Electron Transfer and Stability of the Cytochromeb 6 f Complex in a Small Domain Deletion Mutant of Cytochrome f *

The lumen segment of cytochrome fconsists of a small and a large domain. The role of the small domain in the biogenesis and stability of the cytochromeb 6 f complex and electron transfer through the cytochrome b 6 f complex was studied with a small domain deletion mutant in Chlamydomonas reinhardtii. The mutant is able to grow photoautotrophically but with a slower rate than the wild type strain. The heme group is covalently attached to the polypeptide, and the visible absorption spectrum of the mutant protein is identical to that of the native protein. The kinetics of electron transfer in the mutant were measured by flash kinetic spectroscopy. Our results show that the rate for the oxidation of cytochrome f was unchanged (t 1 2 = ∼100 μs), but the half-time for the reduction of cytochrome f is increased (t 1 2 = 32 ms; for wild type,t 1 2 = 2.1 ms). Cytochromeb 6 reduction was slower than that of the wild type by a factor of approximately 2 (t 1 2 = 8.6 ms; for wild type, t 1 2 = 4.7 ms); the slow phase of the electrochromic band shift also displayed a slower kinetics (t 1 2 = 5.5 ms; for wild type,t 1 2 = 2.7 ms). The stability of the cytochromeb 6 f complex in the mutant was examined by following the kinetics of the degradation of the individual subunits after inhibiting protein synthesis in the chloroplast. The results indicate that the cytochromeb 6 f complex in the small domain deletion mutant is less stable than in the wild type. We conclude that the small domain is not essential for the biogenesis of cytochromef and the cytochromeb 6 f complex. However, it does have a role in electron transfer through the cytochromeb 6 f complex and contributes to the stability of the complex.

Functioning as a plastoquinone-plastocyanin oxidoreductase, the cytochrome b 6 f complex transfers electrons from Photosystem II to Photosystem I in the photosynthetic electron transfer chain of all oxygen-evolving organisms. Cytochrome f is one of the four redox centers and the largest subunit (31 kDa) in the cytochrome b 6 f complex. It transfers electrons from the Rieske FeS 1 protein to plastocyanin. The biogenesis of cytochrome f and the assembly of the cytochrome b 6 f complex involve a complicated process (1,2): starting from the synthesis of the cytochrome f precursor to the translocation of this precursor through the thylakoid membrane, the processing of the precursor to give a mature protein, the covalent attachment of heme to the apoprotein, and the assembly of cytochrome f with the other subunits to form a functional cytochrome b 6 f complex. Cytochrome f is required for the assembly of the cytochrome b 6 f complex because when the petA gene encoding cytochrome f in the chloroplast genome was disrupted, there was no assembled cytochrome b 6 f complex, and the resulting mutant could not grow photoautotrophically (3,14). Biochemical evidence (4) and the three-dimensional structure (5)(6)(7)(8) revealed that there are three domains in cytochrome f: a C-terminal transmembrane span and a small and a large domain. Serving as a transmembrane span, the C-terminal hydrophobic region of cytochrome f (from Gln 253 to Phe 286 in Chlamydomonas reinhardtii) (9 -11) anchors the protein to the thylakoid membrane. Deletion of this region gave rise to a soluble form of cytochrome f (12) that was transported to the lumen and was capable of transferring electrons to plastocyanin, but again there was no assembled cytochrome b 6 f complex (12). Localized in the lumen, the soluble segment has an elongated structure, with a small and a large domain (Refs. 5-8; see Fig. 1). A lysine patch was found in cytochrome f (5-7) from higher plants and C. reinhardtii and was postulated to be involved in the interaction with plastocyanin (5). Mutations within the large domain generated interference with the electron transfer (13)(14)(15)(16)(17)(18)(19)(20)(21), the proton translocation (18), and the maturation of cytochrome f (13,19). When mutants were isolated in which heme attachment was altered in the large domain of cytochrome f, the assembly of the cytochrome b 6 f complex and the insertion of the heme group were disrupted (13). Although these results shed light on the functioning of C-terminal transmembrane span and the large domain of cytochrome f, the role of the small domain remained ambiguous.
The small domain of cytochrome f consists of amino acid residues from Asn 172 to Leu 228 (9 -11), approximately the difference in size between the lumen-soluble segment of cytochrome f and the extramembrane domain of mitochondrial cytochrome c 1 (5). In comparing the primary sequences of cytochrome f from other organisms, the small domain of cytochrome f is less conserved than the large domain (4). Only a few amino acid residues within its sequence are identical among all of the sequences found. From the crystal structure of cytochrome f (5-8), we can see that the backbone of the small domain is distinguishable from the large domain (Fig. 1). Mutations of amino acid residues within the small domain did not reveal any significant difference from the wild type protein in vivo (15)(16)(17). These analyses raised the question about the role of the small domain of cytochrome f in the biogenesis and functioning of the cytochrome b 6 f complex. In a comparison of the structure of the cytochrome b 6 f complex (5-8, 22, 23) with the analogous cytochrome bc 1 complex from mitochondria (24 -26), Soriano et al. (27) speculated that the small domain of cytochrome f might function like the subunit VIII in the cytochrome bc 1 complex to provide structural stability to the complex. Recently, the crystal structure of cytochrome f from C. reinhardtii revealed that cytochrome f could exist in dimeric form (7), and amino acid residues from the small domain are found to be involved in the hydrogen bond formation in the interface of this dimer. However, the relationship of this dimer to the structure of cytochrome f in vivo is not clear.
We were interested in defining the role of the small domain of cytochrome f in the biogenesis and the functioning of cytochrome f and the cytochrome b 6 f complex. To address this question, we have made a small domain deletion mutant of cytochrome f in C. reinhardtii. The mutant is able to grow photoautotrophically and has an assembled cytochrome b 6 f complex. Our studies showed that the electron transfer through the cytochrome b 6 f complex can occur in the mutant but at a slower rate. In addition, the mutant cytochrome b 6 f complex was not as stable as that in the wild type strain.

EXPERIMENTAL PROCEDURES
Strains and Culture Conditions-C. reinhardtii wild type (cc-125), cytochrome f small domain deletion mutant (pFSSD), and ⌬petA mutant (14) strains were maintained on TAP plates (28) under dim light (2 mol of photons m Ϫ2 s Ϫ1 ). For the photoautotrophic growth of strains, the wild type and the pFSSD mutant were grown on HS plates (28) under medium light (35 mol of photons m Ϫ2 s Ϫ1 ). Liquid cultures of the wild type strain and the pFSSD mutant were carried out in HS medium bubbled with 3% CO 2 under 35 mol of photons m Ϫ2 s Ϫ1 light intensity. Bacteria Escherichia coli DH5␣ strain was used for DNA manipulations.
Construction of Cytochrome f Small Domain Deletion Mutant-PCR was performed as described previously (20) to make the construct (Fig.  2). The following are the sequences of the following primers: primer I, 5Ј-CAACTGGAATCCCCTTATAG-3Ј (located 50 bp upstream of petA gene); primer II, 5Ј-CGCTACGTAAATAGTGTTGTTTGA-3Ј (the bold codon encodes Tyr 171 ); primer III, 5Ј-GCCTACGTAACAAACAACAAC-CCTAACGTTGG-3Ј (the bold codon encodes Thr 229 ); and primer IV, 5Ј-GTAGGAGCTGCACAGCAGCC-3Ј (located 13 bp downstream from the petA gene). The underlined sequence is the SnaBI restriction site. The PCR template was the plasmid pJB101⌬H containing the petA gene (14). A 700-bp fragment I was obtained by PCR using primers I and II. A new SnaBI site was introduced at the 3Ј-end of the fragment. Fragment II with a length of 220 bp was obtained by PCR using primers III and IV. A new SnaBI site was also added to the 5Ј-end of fragment II. Fragment I was cut with HindIII and cloned into pUC19 plasmid cut with HindIII and HincII to give pUCFI. Fragment II was cloned into pUC19 cut with HincII to give pUCFII. For the pUCFII, we selected the construct with an orientation where the SnaBI site is adjacent to the HindIII site from the pUC19 polylinker region. The correct sequences of the inserts were confirmed by sequencing in an Applied Biosystems DNA sequencer 377. pUCFII was then digested with SnaBI and EcoRI to give a 220-bp fragment, which was ligated with pUCFI cut with the same enzyme to give pUCFSD. pUCFSD was digested with HindIII and AflII. The insertion was cloned into pJB101⌬H plasmid digested with the same restriction enzyme to give pJBFSD. To make the chloroplast transformation construct pFSSD, the petA gene sequence from pADFI283ST (3) was replaced with small domain deletion petA mutant sequence from pJBFSD by digesting both plasmids with BglII and EcoRV. In the pFSSD construct the sequence of petA gene, which encodes a small domain of cytochrome f from Gln 172 to Leu 228 , was deleted (57-amino acid sequence). An extra codon for valine was added in the petA gene sequence of the small domain deletion mutant between the codon for Tyr 171 and the codon for Thr 229 that resulted from the addition of a new SnaBI site.
Chloroplast Transformation of C. reinhardtii ⌬petA Strain-The ⌬petA strain was transformed as described previously (14) using the biolistic particle delivery system (PDS-1000/He; Bio-Rad). 250 ml of TAP medium was inoculated with the ⌬petA strain and grown for 3 days under dim light. 600 ml of fresh TAP medium was inoculated to ϳ1 ϫ 10 5 cells/ml. Cells grown to ϳ2 ϫ 10 6 cells/ml in the presence of 0.5 mM fluorodeoxyuridine were harvested by centrifugation and resuspended to a concentration of ϳ1 ϫ 10 8 cells/ml. 1 ml of the cell suspension was mixed with an equal volume of (premelted and incubated at 42°C) 0.2% agar in TAP, and 0.7 of the mixture was spread onto two preprepared TAP plates. 4 g of pFSSD DNA was precipitated onto 0.3 mg of gold particles, which were used to bombard the cells on the plates. The bombarded cells were transferred to TAP plates containing 150 g/ml spectinomycin and incubated under dim light at 25°C. PCR was performed to select the transformants as described previously by Berthold et al. (29).
Southern Blotting, Western Blotting, and Heme Staining-C. reinhardtii DNA was prepared by centrifuging cells from 50 ml of culture. The cells were resuspended in 500 l of 2% hexadecyltrimethylammonium bromide, 100 mM Tris-Cl, pH 8, 1.4 M NaCl, 20 mM EDTA, 2% ␤-mercaptoethanol and incubated at 65°C for 60 min. The solution was extracted three times with phenol/chloroform/isoamyl alcohol (24:24:1 v/v/v). The DNA was precipitated with 0.7 volumes of isopropyl alcohol. After digesting the DNA with EcoRV and HindIII, Southern blot hybridization was carried out as described previously (14).
Prior to electrophoresis for Western blotting, protein samples were prepared according to Berthold et al. (30). The protein samples were separated on SDS-polyacrylamide gel with a 15% resolving/5% stacking gel and transferred to nylon membranes. Western blotting was carried out following manufacturer's protocol (ECL Western blotting; Amersham Pharmacia Biotech). Rabbit sera containing polyclonal antibodies generated against spinach cytochrome f, spinach subunit IV, C. reinhardtii Rieske FeS protein (generously provided by Dr. C. de Vitry, Institut de Biologie Physico-Chimique, France), and a synthetic peptide conjugate (corresponding to C. reinhardtii cytochrome b 6 , generously provided by Dr. W. Cramer, Purdue University) were used at a 1:10,000 dilution. Heme peroxidase activity was detected with N,N,NЈ,NЈ-tetramethylbenzidine and H 2 O 2 (31) Oxygen Evolution Measurements-The rate of oxygen evolution was measured at 25°C in an Oxygraph System (Hansatech) according to the manufacturer's instructions. C. reinhardtii cells were resuspended at a chlorophyll concentration of 10 g/ml in HS medium supplemented with 10 mM sodium bicarbonate (14).
Flash Kinetic Measurements-Kinetic measurements were performed on autolysin-treated cells to form a homogenous single-cell suspension, which also allows easy access of inhibitors and uncouplers. Autolysin was prepared as described (28). The pellet of C. reinhardtii cells was resuspended with autolysin solution and incubated at room temperature for 15 min. Cells were then washed twice with HS liquid medium. A home-built single beam kinetic spectrophotometer with microsecond time resolution was used as described previously (14). Flash-induced spectroscopy was done at 25°C under anaerobic conditions maintained by argon flux. Cells were suspended in HS medium, pH 6.8 (28), at a chlorophyll concentration of 30 g/ml. A short 23% P700-saturating flash, having a duration of 3.5 s at half-peak height, was used to avoid multiple turnovers of the cytochrome b 6 f complex. Cytochrome f was monitored as ⌬A 554 -545 nm (14) in the presence of 30 M FCCP. For the measurement of cytochrome f oxidation, 22 M stigmatellin was also added to inhibit the reduction of cytochrome f. Cytochrome b 6 was measured as ⌬A 564 -575 nm (15) in the presence of 30 M HQNO and 30 M FCCP. The slow phase of the electrochromic band shift was measured as the difference of ⌬A 515 nm (18) in the absence and the presence of 22 M stigmatellin. The kinetic data were fit as first order reactions to give the rate constants, which were used to calculate the half-times of the reactions.

Characterization of the Cytochrome f Small Domain Deletion
Mutant-DNA was isolated from the primary transformants, which were able to grow on the TAP plates containing specti-nomycin. The colonies giving the correct size of the DNA fragment by PCR were transferred to HS plates and incubated under dim light at 25°C. Most of the colonies died after 3-4 weeks. Only a few tiny green colonies were found under the microscope. They were restreaked onto the fresh HS plates and TAP plates. After a few weeks it was confirmed that pFSSD mutant was able to grow photoautotrophically. The growth rate measured in HS liquid medium under 35 mol of photons m Ϫ2 s Ϫ1 showed that the pFSSD mutant grew slightly more slowly than the wild type (Table I). Under higher light intensity, i.e. 160 mol of photons m Ϫ2 s Ϫ1 , the slower growth in the pFSSD mutant is more noticeable (doubling time, 9.8 h versus 8 h). Southern blot analysis revealed the presence of the correct insertion in these cells (Fig. 3). In comparison with the wild type strain, the smaller band from the pFSSD mutant showed the correct size of deletion (ϳ0.16 kilobase) from the petA gene sequence. The chlorophyll contents of the wild type and pFSSD mutant are very similar. The chlorophyll (a ϩ b)/cell is 3.8 Ϯ 0.5 ϫ 10 Ϫ15 and 3.2 Ϯ 0.3 ϫ 10 Ϫ15 mol/cell, respectively. The chlorophyll a/chlorophyll b ratio is 2.5 and 2.3, respectively. The presence of the mutant polypeptide was then examined by immunoblotting (Fig. 4). The anti-cytochrome f antibodies identified a band of smaller molecular mass (ϳ25 kDa) in the pFSSD mutant, which is approximately the expected molecular mass for cytochrome f after the deletion of the small domain. The antibodies against cytochrome b 6 , subunit IV, and Rieske FeS protein also identified bands of the same size as the wild type from the pFSSD mutant, which are in contrast to the petA deletion strain where no detection of any of the subunits of the cytochrome b 6 f complex is observed (Fig. 4, lane 2). It is known that the absence of cytochrome f results in a rapid degradation of the cytochrome b 6 f complex subunits (3). Therefore the presence of other subunits in the pFSSD mutant suggests that the cytochrome b 6 f complex was assembled. However, the level of these proteins in the mutant cells was only 40 -30% of the level in the wild type cells, showing that there was less cytochrome b 6 f complex in the mutant. Detection of peroxidase activity using tetramethylbenzidine and H 2 O 2 also indicated the presence of a new 25-kDa heme-containing protein, suggesting that heme had been covalently inserted into the mutant polypeptide of cytochrome f. We conclude from these studies that deletion of the small domain does not affect the translocation of the protein to the lumen, the processing of the precursor protein, and the incorporation of the heme group into the polypeptide and that the small domain is not essential for the assembly of the cytochrome b 6 f complex. Fig. 5 shows the light saturation curves of photosynthesis for the mutant and wild type cells. The quantum yield of O 2 evolution as indicated by the initial slope of the curves is approximately the same for both wild type and the pFSSD mutant. However, the pFSSD mutant showed a light-saturated rate of total O 2 evolution approximately 5 times lower than that in the wild type. These results suggest that the electron transfer between Photosystems II and I is limiting in the mutant under high light conditions. Electron Transfer in Cytochrome f Small Domain Deletion Mutant-To examine the electron transfer properties of the pFSSD mutant, cells were grown in the HS medium until they reached the mid-exponential phase. After centrifugation, the cells were treated with autolysin. We found that after the autolysin treatment the aggregated cells were dispersed into single cells and that the inhibitors and uncouplers, such as HQNO, went into the cells more readily because the cell wall was removed.
It was considered that the deletion of the small domain from cytochrome f might change the redox differential spectrum of the ␣-band of cytochrome f. To examine this, a time-resolved oxidized minus reduced spectrum of cytochrome f was measured within the range of 540 -575 nm for the wild type and the pFSSD mutant. The absorbance maxima of the ␣-band of cytochrome f was found unchanged at 554 nm (Fig. 6). The general shapes of the spectra were identical between the wild type and the pFSSD mutant ( Fig. 6 and inset). This suggests that the heme environment in the pFSSD mutant is similar to that of the wild type.
Cytochrome f oxidation and reduction were monitored by following the difference between flash-induced absorbance changes at 554 and 545 nm. For measuring the oxidation of cytochrome f, stigmatellin was added to inhibit the reduction of cytochrome f and eliminate the interference from the reduction (14). As shown in Table I and Fig. 7 (traces a and b), the half-time (t1 ⁄2 ) of the flash-induced oxidation of cytochrome f in the pFSSD mutant was approximately the same as that of the wild type (ϳ100 s). The same results were obtained by using a 5-ms time scale and an instrumental time constant of 10 s (data not shown). However, the pFSSD mutant had a t1 ⁄2 of cytochrome f reduction at 32 ms, ϳ15 times larger than that of the wild type (Table I and Fig. 7, traces c and d). This in vivo result indicates that the small domain was required for the reduction but not for the oxidation of cytochrome f. Because we were using a low intensity flash (23% P700-saturating), which would avoid multiple turnovers of the cytochrome b 6 f complex,  3 and 4) and was digested by EcoRV and HindIII. Southern blot hybridization was performed using a probe of about 600 bp, which was prepared by digesting the petA gene with AccI and HindIII. the amplitude of the ⌬A 554 -545 nm changes are of similar magnitude in both wild type and the pFSSD mutant. Under a high intensity flash (75% P700-saturating), less photooxidizable cytochrome f in the pFSSD mutant than that in the wild type was found (ϳ2-3-fold less; data not shown), in agreement with the lower levels of the individual cytochrome b 6 f subunits detected by immunoblotting in the pFSSD mutant.
The kinetics of cytochrome b 6 photoreduction were measured by following the difference between flash-induced absorbance changes at 564 and 574 nm. HQNO was added to inhibit the oxidation of cytochrome b 6 by plastoquinone (Fig. 7, traces e and f). In contrast to the reduction of cytochrome f, the halftime for the reduction of cytochrome b 6 was less affected in the mutant, which was 8.6 ms, approximately 2-fold larger than that of the wild type (4.7 ms) ( Table I). This rate of cytochrome b 6 reduction in the pFSSD mutant was faster than the rate of the reduction of cytochrome f by a factor of 3.7 (t1 ⁄2 8.6 ms versus t1 ⁄2 32 ms). In the mutant the amplitude of cytochrome b 6 reduction is smaller by a factor of 1.8 than the amplitude of cytochrome f oxidation, whereas this ratio of amplitudes is close to 1 in the wild type strain. However, under a high intensity flash (75% P700-saturating) both wild type and pFSSD mutant showed a similar ratio (ϳ1:1) between the amplitudes of cytochrome b reduced and cytochrome f oxidized (data not shown).
The slow phase of the carotenoid electrochromic band shift is believed to serve as an indicator for the charge separation across the thylakoid membrane because of the operation of the cytochrome b 6 f complex. The charge separation would include the electron movement from the b L heme to the b H heme in cytochrome b 6 and the proton translocation from the stroma to the plastoquinone reduction site, known as Qi (32,33). The electrochromic band shift was followed by the absorbance change at 515 nm. The slow phase associated with the electrogenic reactions in the cytochrome b 6 f complex is shown as the difference of the traces at 515 nm in the absence and the presence of stigmatellin, which inhibits the cytochrome b 6 f electrogenic activities by blocking the plastoquinol oxidation site (Qo). Fig. 7 (traces g and h) shows that the slow phase of the electrochromic signal in the pFSSD mutant is slower by a factor of 2 as compared with the wild type (Table I). This suggests that the deletion of the small domain reduced the rate of proton translocation because the electron movement between two cytochrome b hemes is probably intact.
The Stability of the Cytochrome b 6 f Complex-From the Western blotting and heme staining gel (Fig. 4), it was obvious that the amount of the cytochrome b 6 f complex in the pFSSD mutant was markedly reduced to 30 -40% of the level found in the wild type strain. This was also confirmed by flash kinetic spectroscopy experiments under a high intensity flash, in which the amplitude of absorbance changes of cytochrome f and cytochrome b in the pFSSD mutant was always 2-3-fold smaller as compared with the wild type (data not shown). The question arose of whether the deletion of the small domain from cytochrome f resulted in an unstable cytochrome b 6 f complex. To examine the stability of the cytochrome b 6 f complex in the pFSSD mutant with respect to the wild type, the rates of degradation of the cytochrome b 6 f complex were studied in the presence of chloramphenicol, an inhibitor of protein synthesis in the chloroplast. Aliquots of the cell culture were removed after the addition of chloramphenicol at different times over a 21-h period, and Western blotting was performed for cytochrome f, cytochrome b, Rieske subunit, and subunit IV. Fig. 8 shows that in the pFSSD mutant the cytochrome b 6 f complex subunits, initially present in lower amounts (as in Fig. 4), undergo a faster degradation than those in the wild type strain. The half-times for the degradation of these subunits are decreased to about 3-6 h (Table II) cilitates the degradation of its subunits, and gives rise to a reduced level of the cytochrome b 6 f complex in the pFSSD mutant (Fig. 4, lane 3, and Fig. 8B, lane 0).

DISCUSSION
The Biogenesis of the Cytochrome b 6 f Complex-Cytochrome f is synthesized in the chloroplast stroma and is then inserted into the thylakoid membrane with the bulk of the protein being localized in the lumen. Because of the unique structure of cytochrome f, in which the Tyr 1 provides the sixth ligand to the heme iron, the processing of cytochrome f and heme attachment are likely to be closely coupled (5)(6)(7)(8). Our results with the pFSSD mutant showed no role for the small domain in these processes. How the heme is inserted into cytochrome f in the chloroplast is still an open question (34,35). If the insertion of the heme group into cytochrome f requires a heme lyase, which was identified in the mitochondria (36), this lyase does not require the small domain of cytochrome f for carrying out its activities. The cytochrome b 6 f complex is assembled in the pFSSD mutant, and at low and medium light intensity, the mutant strain grows photoautotrophically at a slightly slower rate than the wild type. It is clear from these results that the small domain of cytochrome f is not playing an important role in the biogenesis of cytochrome f and the cytochrome b 6 f complex.
Under low light condition, where the limiting factor is the photoactivation frequency of photosystem I and II, the rate of oxygen evolution in the pFSSD mutant is indistinguishable from that of the wild type. However, under higher light intensities, where other limiting factors dominate, the light-saturated rate of O 2 evolution is considerably decreased, pointing to an impairment in the linear electron transfer chain. This is probably due to a combined effect of the lower content of cytochrome b 6 f complex and the slower electron transfer kinetics in the pFSSD mutant.
Electron Transfer through the Cytochrome b 6 f Complex-Electron transfer through the cytochrome b 6 f complex in the pFSSD mutant takes place in a manner different from that in the wild type. Because the oxidation of cytochrome f by plastocyanin is not affected in the intact mutant cell, this indicates that the small domain is not involved in the interaction with plastocyanin. This result is consistent with in vivo studies on the site-directed mutants involving Lys 188 and Lys 189 in the small domain (15)(16). However, observations from experiments with in vitro systems yielded different conclusions. The results from the electron transfer experiments carried out in vitro (17,20,21) and in nebulized cells (16) and from molecular dynamic modeling (37)(38)(39) and the solution structure of the complex between cytochrome f and plastocyanin (40,41) provided evidence that the lysine residues from the small domain do interact with plastocyanin. This discrepancy between in vivo and in vitro behaviors remains unresolved. Nevertheless, the reduction rate of cytochrome f in the pFSSD mutant is slower than that in the wild type, as are the reduction of cytochrome b 6 and the slow phase of the electrochromic band shift. This is similar to the effect observed when the amino acid residues involved in the formation of an internal water chain were altered (18). We noticed that the half-time for the reduction of cytochrome b L is ϳ2-fold larger than that for the slow phase of electrochromic band shift both in the wild type and the pFSSD mutant (Table I). According to the Q cycle model, these two rates should be the same, or the rate of the slow phase of electrochromic band shift should be slower than that of the reduction of cytochrome b 6 . We do not discuss the relationship of these two rates here because these two reactions are measured under different conditions, i.e. in the absence and the presence of FCCP and HQNO, respectively. Further studies are under way to understand the deviation of these two reactions. The oxidation of plastoquinol at the Qo site is generally considered to be nonelectrogenic (32,33), whereas the reduction of plastoquinone at the Qi site is one of the processes contributing to the slow rise of electrochromic band shift. The question is then how proton translocation from the stroma site is affected while the mutation is localized in a region of cytochrome f that is in the lumen. Under the anaerobic conditions in our experiments the plastoquinol/plastoquinone couple and cytochrome b H are in their reduced form before the flash (42), and the only plastoquinone available after the flash is generated from plastoquinol oxidation at Qo site. The cytochrome b 6 reduction signal comes from b L (42), which showed that in the pFSSD mutant the half-time for the reduction of the b L heme is increased. The slower reduction of b L indicates a slower oxidation of plastoquinol at the Qo site, followed by a slower release of plastoquinone from the Qo site, resulting in a slower proton translocation at the Qi site. But this interpretation is not sufficient to explain the amplitude of the slow phase of the electrochromic band shift in the pFSSD mutant, which is very similar to the one in the wild type (Fig. 7, traces g and h) and is not affected by the 1.8-fold decreased amplitude of cytochrome b reduction (Fig. 7, traces e and f). This is unexpected if the electrogenicity would be only due to an electron movement through the low potential chain and a proton movement from the stroma to the Qi site. Instead, a smaller amplitude of cytochrome b 6 reduction would result, by the same factor, in a decrease both in electron movement through the low potential chain and in the provision of plastoquinone from the Qo site to the Qi site. However, the additional unexpected electrogenicity could be explained by an additional proton channel driven by the high potential chain proposed by Joliot and Joliot (43).
We also observed the loss of the concerted reduction of cytochrome f and b 6 in the pFSSD mutant because the reduction of

TABLE II
The half-time of the degradation of subunits of cytochrome b 6 f complex The rates of the degradation of the cytochrome b 6 f complex were measured using immunoblotting as shown in Fig. 8. The intensities of the individual bands were quantified by NIH Image software. The range represents the variation of data collected from two experiments. cytochrome b 6 was much faster than the reduction of cytochrome f. A similar observation was also reported by Ponamarev and Cramer (18) in cytochrome f mutants and earlier by others in studies of the wild type cytochrome b 6 f complex (44). We believe this uncoupling might arise from the electron transfer bypassing cytochrome f under certain conditions (45). The Stability of the Cytochrome b 6 f Complex-The stability of the cytochrome b 6 f complex is affected in the pFSSD mutant where the complex is degraded more rapidly than in the wild type strain. This is consistent with the postulation that the small domain may provide structural stability for the complex (27). However, unlike subunit VIII in the cytochrome bc 1 complex (46), the cytochrome b 6 f complex is still assembled, and the deletion of the small domain of cytochrome f did not abolish the complex completely. The lower level of the cytochrome b 6 f complex in the pFSSD mutant is probably due to a faster degradation of this complex because the rates of the synthesis of cytochrome b 6 and subunit IV are the same as the wild type even in the absence of cytochrome f (3). The faster turnover of the cytochrome b 6 f complex may result from faster degradation of its subunits by protease. As has been shown CIpP, an ATP-dependent protease that has been found in the stroma (47), is able to degrade the assembled form of the cytochrome b 6 f complex (48).
The cytochrome b 6 f complex may exist as a dimer in vivo (49 -51). Interestingly, cytochrome f can also form dimeric associations in the crystal (7), in which amino acid residues from the small domain of one monomer make close contact with residues from another monomer. In either case of cytochrome f forming a dimeric association in a cytochrome b 6 f complex dimer or interacting with other subunits of the complex, the deletion of the small domain may have eliminated part of the interaction between two cytochrome f monomers in the dimeric form of the cytochrome b 6 f complex or between cytochrome f and other subunits of the complex, resulting in a complex which may be "looser" and more prone to the degradation by protease.
In summary, our studies on a small domain deletion mutant of cytochrome f have shown that the small domain is not required for the biogenesis of cytochrome f and the cytochrome b 6 f complex. In the small domain deletion mutant of cytochrome f, electron transfer through the cytochrome b 6 f complex occurs but at a slower rate than that in the wild type strain. The small domain of cytochrome f is also important in stabilizing the cytochrome b 6 f complex.