HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 23, 20355-20366, June 7, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/23/20355    most recent M201104200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Shen, G. Articles by Golbeck, J. H. Search for Related Content PubMed PubMed Citation Articles by Shen, G. Articles by Golbeck, J. H. Social Bookmarking                      What's this?

Assembly of Photosystem I

II. RUBREDOXIN IS REQUIRED FOR THE IN VIVO ASSEMBLY OF F_X IN SYNECHOCOCCUS SP. PCC 7002 AS SHOWN BY OPTICAL AND EPR SPECTROSCOPY^*

Gaozhong Shen, Mikhail L. Antonkine, Art van der Est^§, Ilya R. Vassiliev, Klaus Brettel^¶, Robert Bittl, Stephan G. Zech, Jindong Zhao^**, Dietmar Stehlik, Donald A. Bryant, and John H. Golbeck

From the  Department of Biochemistry and Molecular Biology, Pennsylvania State University, University Park, Pennsylvania 16802, ^§ Department of Chemistry, Brock University, St. Catharines, Ontario, L2S 3A1, Canada, ^¶ Service de Bioénergétique, CEA and URA 2096 CNRS, CEA Saclay, 91191 Gif-sur-Yvette Cedex, France,  Institut für Experimentalphysik, Freie Universität Berlin, Arnimallee 14, 14195 Berlin, Germany, and ^** College of Life Sciences, Peking University, Beijing 100871, China

Received for publication, February 2, 2002, and in revised form, March 13, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The rubA gene was insertionally inactivated in Synechococcus sp. PCC 7002, and the properties of photosystem I complexes were characterized spectroscopically. X-band EPR spectroscopy at low temperature shows that the three terminal iron-sulfur clusters, F_X, F_A, and F_B, are missing in whole cells, thylakoids, and photosystem (PS) I complexes of the rubA mutant. The flash-induced decay kinetics of both P700⁺ in the visible and A₁ in the near-UV show that charge recombination occurs between P700⁺ and A₁ in both thylakoids and PS I complexes. The spin-polarized EPR signal at room temperature from PS I complexes also indicates that forward electron transfer does not occur beyond A₁. In agreement, the spin-polarized X-band EPR spectrum of P700⁺ A₁ at low temperature shows that an electron cycle between A₁ and P700⁺ occurs in a much larger fraction of PS I complexes than in the wild-type, wherein a relatively large fraction of the electrons promoted are irreversibly transferred to [F_A/F_B]. The electron spin polarization pattern shows that the orientation of phylloquinone in the PS I complexes is identical to that of the wild type, and out-of-phase, spin-echo modulation spectroscopy shows the same P700⁺ to A₁ center-to-center distance in photosystem I complexes of wild type and the rubA mutant. In contrast to the loss of F_X, F_B, and F_A, the Rieske iron-sulfur protein and the non-heme iron in photosystem II are intact. It is proposed that rubredoxin is specifically required for the assembly of the F_X iron-sulfur cluster but that F_X is not required for the biosynthesis of trimeric P700-A₁ cores. Since the PsaC protein requires the presence of F_X for binding, the absence of F_A and F_B may be an indirect result of the absence of F_X.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The photosystem (PS)¹ I reaction center complex consists of 12 subunits in cyanobacteria and at least 13 subunits in green algae and higher plants. Only three proteins, PsaA, PsaB, and PsaC, harbor the cofactors involved in transmembrane electron transport (1). These include the primary electron donor, P700 (chlorophyll (Chl) a/a' dimer); the primary electron acceptor, A₀ (Chl a monomer); an intermediate acceptor, A₁ (phylloquinone); and three iron-sulfur cofactors, F_X, F_A, and F_B (each is a [4Fe-4S] cluster). The PsaA and PsaB proteins form a heterodimeric core that binds P700, A₀, A₁, and F_X. The PsaC protein is a 2[4Fe-4S] ferredoxin that binds the two terminal electron acceptors, F_A and F_B (2). Our understanding of the structure and function of PS I is growing at a rapid pace, in part due to the use of powerful genetic tools, their integration with sophisticated optical and magnetic resonance spectroscopies, and the availability of a 2.5-Å structure of PS I from the cyanobacterium Synechococcus elongatus (3).

A wealth of information is available on the metabolic pathways involved in the biosynthesis of the organic cofactors in PS I, including Chl a, -carotene, and phylloquinone. In contrast, little is known of the biogenesis and assembly of the bound [4Fe-4S] clusters, F_X, F_B, and F_A. The biogenesis of PS I is likely to be a multistep process involving apoprotein and cofactor metabolism, cofactor-apoprotein ligation, and assembly and integration in the membrane. The psaA, psaB, and psaC genes are encoded in the chloroplasts in plants (1), making it likely that the F_X, F_B, and F_A clusters are assembled in the chloroplasts of photosynthetic eukaryotes. The F_A, F_B, and F_X [4Fe-4S] clusters have been successfully reconstituted in vitro using ferric iron and inorganic sulfide in the presence of a sulfhydryl reagent (4). These in vitro reconstitution protocols are carried out under decidedly nonphysiological conditions, and due to the toxicity of Fe³⁺ and S², they are unlikely to resemble the process that occurs in living cells. All indications are that the assembly of iron-sulfur clusters involves enzyme-catalyzed metabolic steps (5). Specific enzymes are likely to be required in various stages of iron-sulfur cluster metabolism, including cluster assembly, insertion, and stabilization. The F_X cluster, as an interpolypeptide [4Fe-4S] cluster, may particularly have special requirements for its assembly and insertion due to its unique location between the PsaA and PsaB subunits and due to the unique constraint that it may need to be assembled in a partially assembled, photochemically active PS I core.

In the accompanying paper (6), it was shown that insertional inactivation of the rubA gene, that encodes a rubredoxin in Synechococcus sp. PCC 7002, results in the complete loss of PS I activity. Trimeric PS I complexes could be isolated from the rubA mutant, but the three extrinsic proteins, PsaC, PsaD, and PsaE, of PS I were barely detectable. Although this implies that rubredoxin has a function in the assembly of PS I, it was uncertain whether the rubredoxin functions in cofactor assembly, insertion, and stabilization or in the assembly of the three extrinsic subunits. Using a variety of spectroscopic techniques, it is shown here that the F_X iron-sulfur cluster is not assembled in PS I complexes from the rubA mutant. A preliminary report of some of these results was presented at the XIth International Congress on Photosynthesis in Budapest, Hungary (7).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell Growth and Isolation of Thylakoids and PS I Trimers-- Conditions for the growth of Synechococcus sp. PCC 7002 wild-type and mutant strains as well as the biochemical and physiological properties of the strains employed in the studies reported here are described in detail in the accompanying paper (6). The preparation of thylakoid membranes as well as monomeric and trimeric PS I fractions from isolated thylakoids was prepared as described in detail (6). Thylakoid membranes were prepared in buffer A (50 mM MES-NaOH, pH 6.5, 0.4 M sucrose, 5% (w/v) glycerol, 5 mM CaCl₂, 5 mM MgCl₂, 10 mM NaCl) at a concentration of 0.15 mg Chl ml¹. The oxygen evolution activities of isolated thylakoids from the psaAB and psaAB rubA mutants were 1100 µmol (mg of Chl h)¹, and 780 µmol (mg of Chl h)^-1, respectively. The trimeric and monomeric PS I complexes were collected, resuspended in 50 mM Tris, pH 8.0, containing 0.03% (w/v) n-dodecyl--D-maltopyranoside (DM) and 15% (v/v) glycerol at a Chl concentration of 1.5-2 mg ml¹, and stored at 80 °C until required.

Purification of PS II Complexes from psaAB and psaAB rubA Mutant Strains-- The thylakoid membranes in buffer A at a Chl concentration of 0.15 mg ml¹ were solubilized with DM at a final concentration of 0.5% (w/v) from a stock solution of 20% (w/v). The solution was stirred gently on ice in the dark for 1 h. Solubilized complexes were loaded onto sucrose step gradients (40% (w/v) sucrose, 5 ml; 30% (w/v) sucrose, 5 ml; 25% (w/v) sucrose, 6 ml; 5 ml of the sample solution; the sucrose solutions were prepared with buffer A). The gradients were centrifuged in a Beckman SW27 rotor at 26,000 rpm for 16 h at 4 °C. The PS II complexes were identified as a green band at the interface between the 30 and 40% sucrose layers. The PS II complexes were removed and diluted with buffer B (50 mM MES-NaOH, pH 6.5, 10 mM NaCl) to a sucrose concentration of ~0.4 M. The solution was centrifuged at 100,000 × g 1 h at 4 °C. The green pellet containing the PS II complexes was resuspended with buffer A 1.0 mg of Chl ml¹.

Insertion of the [4Fe-4S] Cluster into the F_X Site in PS I Complexes from the rubA Mutant-- The [4Fe-4S] cluster was inserted into the F_X site in PS I trimers isolated from the rubA mutant according to published procedures (8). Briefly, the protocol relies on the ability of an inorganic [4Fe-4S] cluster to undergo a ligand exchange reaction with the cysteines on PsaA and PsaB to allow insertion of the cluster into the F_X site. The inorganic iron-sulfur clusters are synthesized by adding 10 mM FeCl₃ and 10 mM Na₂S to a solution of 1% (v/v) 2-mercaptoethanol in 25 mM Tris-HCl buffer, pH 8.3. The 2-mercaptoethanol serves the added function of cleaving any potential Cys-S^o-S^o-Cys bonds that may be present in the F_X site and makes free cysteine residues available for ligation to the [4Fe-4S] cluster. The inorganic agents are removed by ultrafiltration or gel filtration chromatography, leaving a reconstituted P700-F_X core.

Flash-induced Transient Absorption Spectroscopy in the Near-IR and Blue Regions-- Transient absorbance changes at 820 nm were measured at room temperature with a laboratory-built dual beam spectrometer as described (9). The sample was contained in a 10 × 4-mm quartz cuvette positioned so that the optical path length was 10 mm. The measuring and reference beams were balanced using a variable density optical filter wheel; the difference signal was amplified with an 11A33 differential comparator and processed using a DSA 610 digital signal analyzer (Tektronix, Beaverton, OR). The upper electrical bandwidth of the detection system was about 10 MHz. Transient absorbance changes in the blue region were measured with a laboratory-built, single beam spectrometer consisting of an Oriel 250-W quartz-tungsten source, two Jarrell-Ash 1/4 meter monochromators (one located before and one after the sample cuvette), and a Schottky-barrier photodiode (PIN 8LC; UDT Sensors, Hawthorne, CA). The sample was contained in a 10 × 10-mm quartz cuvette. A shutter placed in the measuring beam in front of the sample was opened 10 ms before the excitation flash and closed 100 ms after the flash to minimize exposure of the sample to the measuring light. Intensity changes of the measuring light transmitted by the sample were detected by a reverse-biased silicon photodiode (PIN-10D; UDT Sensors), The detector photocurrent was converted to a voltage with a 5,000-Ohm load resistor, and the voltage was amplified (Princeton Applied Research model 113) and processed using a Nicolet 4094A digital oscilloscope. The DC offset from the measuring beam was nulled using a voltage injected into the inverting port of the digital oscilloscope. The upper electrical bandwidth of the detection system was about 100 KHz. The flash excitation for both spectrometers was provided by a frequency-doubled, Q-switched DCR-11 Nd-YAG laser. The flash energy was adjusted by changing the Q-switch delay and/or using neutral density filters. The kinetics were fitted to "sum of several exponentials with base line" using the Marquardt algorithm in Igor Pro (Lake Oswego, OR).

Flash-induced Transient Absorption Spectroscopy in the Near-UV Region-- Measurements were made with a laboratory-built spectrometer in the near-UV and blue spectral regions on PS I complexes from the rubA mutant, similar to Ref. 10. Excitation flashes of 300-ps pulse duration at 532 nm were provided with a Nd-YAG laser (Quantel, France). The measuring light source was a 100-W tungsten halogen lamp filtered by appropriate interference and colored glass filters. The sample was contained in a 4 × 10-mm quartz cuvette oriented with the longer side perpendicular to the direction of the measuring light and the shorter side perpendicular to the direction of the excitation flash. A shutter placed in the measuring light beam in front of the sample was opened 3 ms before the excitation flash and closed 10 ms after the flash to diminish exposure of the sample to the measuring light. Intensity changes of the measuring light transmitted by the sample were detected by a silicon photodiode (S3590 from Hamamatsu, Japan), The detector photocurrent was converted to a voltage with a 3,000-Ohm load resistor, and the voltage was amplified (AM502 from Tektronix, Beaverton, OR) and registered by a DSA 602A digitizing oscilloscope with plug-in 11A33 (Tektronix, Beaverton, OR). The upper electric bandwidth limit of the detection system was about 200 kHz. A Marquardt least-squares algorithm program was used for fitting of the absorption change transients to a multiexponential decay.

CW EPR Spectroscopy of PS I Complexes at X- and Q-bands-- X-band (9.4-GHz) EPR spectra were recorded using a Bruker ECS 106 electron paramagnetic resonance (EPR) spectrometer operating with an ER462ST resonator. Cryogenic temperatures were maintained with an Oxford liquid helium cryostat and an ITC4 temperature controller. The microwave frequency was determined with a Hewlett-Packard 5340A frequency counter. Continuous actinic illumination of the sample was provided by a 150-W model 66057 xenon arc source (Oriel Corp., Stamford, CT), filtered through 2 cm of water, and passed through a heat-absorbing filter to remove the near-infrared wavelengths. Samples used for measurement of F_A and F_B in the wild-type and the rubA mutant contained 0.6 mg of Chl ml¹, 1 mM sodium ascorbate, 30 µM DCPIP, 0.02% (v/v) DM, and 20% (v/v) glycerol in 50 mM Tris-HCl buffer, pH 8.3. For measurement of F_X, the pH of the sample was adjusted to 10.0 by the addition of 1 M glycine buffer, prior to the addition of ~1 mg of dithionite. Samples were frozen in darkness and illuminated inside the EPR cavity. For freezing under continuous illumination, a dry ice-methanol bath was used. Q-band (34 GHz) EPR studies were carried out using an ER 5106 QT-W1 resonator as described (11).

CW EPR Spectroscopy of PS II Complexes at X-band-- The spectrum of the spin-coupled [Q Fe²⁺] signal in PS II was taken by subtracting the spectrum of the light-treated sample from the spectrum of a dark-frozen sample. The dark sample was prepared by adding potassium ferricyanide to a final concentration of 1 mM and duroquinone to a final concentration of 20 µM to 250 ml of PS II complexes isolated from the psaAB and psaAB rubA mutants. The sample was incubated in darkness for 20 min on ice and frozen in an EPR tube with liquid nitrogen under very dim green light. The light sample was prepared by adding 3-(3,4-dichlorophenyl)-1,1-dimethylurea to a final concentration of 50 µM hydroxylamime to a final concentration of 5 mM and sodium acetate to a final concentration of 20 mM from a 4 M stock solution. The photoaccumulation protocol was carried out by illuminating the PS II complexes in an EPR tube for 10 min in an ice water bath. The sample was quickly frozen in liquid nitrogen and illuminated further for 5 min prior to measurement.

Transient EPR Spectroscopy at X-, Q-, and W-bands-- Transient EPR spectroscopy at low temperature was carried out at X-band using a Bruker ER046 XK-T microwave bridge equipped with a dielectric resonator, at Q-band using a Bruker ER056 QMV bridge and a home-built cylindrical resonator (12), and at W-band using a Bruker E680 spectrometer. The temperature was controlled using an Oxford liquid helium gas-flow cryostat. For X-band experiments at room temperature, a Bruker ER041M microwave bridge modified for broadband detection was used with a rectangular resonator and a flat cell. The spin-polarized EPR data were collected in direct detection mode by accumulating the light-induced EPR transients at fixed magnetic field positions over an appropriate range. The samples were illuminated using a Nd-YAG/OPO laser system using either 623- or 532-nm excitation. At W-band, illumination was accomplished using an optical fiber fed into the sample capillary and ending directly above the PS I sample. The samples contained ~1 mM sodium ascorbate and 50 µM phenazine methosulfate as external redox agents and were frozen in the dark for the low temperature experiments.

Electron Spin Echo Spectroscopy at X-band-- Electron spin-echo (ESE) amplitude-modulation curves were obtained by collecting the amplitude of the echo detected at time T   after the second microwave pulse as function of . The microwave pulse lengths were set to 8 ns for the first pulse and 16 ns for the second pulse, resulting in flip angles of about 65 and 130°, respectively, with a microwave field, B₁, of 1.1 mT. The external magnetic field B₀ and the detection time T were adjusted to yield maximum out-of-phase echo intensity. The echo mTmodulation was recorded at 512 points with 8-ns increments in , and 64 traces were averaged to increase the signal/noise ratio (see Ref. 13 for details).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

X-band CW EPR Spectroscopy of F_A and F_B-- SDS-PAGE and immunoblotting indicate that PsaC is missing in thylakoid membranes and trimeric PS I complexes from the rubA mutant (6). It follows, therefore, that the F_A and F_B iron-sulfur clusters must also be missing. The presence or absence of F_A and F_B was determined in PS I trimers isolated from the wild-type and the rubA mutant using low temperature EPR spectroscopy. The experimental protocols involved 1) freezing samples to 15 K in darkness in the presence of ascorbate and DCPIP and illumination inside the EPR cavity to promote one electron from P700 to the bound iron-sulfur clusters and 2) thawing the sample and refreezing to 15 K under continuous illumination to promote two (or more) electrons to the bound iron-sulfur clusters (4). Under the one-electron reduction protocol, PS I trimers from the wild type showed g values of 2.0025 characteristic of P700⁺; 2.044, 1.942, and 1.855 characteristic of F_A; and 2.071, 1.931, and 1.878 characteristic of F_B (Fig. 1 top, dotted line). The F_A/F_B ratio of 3.1:1 is typical of cyanobacterial PS I trimers (2). Under the two-electron reduction protocol, PS I trimers isolated from the wild type showed resonances at g = 2.0025 characteristic of P700⁺ and at g = 2.05, 1.94, 1.92, and 1.87 characteristic of the interaction spectrum of F_A/F_B (Fig. 1 top, solid line). Under both the one-electron and two-electron reduction protocols, trimeric PS I complexes isolated from the rubA mutant showed a light-induced resonance at g = 2.0025 but no signals that could be attributed to the reduction of F_A or F_B (Fig. 1 bottom, dotted and solid lines).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   EPR spectra of PS I trimers isolated from the wild type and the rubA mutant in 50 mM Tris-HCl buffer, pH 8.3, containing 1 mM sodium ascorbate, 30 µM DCPIP, 0.02% (v/v) DM, and 20% (v/v) glycerol. The Chl concentration is 0.6 mg ml1. Dotted lines represent the spectrum after freezing the sample in darkness and illumination at 15 K. Solid lines represent the spectrum after illuminating the sample during freezing to 15 K. EPR conditions were as follows: temperature, 15 K; microwave power, 20 milliwatts; modulation amplitude, 1 mT; microwave frequency, 9.477 GHz.

EPR measurements under the same experimental setup were performed on whole cells, thylakoid membranes, and PS I monomers from both the wild-type and rubA mutant, and the results obtained were identical to the results described above for PS I trimers (data not shown). This indicates that the loss of the F_A and F_B iron-sulfur clusters is not due to the loss of PsaC that might have incurred during the isolation and/or purification protocol. Thus, these findings are consistent with the previous SDS-PAGE and immunoblotting studies (6) that showed little or no PsaC in thylakoid membranes or in PS I trimers isolated from the rubA mutant.

X-band EPR Spectroscopy of F_X-- The absence of F_A and F_B raises the question whether the F_X iron-sulfur cluster is also missing in the rubA mutant. The experimental protocol to detect F_X involves prereducing F_A and F_B with sodium hydrosulfite and freezing the sample to 15 K under continuous illumination to additionally reduce F_X. PS I trimers isolated from wild type showed resonances with g values of 2.15, 1.90, and 1.78 characteristic of F_X as well as g values of 2.05, 1.94, 1.92, and 1.86 characteristic of the interaction spectrum of reduced F_A/F_B (data not shown). In contrast, PS I trimers isolated from the rubA mutant contained no detectable F_A, F_B, or F_X (data not shown). These data show that no chemically reducible F_A or F_B is present in the rubA mutant and that F_X either cannot be reduced using light or is totally absent.

CW EPR Spectroscopy of Photoaccumulated A₁-- The presence of the P700⁺ radical in the light-induced EPR study (Fig. 1B) implies that an acceptor earlier than the iron-sulfur clusters is present and functional in the rubA mutant. To determine whether the A₁ phylloquinone is present and functional, Q-band EPR photoaccumulation studies were carried out on the wild-type and the rubA mutant. Photoaccumulation of A₁ in PS I trimers from the wild type resulted in a spectrum with g values of 2.0064, 2.0034, and 2.0023 and four prominent hyperfine lines from the 2-methyl group characteristic of the phylloquinone semiquinone anion (data not shown; see Ref. 11). Similar studies of PS I trimers from the rubA mutant resulted in a spectrum of photoaccumulated A₁ with a similar g anisotropy but with a less prominent contribution of the hyperfine couplings compared with the wild type (not shown). These data show that phylloquinone is present in the A₁ site of the rubA mutant and that it is functional in accepting electrons from A₀.

Flash-induced Absorbance Changes of P700 in the Near-IR-- If all three bound FeS clusters are missing in the rubA mutant, but A₁ is present, then charge separation and recombination should occur between P700⁺ and A₁ after a single turnover flash. Fig. 2 shows A₈₂₀ decay kinetics measured in PS I trimers isolated from the rubA mutant. Multiexponential deconvolution of the kinetic transients permits the relative contributions of different PS I acceptors to P700⁺ reduction to be tentatively identified (9). The kinetics of P700⁺ reduction in the wild type show contributions arising from components with (1/e) lifetimes () of 23 and 144 ms, both of which are assigned to the back-reaction of P700⁺ with [F_A/F_B] (data not shown). A minor component with  = 2.9 s arises from forward electron donation with reduced DCPIP in PS I trimers that have lost an electron from the bound iron-sulfur clusters. The kinetics of P700⁺ reduction in PS I complexes from the rubA mutant show a striking absence of components with lifetimes in the millisecond time range (Fig. 2). The major contributions arise only from components with  = 13 and 86 µs, both of which are characteristic of charge recombination between P700⁺ and A₁ in PS I trimers devoid of the bound iron-sulfur clusters F_X, F_B, and F_A (10). The initial amplitude of the A₈₂₀ derived from the multiexponential fit corresponds to a Chl/P700 ratio of ~100 and is similar to that measured in wild-type PS I complexes. The analysis of the kinetics of P700⁺ recombination therefore confirms the EPR evidence that electron transfer to the terminal iron-sulfur clusters, F_X, F_B, and F_A, does not occur in the rubA mutant.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Kinetics of flash-induced absorbance changes at 820 nm in PS I trimers isolated from the rubA mutant. Lower panel, experimental points (dots) fitted by a three-exponential decay (dotted lines). The solid line represents the sum of the three exponentials. Upper panel, residuals (difference between experimental and fit curve shown in the lower panel). Sample was suspended in 25 mM Tris-HCl buffer, pH 8.3, containing 10 mM sodium ascorbate, 4 µM DCPIP, and 0.02% (w/v) DM. Conditions were as follows: Chl concentration, 50 µg ml1; optical path length for the measuring light, 10 mm; excitation, about 3.7 mJ cm2 in ~10-ns laser pulses at 532 nm.

Flash-induced P700⁺ reduction in freshly isolated thylakoid membranes of the rubA mutant revealed similar kinetics (data not shown). This indicates that the loss of components with lifetimes in the millisecond time range was not due to the loss of F_X, F_B, and F_A in the isolation and/or purification procedure leading to the PS I trimers.

Flash-induced Absorbance Changes in the Near-UV-- If charge recombination occurs between P700⁺ and A₁ in the rubA mutant, then the kinetics of A₁ oxidation determined in the near-UV will match that of P700⁺ reduction determined at 820 nm. The kinetics of flash-induced absorbance changes at individual wavelengths between 335 and 500 nm were measured with a time resolution of about 2 µs. For example, at 380 nm, where A₁ should be the dominant contributor to absorbance changes in PS I (15), 98% of the initial absorbance increase decayed biexponentially with 1/e lifetimes () of 15 and 95 µs (Fig. 3A). Global analysis of the complete data set (kinetic traces measured at 19 different wavelengths) showed that all transients could be fitted reasonably well with two major exponential phases of  = 15 µs and  = 95 µs, and a minor phase of  = 2.6 ms.² The amplitudes of the three phases obtained from the global analysis are shown as point-by-point spectra in Fig. 3B. The two major components ( = 15 and 95 µs) have very similar optical difference spectra from 335 to 500 nm. Both components display an absorbance increase from 350 to 400 nm, followed by a trough between 420 and 440 nm. The 420-440-nm trough is attributed primarily to the flash-induced P700⁺/P700 difference spectrum, and the absorbance increase between 350 and 400 nm is attributed primarily to the flash-induced A₁/A₁ difference spectrum (16). The fit with  = 15 and 95 µs in the near-UV is very similar to the fit obtained from independent measurements at 820 nm (Fig. 2) and attributed solely to P700⁺/P700. The kinetics and spectra are also highly similar to those of a wild-type P700-A₁ core stripped of the F_X, F_A, and F_B iron-sulfur clusters (10). These results confirm that the PS I complexes from the rubA mutant exhibit charge recombination between P700⁺ and A₁, consistent with the absence of forward electron transfer to F_X, F_B, and F_A.

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Kinetic and spectral analysis of flash-induced UV/blue absorbance changes in PS I trimers from the rubA mutant. A, upper trace, transient measured at 380 nm, fitted by a three-exponential decay with 1/e lifetimes () of 15 µs (43% of the initial amplitude), 95 µs (55%), and 2.6 ms (2%). A, lower trace, residuals (difference between measured transient and fit curve) calculated for the upper trace. B, point-by-point spectra of the amplitudes of the kinetic phases with t = 15 µs (full triangles), 95 µs (full circles), and 2.6 ms (open squares), as obtained from global analysis of transient absorbance changes (as the example shown in A) measured at different wavelengths and also on longer time scales. The sample was suspended in 25 mM Tris-HCl buffer, pH 8.3, containing 5 mM sodium ascorbate, 30 µM DCPIP, and 0.03% (w/v) DM. Conditions were as follows: Chl concentration, ~35 µM; optical path length for the measuring light, 4 mm; excitation, about 500 µJ cm2 in 300-ps laser pulses at 532 nm. At each wavelength, 256 transients were averaged at a repetition rate of 2 Hz. The sample was replaced by a fresh one after 768 flashes.

Room Temperature Transient EPR Spectroscopy at X-band-- The cycling of electrons between P700 and A₁ in the rubA mutant should be particularly evident in the electron spin-polarized transient EPR signal. In wild-type samples at room temperature, electron transfer from A₁ to the iron-sulfur centers generates two sequential spin-polarized EPR spectra due to P⁺A₁ and P⁺(FeS) (12, 17). In Fig. 4, the X-band room temperature, spin-polarized EPR transients in PS I complexes isolated from the wild type and the rubA mutant are compared. The corresponding decay-associated spectra are shown in Fig. 5, and the vertical arrow indicates the field position at which the transients in Fig. 4 were taken. The field position was chosen such that for the wild-type (Fig. 4, bottom; Fig. 5, dashed curves), the signal due to P⁺A₁ is absorptive (positive) and that of P⁺(FeS) is emissive (negative). As can be seen in Fig. 4 (top), the emissive part of the transient due to P⁺(FeS) is missing in the rubA mutant, and as a consequence only one decay-associated spectrum is obtained. It has been shown previously (12) that the emissive part of the transients and corresponding P⁺(FeS) spectrum is also absent in samples in which iron-sulfur clusters F_X, F_B, and F_A have been removed. However, this signal component is present in PS I core preparations containing F_X but lacking the PsaC protein subunit, which binds F_A and F_B. As discussed in Refs. 17 and 12, the transition from absorption to emission is governed by the electron transfer from A₁ to F_X, whereas the decay of the emissive part of the transients reflects the relaxation of the spin polarization. In samples such as the rubA mutant and F_X-depleted P700-A₁ core particles isolated from the wild type, only the signal due to P700⁺A₁ and its decay due to spin relaxation are observed. Thus, it can be concluded that, no forward electron transfer from A₁ occurs within the lifetime of the polarization. Taken together with the optical data, which show only recombination from P⁺A₁, it is clear that forward electron transfer past A₁ does not occur in the rubA mutant.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Comparison of the room temperature transient EPR kinetic traces of PS I complexes from the rubA mutant (top) and wild type (bottom). The vertical arrow in Fig. 5 indicates the field position at which the transients were taken.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Decay-associated transient EPR spectra at two temperatures. Spectra of the rubA mutant (solid curve) and wild-type (dashed curves) extracted from the full time/field data sets as described in detail in Ref. 12. For the wild type sample, the curves corresponding to the states P700+A1 and P700+ FeS are indicated, and for the rubA mutant only P700+A1 is observed. The term FeS is used to indicate one of the three iron-sulfur clusters, FX, FA, or FB, when the identity of the iron-sulfur center acceptor involved is not known with certainty. The vertical arrow indicates the field position corresponding to the transients shown in Fig. 4.

Low Temperature Transient EPR Spectroscopy at X-, Q-, and W-bands-- Spin-polarized EPR spectra of the rubA mutant and wild-type PS I in frozen solution are shown in Fig. 6 for three different microwave frequencies. As can be seen, there are no significant differences between the spectra of the rubA mutant and the wild-type, clearly showing that the lack of the F_X cluster does not perturb the orientation of the electron-active A₁ acceptor. Yet, the spectra of P⁺A₁ at room temperature (Fig. 5) show subtle differences between the two samples. The X-band spectra (Fig. 5) are very sensitive to changes in the line widths, while the spectra at higher frequency (Fig. 6, top and middle) are more sensitive to the orientation of the quinone. The fact that small differences appear in the X-band P⁺A₁ spectra at room temperature (Fig. 5) but not at low temperature (Fig. 6, bottom) suggests that the motion of the phylloquinone in the rubA mutant may be greater due to the absence of F_X and the stromal subunits PsaC, PsaD, and PsaE.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Transient spin-polarized EPR spectra of the state P700+ A1 at low temperature. The spectra of the rubA mutant (solid lines) and wild type are compared at three different microwave frequencies. From top to bottom, 94 GHz (W-band), 35 GHz (Q-band), and 9 GHz (X-band). The spectra have been extracted from the experimental data as described in Ref. 22, and a linear base-line correction was made to the RubA spectra to remove a background signal from 3P700.

Illumination of wild-type PS I at temperatures below ~200 K leads to stable charge separation between P700⁺ and [F_A/F_B] in about two-thirds of the PS I trimers, while an electron cycle between P700 and A₁ occurs in the remaining 1/3 (see Ref. 15 for a review). In transient EPR experiments, only the latter fraction is observed. In the rubA mutant, the fact that stable charge separation does not occur and that only an electron cycle involving A₁ takes place results in much stronger polarization patterns of P⁺A₁ than from analogous wild-type samples as shown in the middle and lower spectra in Fig. 5. (Note that the spectra for PS I complexes from the rubA mutant and wild type in Fig. 6 have been normalized to the same amplitude). The difference in amplitude between the two samples is difficult to quantify accurately, but it is in approximate agreement with the ratio of the electron cycle involving A₁ and stable charge separation to F_A/F_B found in wild type PS I.

The occupancy of the A₁ site can also be probed by transient EPR. One might expect that the lack of F_X could disturb the binding of phylloquinone and that a fraction could be lost from the PS I reaction centers. In this case, charge recombination from A₀ would generate a spin-polarized EPR spectrum of the triplet state of P700 (18) from that fraction without phylloquinone. A weak spectrum due to ³P700 formed by recombination from A₀ is indeed observed in the rubA mutant (data not shown). However, it accounts for at most a small fraction of reaction centers, indicating that no significant loss of the active phylloquinone occurs, despite the fact that F_X is not present.

This conclusion is consistent with flash-induced absorbance changes at 820 nm monitoring the kinetics of P700⁺ with a time resolution of 2 ns.³ According to these measurements, primary pair (P700⁺A₀) charge recombination (  50 ns) occurred in only about 5% of PS I from the rubA mutant. The addition of sodium dithionite at pH 8.3 increased this fraction only slightly. This behavior is unlike that of wild-type P700-A₁ core complexes stripped of the F_X, F_A, and F_B iron-sulfur clusters by chemical means (19, 20), in which about 50% of primary pair recombination was observed after the same dithionite treatment, presumably because of reduction of a significant fraction of the phylloquinone A₁ (10). Since dithionite at pH 8.3 does not reduce A₁ in intact wild-type PS I complexes, it appears that the native environment of phylloquinone is better preserved in PS I isolated from the rubA mutant than in the chemically extracted P700-A₁ core preparations.

Electron-Spin Echo Distance Measurements of P700⁺-A₁-- ESE experiments have been used to determine the distance between P700⁺ and A₁ in wild-type PS I (13). The ESE of a weakly coupled spin-correlated radical pair is phase-shifted by 90° compared with that of a single radical and shows deep amplitude modulations as a function of the pulse spacing (21). The modulation frequency is selectively determined by the spin-spin coupling, the dipolar part of which yields the distance between the spin density centers of the respective radicals. The out-of-phase ESE amplitude modulation curves from PS I trimers isolated from the wild-type and the rubA mutant strains are compared in Fig. 7. The top panel shows data for PS I complexes from the rubA mutant, and the bottom panel shows data for wild-type complexes (22). Comparison of the echo modulation curves shows that the dominant frequency components from both samples are virtually identical.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Comparison of the out-of-phase echo modulation of PS I complexes from wild-type and the rubA mutant at 80 K. Experimental conditions are described under "Materials and Methods."

Simulation of the ESE modulation curves in Fig. 7 yields dipolar coupling constants, D = 172 ± 4 µT for both the wild type and the rubA mutant. In both samples, the isotropic exchange coupling is extremely weak, and only an upper limit of J < 1.0 ± 0.8 µT can be given. Using a simple point-dipole model, the measured dipolar coupling constant corresponds to a distance of 25.3 ± 0.3 Å between P700⁺ and A₁ in the rubA mutant, which is the same within error as the distance of 25.4 ± 0.3 Å established between P700⁺ and A₁ in several preparations of wild-type PS I (13, 23-25). Thus, the absence of the F_X cluster does not influence the position and orientation of phylloquinone in the A₁ site. It can be concluded therefore that the phylloquinone binding site in PS I complexes from the rubA mutant remains largely intact.

Time-resolved Optical Spectroscopy of P700 after Reconstituting F_X-- While the spectroscopic data exclude forward electron transfer beyond A₁ in the rubA mutant, this could be due to either the absence of F_X or to a very slow, forward electron transfer from A₁ to F_X that would not be able to compete kinetically with the fast P700⁺ A₁ charge recombination. To decide between these two alternatives, an iron-sulfur cluster insertion protocol was employed to reconstitute F_X into its site in PS I trimers isolated from the rubA mutant. If F_X is present but cannot be reduced by light, then the insertion protocol should result in the retention of the 17- and 70-µs charge recombination between P700⁺ and A₁; however, if F_X is absent, then the insertion protocol should result in the appearance of a 1-ms charge recombination between P700⁺ and F_X (8).

The reconstitution of F_X was carried out by the addition of ferric iron, sodium sulfide, and 2-mercaptoethanol to PS I trimers isolated from the rubA mutant followed by repurification of the PS I complex. Fig. 8A shows that the lifetime of the major kinetic phase of the flash-induced absorbance change at 432 nm has increased to about 1 ms. There exist slower kinetic phases (only the ~5-ms phase is shown), which probably represent electron donation from DPIP to P700⁺ in a population of reaction centers in which the electron is lost to a solution acceptor. Fig. 8B shows that the ~1-ms lifetime is typical of the P700⁺ F_X back-reaction measured in P700-F_X cores isolated from wild-type PS I (26). In this preparation, the ~100-µs kinetic phase probably represents charge recombination between P700⁺ and A₁ in a population of reaction centers in which F_X has been inadvertently destroyed. These results show that the F_X binding site in the PS I trimers isolated from the rubA mutant is apparently normal and can be reconstituted under the same conditions that work for wild-type core complexes lacking F_X.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Kinetics of flash-induced absorbance changes at 432 nm in PS I trimers isolated from the rubA mutant after reinsertion of the FX iron-sulfur cluster (A) and P700-FX cores prepared by removing PsaC, PsaD, and PsaE from wild-type PS I trimers with 7.8 M urea (B). Lower panels, experimental points (dots) fitted by a two-exponential decay (dotted lines). The solid line represents the sum of the two exponentials. Upper panels, residuals (difference between experimental and fit curve shown in the lower panel). Samples were suspended in 25 mM Tris-HCl buffer, pH 8.3, containing 10 mM sodium ascorbate, 4 µM DCPIP, and 0.02% (w/v) DM. Conditions were as follows: Chl concentration, 15 µg ml1; optical path length for the measuring light, 10 mm; excitation, 2.0 mJ cm2 in ~10-ns laser pulses at 532 nm.

EPR Spectroscopic Detection of the Non-heme Iron in PS II-- The rubA gene is located upstream from genes encoding a protein known to affect PS II biogenesis in plants (HCF135) and the structural genes for several PS II subunits (psbEFLJ). This leads one to question whether the non-heme iron of PS II is inserted normally in the rubA mutant. Moreover, the absence of the F_X, F_B, and F_A iron-sulfur clusters in PS I raises the issue of whether iron incorporation into other soluble and membrane-bound proteins, particularly into the Rieske iron-sulfur cluster of the cytochrome b₆f complex, is similarly affected. One can make several inferences from the growth data in Ref. 6. The rubA mutant can be grown on nitrate as sole nitrogen source, which implies that the iron-sulfur proteins nitrate reductase, nitrite reductase, and ferredoxin are all present. Moreover, because the rubA mutant grows under photoheterotrophic conditions (i.e. in the light with glycerol as carbon source), it is highly unlikely that iron-sulfur centers are absent in type-I NADH dehydrogenase and the Rieske protein in the cytochrome b₆f complex. The latter conclusion is strengthened by the presence of a g = 1.94 EPR resonance characteristic of the Rieske iron-sulfur cluster in whole cells and thylakoid membranes from the rubA strain (data not shown). In PS II, a non-heme Fe²⁺ is located between the primary acceptor Q_A and the secondary acceptor Q_B, with ligands provided by the D1 and D2 polypeptides.

To verify whether interruption of the rubA gene prevents the assembly of this non-heme iron, PS II complexes were isolated from the psaAB and psaAB rubA mutant strains of Synechococcus sp. PCC 7002. As shown in Fig. 9, PS II complexes from both PS I-less mutant strains display an axial EPR spectrum with g values of 1.83 and 1.73. This signal represents the narrow (~20-mT-wide) [Q_A Fe²⁺] EPR signal that arises from Q_A spin-coupled to the acceptor-side iron ferrous iron and is diagnostic of the non-heme iron in PS II (27). The spectrum is nearly identical to that observed in wild-type PS II complexes from the wild-type (data not shown). There is also evidence for a small population of complexes with the broad (40-mT-wide) [Q_A Fe²⁺] EPR signal at g = 1.87 and 1.64. When the non-heme iron is missing, the coupling between Q_A and the non-heme iron is lost, and the decoupled, free Q_A shows a sharp resonance at g = 2.0046. However, there is no evidence for this signal in thylakoids from either the wild type or the rubA mutant (data not shown), thus implying that the PS II non-heme iron occupies all available sites. These results indicate that a deficiency in rubredoxin has no effect on the assembly of the non-heme iron in PS II.

View larger version (12K): [in this window] [in a new window]   Fig. 9.   EPR spectra of QA coupled to the non-heme iron in PS II complexes isolated from the PS I-less strains psaAB and psaAB rubA. Sample conditions were as follows: thylakoid membranes at 500 µg ml1; Chl was reduced with 50 mM dithionite in 100 mM Tris buffer, pH 8.0, and illuminated for 15 min with white light. Spectrometer conditions were as follows: microwave frequency, 9.736 GHz; microwave power, 40 milliwatts; modulation frequency, 100 kHz; modulation amplitude, 20 G; temperature, 6 K. The spectrum represents the average of 16 scans.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Deficiency of Rubredoxin Leads to the Immediate Loss of the F_X Cluster-- As outlined in the companion paper (6), PS I monomers and trimers can be isolated from the rubA mutant that contain the following subunits: PsaA, PsaB, PsaF, PsaI, PsaJ, PsaK, PsaL, and PsaM. Consistent with the absence of the PsaC, PsaD, and PsaE subunits, these PS I reaction centers are completely incapable of light-driven, transmembrane electron transport from cytochrome c₆ to ferredoxin/flavodoxin. It is shown here that the Chl/P700 ratio of PS I trimers isolated from the rubA mutant is similar to that of wild-type PS I complexes, but electron transport does not occur beyond the intermediate electron acceptor, phylloquinone A₁. Evidence to support this conclusion can be summarized as follows: 1) no light-induced or chemically reduced EPR signals from F_X, F_A, and F_B could be detected in membranes or PS I trimers isolated from the rubA mutant; 2) P700⁺ is reduced with lifetimes of ~10 and 100 µs after a single-turnover flash, both of which are characteristic of charge recombination between P700⁺ and A₁ in PS I core complexes devoid of F_X, F_A, and F_B; 3) flash-kinetic absorption spectroscopy in the near-UV exhibits similar kinetics, with a decay-associated spectrum characteristic of A₁; and 4) a single spin-polarized, transient EPR spectrum due to P700⁺ A₁ is observed in PS I complexes from the rubA mutant under conditions where two sequential spectra due to P700⁺ A₁ and P700⁺(FeS) characteristic of functional PS I complexes are found for PS I complexes from the wild type.

These results further indicate that the primary electron donor P700, the primary electron acceptor A₀, and the intermediate electron acceptor A₁ are unaltered in PS I complexes from the rubA mutant. Evidence to support this conclusion includes the unaltered polarization pattern of the transient high field (W- and Q-band) EPR spectra, which indicate that the orientation of phylloquinone in the rubA mutant is identical to that of wild-type PS I, and the out-of-phase electron spin-echo data, which show that the distance between the phylloquinone anion radical and the P700⁺ radical in PS I trimers from the rubA mutant is identical to that for wild-type PS I. In addition, the non-heme iron in PS II is not affected, and the ability of the rubA mutant to grow under photoheterotrophic conditions implies that the iron-sulfur clusters in the type-1 NADH dehydrogenase remain intact. Other cellular processes that depend on proteins containing iron-sulfur clusters are also not affected; for example, the mutant is capable of growth on nitrate as a sole nitrogen source, which indicates that ferredoxin, nitrate reductase, and nitrite reductase must all be functional. Hence, it appears that rubredoxin RubA is required solely for the assembly of the F_X, F_B, and F_A iron-sulfur clusters in PS I.

Information about the order of assembly of the iron-sulfur clusters and stromal polypeptides allows us to refine these conclusions. Our knowledge of cluster and subunit assembly in PS I is derived, in part, from in vitro reconstitution studies performed with P700-F_X cores and recombinant PsaC, PsaD, and PsaE proteins produced in Escherichia coli. From these studies, it is known that prior binding of PsaC is required before PsaD and PsaE will bind to P700-F_X cores (reviewed in Ref. 28). If the F_A and F_B iron-sulfur clusters are missing in PsaC, then apoPsaC does not bind to P700-F_X cores. If the F_X cluster is missing, then PsaC does not bind to P700-A₁ cores.⁴ These in vitro studies, therefore, show that the F_X cluster is a prerequisite for the binding of PsaC, which, in turn, is required for the binding of PsaD and PsaE. Further knowledge of cluster and subunit assembly in PS I is derived from in vivo studies of mutants lacking PsaC, PsaD, and PsaE. In these studies, it was shown that an interposon mutant of psaC in Synechocystis sp. PCC 6803 contains F_X but lacks PsaC, PsaD, and PsaE (29). The isolation of a P700-F_X core from the PsaC mutant using either DM or Triton X-100 shows that the F_X cluster is stable in the absence of PsaC, PsaD, or PsaE (29). Hence, one can exclude the possibility and that the absence of F_X in the RubA mutant is due to instability of the iron-sulfur cluster in the absence of the stromal polypeptides. Further, a deletion mutant of psaD Synechocystis sp. PCC 6803 lacks PsaD, but it contains F_X and wild-type levels of PsaC and PsaE (30). Similarly, deletion mutants of psaE in Synechococcus sp. PCC 7002 (31) and Synechocystis sp. PCC 6803 (32) lack PsaE, but they contain F_X and wild-type levels of PsaC and PsaD. These in vivo studies, therefore, show that PsaC, PsaD, and PsaE are not required for the assembly of the F_X iron-sulfur cluster. Were the absence of RubA to preclude the biosynthesis of F_X, the phenotypic consequences would be precisely what is observed: the synthesis of a P700-A₁ core lacking F_X, PsaC, PsaD, and PsaE. One can therefore conclude from the results presented here that RubA is involved in the assembly of F_X. Since F_X is required for PsaC binding, the absence of F_A and F_B may simply be a secondary consequence of the absence of F_X, Finally, an important corollary to these conclusions is that F_X is not required for the biosynthesis of trimeric P700-A₁ core complexes or for their assembly in the thylakoid membrane of Synechococcus sp. PCC 7002.

Role of Rubredoxin in F_X Assembly in PS I-- The ability of the rubA mutant to grow photoheterotrophically on glycerol with nitrate as nitrogen source implies that other iron-sulfur proteins required for growth and metabolism are synthesized normally. Were RubA to be involved in generalized iron mobilization or general iron-sulfur cluster biosynthesis, then one would expect that the mutant would show a severely diminished growth rate, which was not found in the rubA mutant (6). Nevertheless, it has recently been shown that rubredoxin-2 can reduce an unusual bacterioferritin in Desulfovibrio desulphuricans (32). This observation suggests that rubredoxins may play multiple roles in iron metabolism and oxygen detoxification in anaerobes. Perhaps the large amount of iron required for the abundant PS I complex requires a high electron throughput to liberate iron from a suitable storage protein. Were this to be the case, then RubA would be involved in redox chemistry that would supply iron for iron-sulfur cluster biosynthesis in PS I.

An alternative hypothesis is that RubA might be directly involved in the biosynthesis of the F_X iron-sulfur cluster in PS I. Although experimental data on the temporal assembly of the electron transfer cofactors P700, A₀, A₁, and F_X are lacking, we propose that the insertion of the F_X cluster occurs after the assembly of the membrane-embedded polypeptides PsaA and PsaB but prior to the attachment of the stromal polypeptides, PsaC, PsaD, and PsaE. The interpolypeptide iron-sulfur cluster F_X is ligated by four cysteines from the highly conserved h-i loop regions of PsaA and PsaB. This binding by the cysteines requires precise positioning of the two loop regions relative to one another, which in turn requires that the transmembrane helices bridged by the loops have the correct orientation. Since these helices interact with helices from both the amino-terminal domain that carry the antenna chlorophylls as well as the carboxyl-terminal domain that carry the electron transfer cofactors, it is reasonable to propose that the PsaA/PsaB heterodimer is in a physiologically mature conformation in the membrane prior to the insertion of the F_X cluster. It is also reasonable that the chlorophylls, carotenoids, and lipids would be incorporated into their binding sites at this stage of reaction center development. In this scenario, the F_X cluster would be one of the last cofactors to be assembled on the PsaA/PsaB heterodimer. The formation of trimers is likely to be the last step in the process, since mutants unable to form trimers are completely functional in electron transfer (33, 34).

The biogenesis of PS I occurs in growing and dividing cyanobacterial cells. If a fully developed P700-A₁ core complex is present prior in the insertion of F_X, then the reaction center will be partly functional, undergoing photochemical charge separation and recombination between P700 and A₁ in the light. The assembly and/or insertion of the F_X cluster must therefore occur in an environment in which a reductant with an E_m below about 700 mV (A₁) is within electron transfer distance. Given the likely need to maintain tightly coupled redox chemistry in the proposed scaffolding protein that carries out the mechanistically complicated biosynthesis of an iron-sulfur cluster (35, 36), we propose an alternative hypothesis that the diversion of the electron from A₁ may be necessary to prevent overreduction of the transient cluster during biogenesis. A possible function of RubA may therefore be to serve as an electron shunt that would divert electrons from A₁ until the labile iron-sulfur cluster can be assembled and inserted into the F_X site. These issues would not arise during the biosynthesis of the water-soluble PsaC protein or any other iron-sulfur protein, consistent with the finding that rubredoxin is involved only in F_X cluster biosynthesis. RubA differs structurally from the rubredoxins of anaerobes by having a flexible linker or hinge domain that connects the rubredoxin domain to a membrane anchor that is predicted to be a transmembrane -helix (6). The high degree of sequence conservation in these two structural motifs across large evolutionary distances (from prokaryotes to eukaryotes) argues strongly for the functional importance of these two structural motifs. If the function of RubA is to intercept electrons from A₁ during F_X cluster biogenesis, then the rubredoxin domain would necessarily have to be bound to PS I monomers in close proximity to the A₁ binding site. Movement of the rubredoxin domain could be facilitated by the flexible hinge domain in such a model.

In characterizing a closely related rubredoxin from the cryptomonad Guillardia theta Wastl et al. (37, 38) recently reported that the eukaryotic homolog of RubA is associated with PS II complexes in BBY and PS II core complexes isolated from the grana membranes of spinach chloroplast thylakoids. These authors identified a g = 4.3 resonance by EPR spectroscopy of a fraction enriched in PS II core dimers that was attributed to rubredoxin. However, it is clear from our studies that, regardless of the subcellular localization of rubredoxin in chloroplast thylakoids, RubA in the prokaryotic cyanobacterium Synechococcus sp. PCC 7002 is ultimately involved with iron-sulfur cluster assembly in PS I and is clearly not associated structurally or functionally with PS II. The very high sequence similarity between the cyanobacterial RubA proteins and their eukaryotic homologs likewise implies a conserved function.

Application of the rubA Mutant in PS I Research-- The trimeric P700-A₁ core isolated from the rubA mutant is ideal for spectroscopic studies of the early events in photochemical charge separation and charge recombination. While it is possible to isolate a P700-A₁ core chemically from wild-type PS I complexes, the procedure utilizes a multistep process, involving first the removal of PsaC, PsaD, and PsaE stromal subunits using 6.8 M urea, followed by the removal of the F_X cluster using 2 M urea and 5 mM potassium ferricyanide (summarized in Ref. 8). The 6.8 M urea step represents a careful balance between the removal of the stromal polypeptides and the oxidative denaturation F_X, and the 3 M urea step represents an equally careful balance between the oxidative denaturation of F_X and the detergent-induced removal of phylloquinone (see Fig. 8B). The use of the rubA mutant avoids these difficulties by allowing the isolation of a fully intact P700-A₁ core that can be used directly for spectroscopic characterization of the quinone intermediate. Alternatively, this complex can be used as the starting material for further resolution or rebinding studies of the cofactors and subunits of PS I.

One of the most important issues that can be addressed in this way is the route of electron transfer in PS I among the pseudosymmetrically located set of cofactors. Because the two branches of cofactors in PS I converge at F_X, it is impossible to determine which pathway is used after electron transfer to F_X has occurred. In the rubA mutant, electron transfer to F_X does not occur, and thus the time window available to determine which path an electron has taken is extended by roughly 2 orders of magnitude. In this regard, it is important to note that recent mutation studies of Chlamydomonas reinhardtii (39) have shown that at low temperature the quinone bound to the PsaA protein subunit is observed by EPR. In these samples, irreversible charge separation occurs in a significant fraction (~) of the reaction centers, but it is unknown which quinone is involved in the electron transfer in this fraction. In the rubA mutant, there is no irreversible charge separation, and the entire population of A₁ is observed by EPR. The fact that the spin-polarized EPR spectra of the rubA mutant shown in Fig. 6 are virtually identical to the corresponding wild-type spectra points toward preferential (and possibly unidirectional) transfer along the PsaA branch at low temperature. This conclusion is also supported by calculations (data not shown) that indicate that the differences between the possible P⁺A₁ spectra are substantial at Q- and W-band. Point mutation studies in the rubA background will help to confirm this result. In addition, the rubA mutant would also be useful to test the conclusions regarding bidirectionality obtained (14) using the spin polarization decay of the P⁺A₁ state, which is obviously dominated by uncontrolled spin relaxation from reduced iron-sulfur centers. These spin relaxation contributions would be absent in the rubA mutant.

	ACKNOWLEDGEMENT

We thank Dr. Wolfgang Lubitz for generously allowing use of the EPR facility.

	FOOTNOTES

^* This study was supported by National Science Foundation Grants MCB-9723469 and MCB-0077586 (to D. A. B.), United States Department of Agriculture Award 2001-35318-10125 (to J. H. G.), the Deutsche Forschungsgemeinschaft (SFB 498, TP A3, and SPP "High Field EPR"), Fonds der Chemischen Industrie (to R. B.), the Natural Sciences and Engineering Research Council, the Canada Foundation for Innovation and The Ontario Innovation Trust (to A. v. d. E.), the Humboldt Foundation (to S. G. Z.), and China State Key Project G19980101 (to J. Z.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: S-310 Frear Bldg., Pennsylvania State University, University Park, PA 16802. Tel.: 814-865-1662; Fax: