relevant when the function of the quinone acceptor complex in the two-subunit RC isolated from Cfl. aurantiacus is compared to the electron transfer properties of the LM complex from Rb. sphaeroides.

Dissociation of the H subunit from the RC of Rh. sphaeroides dramatically reduces the QA to Qe electron transfer rate as well as the affinity of UQ-10 for the QB site and abolishes the oscillatory behaviour of semiquinone formation [18]. Although the Cfl. aurantiacus RC in detergent suspensions shares with the Rb. sphaeroides LM complex some instability to oxidation of the semiquinone anions Q; and QB, the P'QA QB - P+QAQB electron transfer equilibrium does not seem to be affected in isolated Cfl. auvantiacus RCs as compared to intact membranes. When the side reaction removing electrons from the reduced acceptor complex is taken into account, the rate constant for charge recombination from the P.'QAQ, state is close to that observed in vivo. Moreover the thermodynamic parameters derived for the P'QA QH - P'QAQK equilibrium in detergent suspensions of RCs give essentially the same AG, independent of the isoprene homol- ogue a t the Qu site and therefore of the extent of semiquinone instability. Finally, the physiological behaviour observed in proteoliposomes suggests that in Cfl. aurantiacm a third sub-

634

unit is not required for proper functioning of QB when the RC is embedded into a lipid bilayer. In line with this conclusion, a rapid electrogenic phase associated with QB protonation has been recently observed upon photoactivation of Cfl. auruntiacus RCs incorporated into proteoliposomes [51], an observation confirming the functional integrity of the QB site.

G.V. wishes to thank Dr Armeii Mulkidjanian, Dr Alexey Se- mcnov, Dr Michail Verkhovsky and Prof. Vladimir Skulachev of the Moscow State University for many stimulating discussions and access to unpublished material. The authors are grateful to Prof. Giorgio Forti (University of Milano) for helpful advice and assistance with the low-temperature measurements. Hoffnmnn-La Roche (Basel, Switzerland) is gratefully acknowledged for providing mcnaquinone samples. This work was supported by Minister0 dellu Pubblica lstruzione and by Consiglio Nazionale delle Ricerche (Progetto Finul- izzuto Biorwnologic~ e Bioinstrumentazione) of Italy.

RE t; E REN CES 1. Deisenhofer, J., Epp, O., Miki, K., Huber, R . & Michel, H. (1985)

2. Michel, H., Epp, 0. & Deisenhofer, J . (1986) EMBO J . 5, 2445 -

3. Deisenhofer, J . & Michel, H. (1989) EMBO J . 8, 2149-2170. 4. Allen, J. P., Fehcr, G., Yeates, T. O., Komiya, H. & Rces, D. C.

5. Allen, J. P., Feher, G., Yeates, T. O., Komiya, H. & Rees, D. C.

6. Feher, G., Allen, J . P., Okamura, M. Y. & Rees, D. C. (1989)

7. Bruce, B. D., Fuller, R. C. & Blankenship, R. E. (1982) Proc.

8. Kirmaier, C., Holten, D., Feick, R. & Blankenship, R. (1983)

9. Kirmaier, C., Blankenship, R. & Holten, D. (1986) Biochim. Bio-

10. Vasmel, H. &Amesz, J. (1983) Biochim. Biophys. Actu 724,118-

11. Hale, M. B., Blankenship, R. & Fuller, R. C. (1983) Biochim.

12. Pierson. B. K., Thornber, J. P. & Seftor, R. E. B. (1983) Biochim. Biophys. Acta 723, 322- 326.

13. Kutuzov, M. A., Levina, N. B., Abdulaev, N. G. & Zolotarcv, A. S. (1990) in Molecular biology of’ membrane-bound complexes in photosynthetic bacteria (Drews, G. & Dawes, E. A,, eds) pp. 299 - 303, Plenum Press, New York.

14. Pierson, B. K. & Thornber, J. P. (1983) Proc. Nut1 Acad. Sci. USA

15. Ovchinnikov, Yu. A.. Abdulaev, N. G., Zolotarev, A. S., Shmukler, B. E., Zargarov, A. A., Kutuzov, M. A., Telezhinskaya, I . N. & Levina, N. B. (1988) FEBS Lett. 231, 237 - 242.

16. Ovchinnikov, Yu. A,, Abdulaev, N. G., Shmukler, B. E. & Zolotarev, A. S. (1988) FEBS Lett. 232, 364-368.

17. Shiozawa, J. A,, Lottspeich, F., Oesterhelt, D. & Feick, R. (1989)

18. Debus, R. J., Feher, G. & Okamura, M. Y. (1985) Biochemistry

19. Clayton, R. K. & Yau, H. F. (1972) Biophys. J . 12, 867-881. 20. Hsi, E. S. P. & Bolton, J. R. (1974) Biochim. Biophys. Acta 347,

Nature 318, 61 8-624.

2451.

(1987) Proc. Nutl Acad. Sci. USA 84, 5730-5734.

(1987) Proc. Natl Acad. Sci. U S A 84, 6162-6166.

Nature 339, 1 11 - 1 16.

Nutl Acad. Sci. USA 7Y, 6532-6536.

FEBS Lett. 158, 73 - 78.

phys. Acta 850,275-285.

122.

Biophys. Acta 723, 376-382.

80, 80 - 84.

ELI!. J . Biochem. 180, 75 - 84.

24,2488 -2500.

1 26 - 133.

21. McElroy, J . D., Mauzerall, D. C. & Feher, G. (1974) Biochim.

22. Mancino, L. J., Dean, D. P. & Blankenship, R. E. (1984) Biochim.

23. Kleinfeld, D., Okamura, M. Y. & Feher, G. (1984) Biochim.

24. Gunner, M. R., Robertson, D. E. &Dutton, P. L. (1985) J . Chem.

25. Shopes, R. J. & Wraight, C. A. (1987) Biochim. Biophys. Acta

26. Wraight, C. A., Shopes, R. J. & McComb, J. C. (1987) in Progress in photosynthesis research (Biggins, J., ed.) vol. 11, pp. 387- 396, Martinus Nijoff, Dordrecht.

27. Sebban, P. & Wraight, C. A. (1989) Biochim. Biophys. Acta 974,

28. Baciou, L., Rivas, E. & Sebban, P. (1990) Biochemistry 29,2966-

29. Gao, J., Shopes, R. J. & Wraight, C. A. (1991) Biochim. Biophys.

30. Vcnturoli, G. & Zannoni, D. (1988) Eur. J . Biochem. 178, 503-

31. Zannoni, D. & Ingledew, W. J . (1985) FEBS Lett. 193, 93-98. 32. Shiozawa, J . A,, Lottspeich, F. & Feick, R. (1987) Eur. J . Biochem.

33. Kutuzov, M. A. (1990) Thesis, Moscow State University. 34. Bevington, P. R. (1969) Data reduction and error unuljisis,for the

physical sciences, pp. 204-246, McGraw-Hill, New York. 35. Gopher, A., Rlatt, Y., Schonfeld, M., Okamura, M. Y., Feher,

G. & Montal, M. (1985) Biophys. J . 48, 311-320. 36. Blankenship, R. E., Trost, J. T. & Mancino, L. J. (1988) in The

photosyn flzetic bacterial reaction center: structure and dynumics (Breton, J. & Vermeglio, A, , eds) pp. 119-127, Plenum Press, New York.

37. McComb, J . C., Stein, R. R. & Wraight, C. A. (1990) Biochim. Biophys. Actu 1015, 156-171.

38. De Vault, D. (1 984) Quantum-mechunical tunneNing in biological systems, Cambridge University Press, Cambridge.

39. Clayton, R. K . (1978) Biochim. Biophys. Acta 504, 255-264. 40. Hoff, A. J. (1979) Phys. Rep. 54, 75-200. 41. Jortner, J. (1 980) Bioclzim. Biophys. Actu 594, 193 - 230. 42. Sarai, A. (1980) Biochim. Biophys. Acta 589, 71 -83. 43. Kakitani, T. & Kakitani, H. (1981) Biochim. Biophys. Actu 635,

44. Bixon, M. & Jortner, J. (1986) J . Phys. Chem. YO, 3795-3800. 45. Fehcr, G., Okamura, M. & Kleinfeld, D. (1987) in Protein struc-

ture: molecular und electronic reactivity (Austin, R., Buhks, E., Chance, B., Dc Vault, D., Dutton, P. L., Frauenfelder, H. & Gol’danskii, V. I., eds) pp. 399-421, Springer-Verlag, New York.

46. Blankenship, R. E. & Parson, W. W. (1979) Biochim. Biophys.

47. Wraight, C. A. (1979) Biochim. Biophys. Actu 548, 309-327. 48. Prince, R. C., Dutton, P. L. & Bruce, J . M. (1983) FEBS Lett.

49. Shopes, R. J. & Wraight, C. A. (1986) Biochim. Biophys. Acta

50. Glasstone, S., Laidler, K. J. & Eyring, H. (1941) The theory qf rute processes, McGraw-Hill. New York.

51. Semenov, A. Yu., Kutuzov, M. A., Mamedov, M. D., Shinkarev, V. P. & Verkhovsky, M. I. (1990) in Sixth European Bioener- getics Conference. Short reports, vol. 6, p. 43, Elsevier, Amsterdam.

Biophys. Acta 333, 261 -277.

BiophJis. Acta 764, 46-54.

Biophys. Acta 776, 126- 140.

Phys. 90, 3783 - 3795.

893,409-425.

54-65.

2976.

Act0 1056, 259-272.

509.

167, 595-600.

498 - 514.

Acta 545,429 - 444.

160, 273-276.

848, 364-371.

Temperature dependence of charge recombination from the P+QA- and P+QB- states in photosynthetic...

Documents

Transcript of Temperature dependence of charge recombination from the P+QA- and P+QB- states in photosynthetic...