The EMBO Journal vol.15 no.14 pp.3498-3506, 1996

The chloroplast ycf7 (petL) open reading frame of Chiamydomonas reinhardtii encodes a small functionally important subunit of the cytochrome b6f complex Yuichiro Takahashi1l2, Michele Rahire2, Cecile Breyton3, Jean-Luc Popot3, Pierre Joliot3 and Jean-David Rochaix2'4 'Department of Biology, Faculty of Science, Okayama University, 3-1-1 Tsushima-naka, Okayama 700, Japan, 2Departments of Molecular Biology and Plant Biology, University of Geneva, 30 quai Emest-Ansermet, CH-1211 Geneva 4, Switzerland and 3Institut de Biologie Physico-Chimique and College de France, CNRS URA1 187, 13 rue Pierre-et-Marie-Curie, F-75005 Paris, France 4Corresponding author

The small chloroplast open reading frame ORF43 (ycf 7) of the green unicellular alga Chlamydomonas reinhardtii is cotranscribed with the psaC gene and ORF58. While ORF58 has been found only in the chloroplast genome of C.reinhardtii, ycf 7 has been conserved in land plants and its sequence suggests that its product is a hydrophobic protein with a single transmembrane a helix. We have disrupted 0RF58 and ycf 7 with the aadA expression cassette by particlegun mediated chloroplast transformation. While the ORF58::aadA transformants are indistinguishable from wild type, photoautotrophic growth of the ycf 7::aadA transformants is considerably impaired. In these mutant cells, the amount of cytochrome b6f complex is reduced to 25-50% of wild-type level in mid-exponential phase, and the rate of transmembrane electron transfer per b6f complex measured in vivo under saturating light is three to four times slower than in wild type. Under subsaturating light conditions, the rate of the electron transfer reactions within the b6f complex is reduced more strongly in the mutant than in the wild type by the proton electrochemical gradient. The ycf 7 product (Ycf 7) is absent in mutants deficient in cytochrome b6f complex and present in highly purified b6f complex from the wild-type strain. Ycf 7-less complexes appear more fragile than wildtype complexes and selectively lose the Rieske ironsulfur protein during purification. These observations indicate that Ycf 7 is an authentic subunit of the cytochrome b6f complex, which is required for its stability, accumulation and optimal efficiency. We therefore propose to rename the ycf 7 gene petL. Keywords: chloroplast open reading frame/cytochrome b6f complex/chloroplast transformationlChlamydomonas reinhardtii

Introduction The cytochrome b6f complex of the thylakoid membranes mediates both linear electron transport between photosystem II (PSII) and photosystem I (PSI) and cyclic electron transport around PSI. It comprises four large

subunits encoded by the chloroplast genes petA, petB and petD and by the nuclear gene PetC (Hauska et al., 1983; Malkin, 1992; Hope, 1993). The PetA protein (31-33 kDa) binds a c-type heme (cytochrome f) and is predicted to contain a single transmembrane a helix at its C terminus. The PetB protein (23-26 kDa) binds two b-type hemes (cytochrome b6) and probably forms four transmembrane helices. The petD gene encodes a 17-20 kDa protein, designated subunit IV, which is predicted to contain three transmembrane helices and appears to be involved in quinone binding (Doyle et al., 1989). The PetC protein, the so-called Rieske protein, contains a high potential 2Fe-2S cluster. Although the Rieske protein features a hydrophobic region at its N-terminus that could form a transmembrane ax helix, extraction experiments strongly suggest that it does not span the membrane (Breyton et al., 1994). The complete nucleotide sequences of the chloroplast genomes of six vascular plants and five algae have been determined (for review see Reardon and Price, 1995). Sequence comparisons revealed the presence of several conserved open reading frames that may encode functional polypeptides. Use of an antiserum raised against a synthetic oligopeptide predicted from one of these conserved open reading frames revealed that a small protein of -4 kDa copurifies with the cytochrome b6f complex from maize (Haley and Bogorad, 1989). This ORF has been designated petG. N-terminal sequence analysis and immunolabelling of small polypeptides present in preparations of the cytochrome b6f from the unicellular green alga Chlamydomonas reinhardtii revealed the presence of the PetG protein and identified an additional, nuclear-encoded protein of -4 kDa, first designated PetX (Pierre and Popot, 1993; Schmidt and Malkin, 1993), but recently renamed petM (de Vitry et al., 1996). Both PetG and PetM appear to contain a single transmembrane ax helix (Breyton et al., 1994; de Vitry et al., 1995, 1996). Their functional role is not known. Directed mutagenesis is a powerful approach for examining the functional role of chloroplast-encoded proteins. C.reinhardtii is ideally suited for this purpose: biolistic chloroplast transformation is well established in this alga (Boynton et al., 1988), and the chloroplast aadA expression cassette conferring spectinomycin resistance (Goldschmidt-Clermont, 1991) can be used for disrupting specific chloroplast genes and for selecting transformants. In the present study, we demonstrate by immunochemical as well as by biochemical means that a small subunit of the cytochrome b6f complex is encoded by the hypothetical chloroplast open reading frame 7, ycf 7, in the psaC operon of C.reinhardtii. Generation and analysis of chloroplast transformants with disrupted ycf 7 show that the Ycf 7 protein is important for photoautotrophic growth as well as for electron transfer efficiency and stability of the cytochrome b6f complex.

39©Oxford University Press 3498

Role of ycf 7 in photosynthesis

Results The psaC operon of C.reinhardtii contains ORF58 and ycf7 Two chloroplast psaC transcripts of 0.45 and 1.1 kb have been detected previously in C. reinhardtii (Takahashi et al., 1991). They have a common 5' end and the 3' ends map 120 and 750 nucleotides downstream of the psaC 3' end. Sequencing of the 1076 bp NcoI-XbaI fragment revealed the presence of two small open reading frames, ORF58 and ORF43, which are cotranscribed with psaC (Figure 1). Whereas ORF58 is not conserved in the chloroplast genomes from other plants and algae, ORF43 homologues are present in the plastid genomes of plants and correspond to the hypothetical chloroplast open reading frame ycf 7 (Figure 1). Comparison of the ycf 7 sequences reveals that the C.reinhardtii and rice genes contain upstream initiation codons that define putative extra N-terminal sequences of 12 and 6 residues respectively (Figure 1). There is little sequence homology in these N-terminal sequences and it is not known whether translation of ycf 7 RNA is initiated at the first or the second AUG. The ycf 7 product of C.reinhardtii is predicted to contain an additional valine residue near the C-terminus that is absent in its homologues in plants (Figure 1). Assuming initiation to take place at the second AUG, the predicted mass of Ycf 7 would be 3.4 kDa. Sequence identities between the Ycf 7 proteins of C.reinhardtii and of other organisms range between 53 and 58% (Figure 1). Except for its C-terminal end, the ycf 7 product consists mostly of hydrophobic amino acid residues. Its high hydrophobicity, 17 = 2.2 kcal/ residue on the GES scale of Engelman et al. (1986), suggests that it may form a transmembrane a-helix (Popot and de Vitry, 1990). The presence of basic residues at the C-terminus suggests that this end of Ycf 7 lies in the stroma (Gavel et al., 1991).

Disruption of ORF58 and ycf7 To study the functions of ORF58 and ycf 7 we disrupted these two chloroplast open reading frames with the chloroplast aadA expression cassette which confers resistance to spectinomycin (Goldschmidt-Clermont, 1991). A 4.6 kb SalI-PstI fragment containing the psaC operon was cloned into the Bluescript plasmid and the aadA cassette was inserted in both orientations at the DrallI or SnaBI restriction site to inactivate ORF58 or ycf 7 respectively (Figure 2). These plasmids were used to transform wildtype C.reinhardtii cells with a particle gun. Transformants were selected by their resistance to spectinomycin. Figure 2 shows a Southern blot analysis of DNA from wild type and the transformants. Total DNA was digested with EcoRI and PstI and hybridized with the wild-type 1.8 kb Sall-XbaI fragment. This probe hybridizes with a 5.8 kb fragment of wild-type DNA and with two fragments of the DNA of the transformants, because the aadA cassette contains a PstI site. The ORF58::aadA transformants, with the aadA cassette in the same and opposite orientations relative to the transcription of the psaC operon, gave rise to two bands of 3.8 and 3.9 kb, and 2.9 and 4.8 kb respectively. In the ycf 7::aadA transformants the corresponding fragments had sizes of 3.5 and 4.2 kb, and 3.2 and 4.5 kb respectively, as predicted from the known

psaC

ycf7

ORF58 V77m3mm

P7zEisU

100 ORF5 8

MTCTRTLLVLNRMLALKANFVLHSVKSYLNVSGVNKTLPAKGSAGVKNGFILSHLGCQ YCF7 a

b c

d e

f

MIFDFNYIHIFMLTITSYVGLLIGALVFTLGIYLGLLKVVKLI MEFN- PFMLTITSYFGFLLAALTLTLALFIGLNKI-RLI MLTITSYFGFLLAALTITSALFIGLSKI-RLI MLTITSYFGFLLAALTITPALFISLNKI-RLI MLTITSYFGFLVAAFTITSALFIGLNKI-RLI MLTIISYFLFLIGALTLALVLFIGLNKI-QLI .

..

....

....

.

.

Fig. 1. Map of the chloroplast psaC operon of C.reinhardtii comprising psaC, 0RF58 and ycf 7. The 0.45 and 1.1 kb transcripts are indicated. The sequence of the 0RF58 product is shown together with a sequence comparison of the ycf 7 protein in C.reinhardtii (a), rice (b), tobacco (c), maize (d) (Haley and Bogorad, 1989), Cuscuta reflexa (e) (L.G.Haberhausen; EMBL/GenBank; DDBJ data banks) and liverwort (f). Stars and dots indicate identical and related amino acids.

restriction map of the psaC region. No wild-type fragment of 5.8 kb was detected in these mutants, indicating that all copies of their chloroplast genome contained the aadA cassette. Hybridization of the same DNAs with an aadA probe yielded signals consistent with the structures shown in Figure 2. To avoid any polar effects on the expression of genes located downstream of the psaC operon, the transformants containing the aadA cassette transcribed in the opposite direction were used. Growth rates of the transformant and wild-type cells were compared on acetate and minimal medium (Figure 3). These spot tests revealed that the growth patterns of the ORF58::aadA transformants and wild type are the same on minimal medium. In contrast, the photoautotrophic growth of several independently isolated ycf 7:: aadA transformants was considerably impaired, suggesting that Ycf 7 is involved directly or indirectly in photosynthetic electron transfer.

ycf7 encodes a novel subunit of the cytochrome b6f complex To test whether ycf 7 is expressed as a chloroplast protein, this ORF was cloned into the expression vector pET 3a and the recombinant protein produced in Escherichia coli was used to raise antisera (see Materials and methods). Figure 4 shows an immunoblot of an SDS-polyacrylamide gel of whole cell proteins from wild type, from the control strain CT (in which the aadA cassette is inserted at the Sall site upstream of the psaC gene; see Figure 2) and from two ycf 7::aadA transformants labelled with antisera raised against cytochrome f and against the Ycf 7 protein. A polypeptide with an apparent molecular mass of 3.4 kDa was detected with the latter antiserum in the wild-type and control strains but not in the two ycf 7: :aadA transformants, indicating that the product of ycf 7 accumulates in wildtype cells. To localize the Ycf 7 protein within cells, wild-type cells were broken and fractionated by centrifugation. Ycf 7 fractionates with the pellet and is absent from the soluble fraction, as expected if it is a membrane protein (Figure SA). Upon fractionation of the pellet by discontinuous

3499

Y.Takahashi et aL ORSS8 aadA WT 1

2

3

4

ycf7 aadA 6

5

7

8 9 10 11

...8.

45

3.8 -1.

-.

A

_-

",-

_

ago

am

42 _ --- 32 ....3 2

..'.

t, i..

_

00 0 0 * 0*00

2

Fig. 3. Growth patterns of cells of wild type and transformants. Cells were spotted on either TAP medium in the dark (left) or on minimal medium under 30 gE/m2/s white light illumination (right). First row: wild-type, control strain containing aadA at the Sail site upstream of psaC (Figure 2), F15 (PSI mutant). Second row: ORF 58::aadA transformants D-1, D-2, D-4. Third row: ycf 7::aadA transformants 7-1, 7-2, 7-3, 7-4, 7-5. Fourth row: ycf 7::aadA transformants 7-6, 7-7, 7-8, 7-9.

4.8 0r 9 38 .0 Ln x

CL

F-

=

K

CL

1-

u~

2 4'-'

3.5

cytf

o

u

>%

>%

a_Ms.....

04

Fig. 2. Southern blot analysis of the 0RF58 and ycf 7 deficient transformants. Total DNA from wild type and the transformants was digested with EcoRI and PstI, electrophoresed on a 0.8% agarose gel, transferred to nitrocellulose membrane and hybridized with the 1.8 kb SalI-.XbaI fragment (upper) or with an aadA probe (lower). Wild type (WT), lanes 1, 2 and 3 and 4, 5 and 6 contain DNA from the 0RF58::aadA transformants with the aadA cassette in the same and opposite orientation, respectively, relative to the transcription of the psaC operon. Lanes 7, 8 and 9 and 10 and 11 contain DNA from the ycf 7::aadA transformants with the aadA cassette in the same and opposite orientation, respectively, relative to the transcription of the psaC operon. The map of the region of the psaC operon with the inserted aadA cassette together with the predicted sizes of the PstIEcoRI fragments are shown in the lower part of the figure.

sucrose density gradient centrifugation, Ycf 7 co-purifies with the thylakoid fraction (Figure 5A). To test whether Ycf 7 is indeed a transmembrane protein, thylakoid membranes were treated with either 2 M NaCl, 0.1 M Na2CO3 (pH 11.5) or 2 M KSCN in order to wash out extrinsic proteins, as described previously for other cytochrome b6f subunits (Breyton et al., 1994). Under all conditions tested, Ycf 7 remained associated with the membranes, indicating that it is indeed an integral protein (data not shown). The association of Ycf 7 with a specific complex of the photosynthetic electron transfer chain was further probed by analyzing the polypeptide composition of several mutants that lack one or the other of the photosynthetic complexes (PSII, the cytochrome b6f complex or PSI). The immunoblot in Figure SB reveals that, while Ycf 7 accumulates normally in the PSI and PSII mutants, it is

3500

Ycf7 Fig. 4. Immunoblot analysis of cytochrome f and Ycf 7 content in cells of wild type (WT) and transformant strains. Total cell proteins (5 jg chlorophyll) of wild type, a transformant with the aadA cassette inserted at the Sall site upstream of psaC (see Figure 2) and two ycf 7::aadA transformants, ycf 7-1 and ycf 7-3, in exponential growth phase were separated by SDS-PAGE. The gels were blotted, reacted with antisera against cytochrome f and Ycf 7 and revealed by the ECL method.

not detectable in mutants FuD6 and F18, both of which lack cytochrome b6f (Lemaire et al., 1986). These results suggest strongly that Ycf 7 is associated with the

cytochrome b6f complex. To test this hypothesis further, the cytochrome b6f complex was isolated from wild-type cells as described elsewhere (Pierre et al., 1995). Ycf 7 comigrates with the complex throughout the purification (not shown). The immunoblot in Figure SC shows that the Ycf 7 protein is enriched in cytochrome b6f complex preparations as compared with wild-type thylakoids. We conclude that the ycf 7 product represents a novel subunit of the cytochrome b6f complex. The presence of Ycf 7 was not revealed by Edman degradation of low Mr polypeptides from the purified b6f complex from C.reinhardtii separated by SDS-PAGE (Pierre and Popot, 1993), suggesting that its N-terminus is blocked, either naturally or artefactually.

Role of ycf 7 in photosynthesis

A T

P

-:

S TK

V7--7

-

LL -

oytf am

Ycf 7 '

vctf7 100 75 50 10

D

A-

cytf

WT

C

B

WT M

L-

mmd~

_

tf

t

"i

cyf

cytf

Ycf7_

41.

m m

cytb6

tm,, 40:4mo

Fe-S

a

w

f

..,

,.

_

Fig. 5. Subcellular localization of the Ycf 7 protein examined by immunoblot analysis. Subcellular fractions were analysed by SDSPAGE followed by electrotransfer and immunolabelling for cytochrome f and Ycf 7. (A) Broken cells (T) were separated into a pellet (P) containing the membrane fraction, a supematant (S) containing soluble proteins and thylakoids (TK) which were purified from the pellet fraction by discontinuous sucrose density gradient centrifugation. Proteins from each fraction corresponding to 5 .g chlorophyll were loaded. (B) Total cell proteins (5 ,ug chlorophyll) from five strains (wild type; FuD7, lacking PSII; FuD6 and F18, lacking cytochrome b6f; F15, lacking PSI). (C) Thylakoid membrane proteins (10 gg) from wild type (WT) and one ycf 7-deficient transformant (M) as compared with the proteins (2 ,ug) of the cytochrome b6f complex purified from wild type (b6f).

Absence of Ycf 7 leads to a reduced accumulation of the cytochrome b6f complex As already seen in Figure 4, the amount of cytochrome f in the ycf 7-deficient transformants appears to be reduced as compared with wild type. We have estimated by immunodetection the levels of several subunits of the cytochrome b6f complex in the ycf 7::aadA transformants by comparing the intensity of labelling with that obtained using a dilution series of wild-type extracts. This immunoblot (Figure 6) reveals that, in cells in exponential phase, cytochrome f, cytochrome b6, the Rieske protein and subunit IV accumulate to 25-50% of the levels observed in wild-type cells. Loss of the Rieske Fe-S protein during purification of the cytochrome b6f complex from the ycf 7::aadA transformant The Ycf 7-less complex from the ycf 7-deficient transformant was isolated using a protocol developed for wildtype cells: following selective solubilization by Hecameg, the complex was purified by sucrose density gradient centrifugation and hydroxylapatite chromatography (see Pierre et al., 1995). The resulting preparation contained highly enriched cytochromef and cytochrome b6, both of which underwent the same changes of redox state as in the wild type upon oxidation or reduction with potassium ferricyanide, sodium ascorbate or sodium dithionite (data not shown). The polypeptide compositions of the complex from the ycf 7::aadA transformant and from wild type were compared using immunoblotting. As shown in Figure 7, the relative amounts of cytochrome f, cytochrome b6, subunit IV, PetG and PetM are comparable in the two complexes, indicating that these subunits are assembled normally even in the absence of the ycf 7 product. As expected, the latter was absent in the complex from the

_M

_

sub

Fig. 6. Accumulation of cytochrome b6f complex subunits in the ycf 7-deficient cells. The amount of total protein loaded (corresponding to 5 ig chlorophyll) from ycf 7 and wild-type cells (100) were the same; ycf 7, transformant ycf 7-1; decreasing amounts (100, 50, 25 and 10% of that in lane 2) of wild-type proteins were separated and probed with antisera against cytochrome f, cytochrome b6, the Rieske protein (Fe-S) and subunit IV. cytf w

M

cytb6 M

w

subliV w

M

Fe-S w

M

Ycf 7 w M

PetG w M

PetM w M

Fig. 7. Subunit composition of cytochrome b6f complex purified from wild type (W) and ycf 7-1 transformant (M). The same amount of cytochrome b6f complex (50 pmol) was loaded on each lane. The antisera used for immunoblotting are indicated at the top of each lane. Wedges indicate the bands of cytochrome f, cytochrome b6 and Ycf 7. The gel system of Schagger and von Jagow was used for cytf, cytb6, subIV, Fe-S (Rieske protein) and ycf 7, and 12-18% SDS/8 M ureaPAGE was used for the lanes corresponding to PetG and PetM.

ycf 7::aadA transformant. It is worth noting that, while disruption of the petA, petB, petD or petG genes prevents accumulation of other subunits of the cytochrome b6f complex (Kuras and Wollman, 1994; Berthold et al., 1995), that of ycf 7 does not have the same pleiotropic effect. The Rieske protein, however, is undetectable in the purified mutant complex. Since the complex is enzymatically active in vivo (see below) and the Rieske protein is present both in the cells (Figure 6) and in the thylakoid membranes (data not shown) from the ycf 7::aadA transformant (Figure 6), it must dissociate from the complex 3501

Y.Takahashi et al.

either during the solubilization of the complex and/or during the subsequent purification steps. Analysis of individual fractions after sucrose density gradient centrifugation revealed the presence of some Rieske protein in the upper part of the gradient, which apparently had dissociated from the complex at an early stage. The rest of the Rieske protein (30-50%) was still associated with the complex which, however, migrated as a monomer rather than the wild-type dimer (data not shown). After purification of the complex by hydroxylapatite chromatography, all of the Rieske protein was lost (Figure 7). As expected, the electron transfer activity from plastoquinol to plastocyanin of the purified complex was undetectable (data not shown). Under the same conditions of purification, the wild-type cytochrome b6f complex remains a dimer (Breyton et al., 1995), retains a stoichiometric amount of the Rieske protein throughout the purification (Pierre et al., 1995 and Figure 7) and is enzymatically highly active (Pierre et al., 1995). These results suggest that the absence of Ycf 7 weakens, either directly or indirectly, the binding of the Rieske protein to the cytochrome b6f complex. The fact that the complex from the ycf 7::aadA transformant migrates as a monomeric species retaining some Rieske protein during the sucrose gradient fractionation suggests that loss of the Rieske protein might follow monomerization.

Stability of the cytochrome b6f complex is impaired in the ycf7::aadA transformants Figure 8 displays the fluorescence transients of wild-type and ycf 7::aadA cells recorded at different cell concentrations in the cultures. The fluorescence patterns in wild-type cells did not change significantly under these conditions. In contrast, the fluorescence transients of the ycf 7-1 transformant no longer declined after reaching the maximum fluorescence yield when the cells were in stationary phase (9X 106 cells/ml), indicating an inability to reoxidize the plastoquinone pool. Immunoblot analysis revealed that cytochrome f accumulates normally in wild-type cells both in lateexponential (6X 106 cells/ml) and stationary phase (Figure 8B). In contrast, the accumulation of cytochrome f in ycf 7::aadA cells, which is already lower than in wildtype cells during the exponential phase (lane 1), diminished considerably when the cells reached stationary phase (lane 3). These results suggest that Ycf 7 is required for stable accumulation of the cytochrome b6f complex, especially in ageing cultures. The rate of the electron transfer reactions within the cytochrome b6f complex is inhibited more strongly in ycf 7-deficient cells than in wild-type cells by the resting proton electrochemical gradient Previous research has shown that flash-induced electrochromic absorbance changes at 515 nm in dark-adapted algae reflect thylakoid membrane potential changes (Junge and Witt, 1968; Joliot and Delosme, 1974). Some of these changes are associated with the electron and proton transfer reaction occurring within the cytochrome b6f complex and can therefore be used to monitor its activity in vivo (Delosme, 1991). We compared the activity of the b6f complex in wild-type and ycf 7-deficient cells. Thylakoid

3502

A

B -. _

-i1 L

cyttf

_-

I

_

damw* _

Fig. 8. Reduced accumulation of cytochrome b6f complex in the ycf 7::aadA transformants in stationary phase. (A) Fluorescence induction transients of each cell culture were measured after 10 min dark adaptation of the cells. (B) Accumulation of cytochrome f in ageing cells from wild type and the ycf 7-3 transformant. Total cell proteins (5 gg chlorophyll) were loaded on each lane. Curves and lanes 1-3 refer to wild-type cultures at 6.OX 106, 9.Ox 106 and 1.2x 107 cells/ml and to the cultures of the ycf 7-3 transformant at 3.4X 106, 7.7X106 and 9.0X 106 cells/ml respectively.

membrane potential measurements were performed in living C.reinhardtii cells in which PSII reaction centres had been blocked by anaerobiosis. Positive charges were delivered to the reduced plastocyanin pool by exciting PSI reaction centres with light flashes (see Materials and methods). Under these conditions, the kinetics of the flash-induced membrane potential increase display two sequential phases (Joliot and Delosme, 1974). A first phase (phase a), completed in