The Plant Cell, Vol. 3, 203-212, February 1991 O 1991 American Society of Plant Physiologists

Biosynthesis of the Chloroplast bsf Complex: Studies in a Photosynthetic Mutant of Lemna

Barry D. Bruce’ and Richard Malkin2 Department of Plant Biology, University of California, Berkeley, Berkeley, California 94720

The biosynthesis of the has been studied in a mutant, no. 1073, of Lemna perpusilla that contained less than 1% of the four protein subunits when compared with a wild-type strain. RNA gel blot analyses of the mutant indicated that the chloroplast genes for cytochrome f, cytochrome b6, and subunit IV (petA, petB, and petD, respectively) are transcribed and that the petB and petD transcripts undergo their normal processing. Analysis of polysomal polyA+ RNA indicated that the leve1 of translationally active mRNA for the nuclear-encoded Rieske Fe-S protein (petC) was reduced by >lOO-fold in the mutant. lmmunoprecipitation of in vivo labeled proteins indicated that both cytochrome f and subunit IV are synthesized and that subunit IV has a 10-fold higher rate of protein turnover in the mutant. These results are discussed in terms of the assembly of the cytochrome complex and the key role of the Rieske Fe-S protein in this process.

INTRODUCTION

All organisms capable of photosynthetic autotrophic me- has only started to receive attention. Recent studies have tabolism or heterotrophic respiratory metabolism share the begun to investigate the questions of how these proteins ability to couple electron transport to proton translocation are inserted into a membrane, when and where prosthetic across a membrane. In all cases, this coupled electron groups are attached, and what directs the assembly of the transport is carried out in distinct multisubunit membrane individual subunits into a functional complex. complexes. The most widespread and well studied of The present work considers the mechanism of assembly these complexes is the cytochrome bcl (complex 111) and of the chloroplast cytochrome b6f complex using a pho- bsf class of electron transport complexes. These com- tosynthetic mutant of the higher plant Lema perpusilla plexes function as quinol/cytochrome c (plastocyanin),ox- no. 1073. This mutant was originally isolated as a flowering idoreductases. In all known organisms, this complex con- mutant, and subsequent biochemical characterization in- tains a b-type cytochrome containing two hemes of differ- dicated that it was a photosynthetic mutant lacking the ent potentials, a c-type cytochrome, and a high potential thylakoid membrane cytochrome bsf complex (Shahak et 2Fe-2S center bound by a protein known as a Rieske al., 1976; Malkin and Posner, 1978; Lam and Malkin, Fe-S protein (Malkin, 1988). In chloroplasts, there is an 1985). However, this mutant was not defective in heme additional subunit (subunit IV) that has been shown to bind biosynthesis or in heme insertion because its thylakoids plastoquinone (Doyle et al., 1989). In mitochondria and showed normal amounts of another chloroplast cyto- purple bacteria, the protein sequence homologous to sub- chrome, cytochrome b-559. Lam and Malkin (1985) also unit IV has become fused with the cytochrome b protein, used protein gel blotting to show that not only were the yielding a larger protein with a possible added role in prosthetic groups lacking in the mutant but that the apo- quinone binding (Hauska et al., 1988). proteins for cytochrome f, cytochrome bs, Rieske Fe-S Although extensive studies have been carried out on the protein, and subunit IV were absent. mechanism of electron transfer and proton translocation The present work considers the mRNA levels for the catalyzed by cytochrome bcl/b6f complexes (Rich, 1986), cytochrome complex subunits, the transient synthesis of the mechanism by which these complexes are assembled the corresponding polypeptides, and the fate of these proteins in the thylakoid membrane of L. perpusilla no. ’ Current address: Department of Botany, University of Wisconsin, 1073. These results are considered in relation to the Madison, WI 53706. mechanism of assembly of this multimeric complex in To whom correspondence should be addressed. thylakoids. 204 The Plant Cell

LemnaperpusilloWT Lemnaperpusilla 1073 accumulate Rieske <3%th do t d e an ,protei n accumulated to <2% of the levels of these subunits relative to wild-type 125 25 10 5 1 125 25 10 5 1 membranes. The amount of another photosynthetic mem- brane protein, light-harvesting chlorophyll-protein complex I; II (LHCP II), that accumulated in wild-type and mutant thylakoids was identical (data not shown), although there a sligh s t wa decreas e amounth n i ef subuni o t f o I I t photosystem (thepsaI D gene product mutane th n )i t (see Figure 2).

RN l BloAGe t Analysi Cytochrome th f so e b6f Genes

To determine whether the lack of protein accumulation in the mutan failura t o membranet mRNn ei e du As wa s accumulation l bloge t A analyseRN , s were don totan eo l RNA and a polysomal polyA+ RNA fraction. Figure 3 shows the results of an analysis of the chloroplast-encoded sub- cytochrome unitth f so f complexe eb . This blot contained equal amounts of total RNA that was ethanol precipitated before loading ont gele oth . Therefore slighe th , t differ- ences in intensity cannot necessarily be interpreted as a CYTOCHROMEf difference in the absolute RNA levels. It is clear from this Figur . Quantitative1 e Protein Blot Analysi perpusilla. L f so Wild- figure that all plants, including the closely related Type (WT Mutand an ) t (1073) Thylakoid r Cytochromsfo e. f L. gibba, accumulated a single predominant mRNA tran- scrip f approximatelo t r peffo Ab k (cytochrom 1 y4. . ef) Proteins were resolve SDS-PAGy db electroblotted Ean d onto nitrocellulose. The blot was probed with antiserum raised against spinach cytochrome r. The numbers represent the amount of chlorophyl microgramn i l s loade eacdn i h laneantibode Th . s ywa visualized with goat anti-rabbit horseradish peroxidase. VvT 1073 0 8.01.2 0 4 6 0 8.01.0 2 6 0 4 6 0 6

Cytochrome RESULTS

Subunit IV Protei l BlonGe t Analysi Thylakoif so d Protein Lemnasn i

To determine accurately the level of the subunits of the cytochrome comple . perpusillaL n xi . 107no 3 versue sth PSI-subunit II wild-type/., perpusilla, quantitative protein gel blotting was done varyiny B . amoune gth f proteio t n loaded onte oth gel before blotting, it was possible to compare the signals Rieske Fe-S of the wild type and mutant and, thus, provide an estimate relative th n o e amount f proteiso n that accumulatee th dn i respective thylakoids. These blots were normalizee th n 6 do 1. 0 8. 0 2 0 4 0 6 6 1. 0 8. 0 2 0 4 0 6 basis of the amount of chlorophyll loaded. Figure 1 shows Figur . eQuantitativ2 e Protein Blot Analysi . perpusillaL f so Wild- a quantitative protein gel blot using a cytochrome f anti- Type (WT) and Mutant (1073) Thylakoids for Subunits of the body from spinach. By comparing the signal intensity from Cytochrom f Complex6 eb . the 1 HQ lane of the wild type versus the 125 n<3 lane for Analyses were carried out as in Figure 1 using spinach antibodies the mutant, it was possible to conclude that cytochrome f raised against cytochrome £>6, subunit IV, the Rieske Fe-S protein, accumulated in the mutant thylakoids to <1% of the level controlanda s a , , subuni I (thI t e psaD gene product f spinaco ) h that it occurred in the wild-type thylakoids. Extending this chloroplast photosyste mI (PSI) . Becaus f contaminatioeo y nb typ f analysieo othee th ro s t subunit cytochrome th f so e nonspecific reactions with the Rieske antibody, the presence of complex s showa , n Figuri n , indicate2 e de th tha n i t the cross-reacting produc indicates i t arrobottoe e th th wy n di mb mutant, cytochrome t>6 accumulated to <1%, subunit IV panel. Cytochrome £> 6f Comple x Assembl5 20 y

ro ro ro resula gene s f theio tswa r cotranscription int singloa e i>- ro- large polycistronic transcript that subsequently underwent o o o o a comple post-transcriptiof o t xse n processin splicind gan g 7> '55 steps. This cotranscriptio post-transcriptionad nan l proc- o => ^ essin bees ha gn observe maizen di , a spinachpe d an , 3. .O O. O. 5 S a s i fc (Heinemeye t al.e r , 1984; Berend t al.se , 1986; Roct ke o> a. a. O> Q. Q. L C . Q > 0 al., 1987) interestins i t I . noto gt e thamRNe th t A accu- mulatio r botnfo f thesho e genemutane th n si t appearo st aboue b t twic wilee thath d f typeo t . Whethee r thidu s si increasn a o t transcriptionan ei l activit r mRNyo A stability t knownisno . From these conclude e studiesb n ca t i , d that transcripts for all three chloroplast cytochrome complex genes accumulate in the L. perpusil/a mutant. Although thers accumulatiowa e f mRNAo e n th r fo s chloroplast genes in the preceding experiment, it was still 3.94kb possible thamutane th tdefectivs wa t transcriptioe th en i n or accumulation of the mRNA for the nuclear-encoded 1.97kb subunit, the Rieske Fe-S protein. In fact, RNA gel blot analysis of total RNA for the Rieske Fe-S transcript indi- cated that the mRNA level in the mutant was <10% of that founwile th d dn typi e (dat t shown)ano . However, this signal was quite low and required hybridizations under low-stringency conditions. To increas hybridizatioe eth n signal, polysomal polyA+ RNA was isolated from the mutant and wild type. The l bloresulge t A f analysithio t sRN shows i Figurn i . e4 Equal werA loadingRN e f useo s o allot d r direcwfo t comparison between transcript accumulation in the wild Cytochrome f Cytochrome b« Subunit H type and mutant. After a 1-day exposure, a single tran- Figure 3. RNA Gel Blot Analysis of Chloroplast-Encoded Cyto- script was seen in the wild type, but only a very faint band was observed in the mutant. After a 5-day exposure, a chrome bsf Components. signa observes mutanlwa e th n di t wit same hth e molecular Whole Lemna tissu s frozeewa liquin ni d nitrogen beforA eRN wile th weighd n i type s a t . Scanning densitometre th f yo isolation s isolatewa . TotaA d RN usinl guanidinium/phenoga l 1-day exposure indicated that the transcript in the mutant procedure. Twenty micrograms of total RNA was denatured and subjecte formaldehydo dt e agaros l electrophoresisege , blotted only accumulate wile levee th 0.8o dt th n i f typel% o . onto nitrocellulose, and hybridized with one of the following 32P RNA dot blots gave a value of about 1% for the mRNA hexamer-labeled cDNA clones 400-ba : p fragment encodina gpe accumulation in the mutant (data not shown). Whether the cytochrome / . Barkan)(courtesA . Dr 300-ba f , o y p fragment lac f mRNo k n effecAa n accumulatioo o t t e du s wa n encoding spinach cytochrome b6 (courtes Bohnert). H . Dr d f yo ,an transcription activity or on the stability of the mRNA is not a 420-bp fragment encoding pea subunit subunit IV (courtesy of known because run-on transcription analysi t beeno ns sha Dr. A. Barkan). Hybridization was at 42°C in 5 x SSC, 40% done. However l bloge t A analysi otheo ,RN tw f rs o nuclear formamide for 48 hr. mRNA sizes are based on isolated avocado photosynthetic genes, cab and rbcS, encoding LHCP II ribosomal RNA standards (courtesy of Dr. M. Tucker). WT, wild and the small subunit of ribulose-1,5-bisphosphate carbox- type; 1073, mutant. ylase, showed no change in the mRNA level between the mutant and wild-type strains (data not shown). Therefore, it can be concluded that the lesion in this mutant does not There was, in addition, a second, larger minor band found affect the accumulation of mRNA for all nuclear-photosyn- in the mutant. Whether this band represents a second, thetic genes. large At transcriprpe t clearno s .ti Fro intensite mth e th f yo 4.1-kb band t i woul, d appear thapefe th tA transcript accumulated to the same level in all three plants. Immunoprecipitation of Thylakoid Proteins Equivalent blots were probed with clones that are spe- cific for the coding region for pefB (cytochrome b6) and l Becausbloge t A analyseeRN s indicated that mRNAr sfo oefD (subunit IV). Both of these probes detected at least chloroplase th t subunit cytochrome th f so e comple- ac x 12 different transcripts in all three organisms. The large cumulate perpusilla. L e th n di f interes o mutant s wa o t t i , number of different sized transcripts for both of these determine whether these transcripts were translationally Plane Th t Cel l6 20

ro CO polyclonal antibodies raised against spinach proteinse w , fe were abl shoeo t w tha mutane tth activ s synthesii te th en i s * £ of cytochrome f and subunit IV of the cytochrome complex .o , respectively)kD (molecula7 1 d an ,r D k masse 3 3 f so = = '35 5 '5 ' althoug e casth f cytochromeo n i h e leveth f thi, o elf s 8 protein accumulated to a lower level than that observed in Q. ^X ^X wild-type th e strain. Thi relates e mor e b resul th y edo t ma t 4) 4) fc o5 Q. Q. Q. Q. rapid rat f turnovee o subunite th f o r s observee th n di mutant strain (see below) controa s A .thesn i l e experi- ments, the syntheses of LHCP II (molecular mass of 28 kD) and plastocyanin (data not shown) were measured by immunoprecipitatioe th n procedure resulte th d s an ,show n in Figure 5 indicate that LHCP II is synthesized at com- parable level botsn i h strains complee .Th x blotting pattern seen with the LHCP II antisera, such as the diffuse band at 33 kD, is probably a contamination of D-1 and D-2 from 3.94kb photosystem II that cross-reacts with the antisera. When similar experiments were done using polyclonal 1.97kb r antisera raised against spinach cytochrome r t>Riesko 6 e R H K 22 2

97.2- 66.2-

- cytochrome f - LHCP II

- cytochrome b 6 21.5- -Rieske Fe-S -SubuniV I t 14.4-

expy da . 5 y expda . 1 |n-LHCIl|| g-cyl/ ||a-cyt ly || ct-Ricskc ||a-Sub. IV |

Figur . Synthesi5 e f o Chloroplass t Membrane Proteiny b s Rieske Fe - S L. perpusilla. Intact L. perpusilla wild-type (WT) and mutant no. 1073 (1073) Figure 4. RNA Gel Blot Analysis of the Rieske Fe-S Protein. plants individua 0 (fiv1 eo t l plantlets) w werlo n ei r labeleh 2 r dfo 35 RNA gel blot analyses were carried out as in Figure 3 using light with 1 mC of S-TransLabel (ICN, Costa Mesa, CA). Whole polysomal polyA+ RNA, isolated using the procedure of Lincoln et plants were homogenize totad dan l proteins were acetone precip- al. (1987). This probes blowa t d wit 850-bn ha p EcoRI fragment itated. This total Dlant protein fraction was resuspended and encoding pea Rieske Fe-S protein (courtesy of Dr. J. Gray). WT, denatured wit SOS% h2 . This total protein suspensio pres n-wa wild type; 1073, mutant. cleared with 80 nL of preimmune serum and 100 ^L of protein A- agarose beads (Bio-Rad). Specific proteins were immunoprecipi- tated by the addition of polyclonal antisera raised against spinach proteins designated LHC ^L)0 (1 ,PI I cytochrom nL),5 (2 e f cyto-

chrome b6 (60 nL), Rieske (60 nL), and 100 ^L of protein A- active studo T . y this developee w , procedurda - im o et agarose beads. This immunoprecipitate was denatured and re- solve SDS-PAGy db visualized Ean fluorographyy db . Each lane munoprecipitate a specific protein from entire in vivo la- represents an immunoprecipitation from the same amount of plant beled plants. This procedure is very sensitive and can tissue base chlorophyln o d l content bane e Th th d. l seeal n ni detect proteins that are made in very small amounts and/ lanes at 43 kD corresponds to the leading edge of the very large or are very rapidly turned over. As shown in Figure 5, by amount (>250 ng/\ane) of the heavy chain of IgG used in the measuring the levels of immunoprecipitated proteins using immunoprecipitation. Cytochrome £>6 / Complex Assembly 207

Fe- t possiblSno proteins deteco et wa synthesit y i , an t s turnove mutane f LHCth o r comparen s i a PI tI d wite hth of these proteins in either the wild type or mutant. The wild type. We conclude from these results that the chlo- lac f Rieskko e Fe-S protein synthesi expectes swa - dbe roplast-encoded subunits of the cytochrome complex are cause very little mRNA accumulated for this protein. How- synthesized at the same rate in the mutant and wild type

ever, in the case of cytochrome £>6, the lack of detectable and that turnover is accelerated in the mutant. The fact protein could be due to low affinity of the heterologous that the pattern of protein degradation for subunit IV seen antibody used or could reflect a low level of protein syn- in Figur differens i e6 t betwee wile nth d typmutand ean t thesis. The latter explanation is supported by our results suggests that accessibilit proteaseo yt differens i e th n i t that suggest that post-transcriptional regulation of the mutant.

synthesis or assembly of cytochrome t>6 may occur. In our studies, we have found that in vitro labeling of isolated spinach chloroplasts with 35S-TransLabel and subsequent isolation of the cytochrome complex did not result in the insertion of labeled cytochrome £>6 into the resolved com- Lemna perpusillaT W plex even though subuni cytochromd an V I t ewerf e shown by this procedur e labeleb d inserteo et an d d inte th o 0.0 0.5 1.0 3.5 6.5 17 24 40 complex. This result suggests tharate f turnoveth te o f o r 31.0- the cytochrome b6 subunit is much lower than that of the other subunits and that the complex is able to undergo subunit exchange without requiring de novo synthesis of 2 21.5- the entire complex. 14.4-

Pulse-Chase Determinatio f Proteino n Turnover CHASE TIME (hrs.) To determinchloroplast-encodee th y wh e d cytochrome subunits fail to accumulate despite having normal levels of mRNA and comparable levels of protein synthesis, the rate Lemna perpusilla 1073 of turnover of several proteins in whole plants was meas-

ured. This was done by giving entire plants a 1-hr radiola- 0.0 0.5 1.0 3.5 6.5 17 24 40 beling pulse of 35S-TransLabel, followed by extensive washin chasind gan g wit approximatn ha e 1000-fol- din 31.0- creas coln ei d methionin cysteined ean t variouA . s time chasepointe th proteie n si th , f interesno immunos wa t - 21.5- precipitated fro msolutioa n containing total plant proteins. By this procedure turnovee th , r rate r subunith e fo n i V I t 14.4- wile mutanth d d typs measuredan t ewa shows A . n ni Figure 6, subunit IV in the wild type could be detected up to 17 hr after the chase addition, whereas in the mutant this subunit disappeared after approximatel. hr 3 o t r h y1 CHASE TIME (hrs.) Approximate half-times for turnover of subunit IV in the mutant and wild type were 2 hr and 24 hr, respectively. Figur . eTurnove6 Cytochrom e f Subuniro th f o V I t e bef Complex in L. perpusilla. Thus, this subunit turns over more rapidly (10 times) in the mutant as compared with the wild-type strain. Intact L. perpusilla wild-type (WT) and mutant no. 1073 (1073) To verify that the mutant was not simply more active in plants (five to 10 individual plantlets) were labeled for 1 hr in low 35 protein turnover or contained an elevated level of pro- light with 1 mC of S-TransLabel. After residual label was re- teases, we also measured the turnover rate for plastocy- move washingy db , cold methionin cysteind ean e were addeo dt anin, as shown in Figure 7. Plastocyanin showed no sig time -th e t poinA . 2t 0mM indicated , whole plants were homoge- nize totad dan l proteins were acetone precipitated. This total plant nificant difference in protein turnover in the mutant when protein fractio resuspendes nwa denatured dan d wit SDS% h2 . compared with the wild type and could be detected up to Subuniimmunoprecipitates wa V I t additioe th y db f polyclonano l 24 hr after the chase addition in both strains. We also antisera raised against spinach subuniproteid an V I nt A-agarose studied whether this turnove specifis wa rr protein fo c s beads. This immunoprecipitate was denatured and resolved by destine r insertiofo d n inte thylakoith o d membrany b e SDS-PAG visualized Ean fluorographyy db . Each lane represents measurin turnovee gth f LHCo r PI (datI t shown)ano . This an immunoprecipitation from the same amount of plant tissue result indicated tha increaso t n therrat e s f th eo e wa n ei base chlorophyln do l content. 208 The Plant Cell

Lemna perpusillaT W DISCUSSION

0 4 4 2 7 1 5 6. 5 3. 0 1. 5 0. 0 0. The results fro presene mth t photosynthetie studth f yo c 14.4- mutant L perpusilla no. 1073 indicate that the mutation in

a-SubunitIV a-LHCP CHASE TIME (hrs.)

Lemna perpusilla 1073

0.0 0.5 1.0 3.5 6.5 17 24 40 97.4 66.2

42.7

CHASE TIME (hrs.) Figur . eTurnove7 f Plastocyaniro perpusilla.. L nn i 31.0 Turnover of plastocyanin in wild-type (WT) and mutant (1073) L. perpusilla plantlets was measured as in Figure 6 using an antibody raised against spinach plastocyanin. 21.5

14.4

Subunit Membrane Association

determino T e whether subuni associates i V I t d wite hth thylakoid, it was necessary to in vivo label whole plants an isolatdo t e thylakoid membranes before immunoprecip- itation. To test further the association with thylakoids, the WT 107T 107W 33 membranes were treated with the chaotropic agent NaBr. Washing thylakoid membranes with NaBr removes periph- Figure 8. Association of Subunit IV and LHCP II with Thylakoid eral proteins, such as the CF1 complex (Kamienietzky and Membranes in L. perpusilla. Nelson, 1975). Figure 8 shows the results of immunopre- Intact L. perpusilla wild-type (WT) and mutant no. 1073 (1073) cipitation of in vivo labeled subunit IV and LHCP II from plant sindividua0 (fiv1 o et l plantlets) w werlo n i e r labeleh 5 r dfo NaBr-washed thylakoids. It is clear that both proteins are lighf 35o S-TransLabel t C witm h1 . Intact chloroplasts were iso- associated witthylakoie hth d membrane because NaBr lated fro labelee mth d plant werd san e washed twic removo et e washin t lea no theio d t d grdi removal face t.Th that there any residual label. These chloroplasts were lysed and the thyla- was less labeled subunit IV in the mutant thylakoids is to koids were collecte centrifugationy db thylakoie Th . d membranes be expected because this protein is more rapidly turned were washed twice wit chaotropie hth c agen . Aftet M NaB 3 r t ra ovewould an r d not, therefore, accumulat same th eo et the NaB removes rwa washingy db , total thylakoid proteins were wile th d leven typei s a l . (See r timexampler h ,fo e5 6. e th , acetone precipitated. This total thylakoid protein fractios nwa resuspended and denatured with 2% SDS. Specific proteins were point in the pulse-chase experiment shown in Figure 6.) immunoprecipitated by addition of polyclonal antisera and protein LHCP II, another known integral membrane protein, was A-agarose beads. This immunoprecipitate was denatured and immunoprecipitated equivalently fro thylakoide mth s iso- resolved by SDS-PAGE and visualized by fluorography. Each lane lated fro mutane m th wil d dan t type becaus rate ef th eo represents an immunoprecipitation from the same amount of plant turnove thir fo rs sam e proteith bots en i nhwa plants. tissue base chlorophyln do l content. Cytochrome bsf Complex Assembly 209

this organism affects the transcription and/or mRNA ac- the stable assembly of the thylakoid cytochrome complex cumulation of the petC gene encoding the Rieske Fe-S and argue against a concept of a stable complex that lacks protein of the cytochrome b6f complex. A possible expla- any one component of the cytochrome complex. Assembly nation involving a second mutation that affects the accu- of the chloroplast cytochrome complex appears to operate mulation of chloroplast-encoded subunits cannot be ex- by an “all-or-none” mechanism whereby all components cluded by this work but appears less likely because this must be present for the stabilization of the complex. mutation has been induced by x-rays that presumably In addition to the few cytochrome b6f complex mutants generate nuclear mutations. The three chloroplast-en- in higher plants, severa1 photosynthetic mutants have been coded subunits, cytochrome f, cytochrome b6, and subunit isolated from the green alga Chlamydomonas reinhardtii IV, appear normal in relation to their transcription, mRNA that are defective in the cytochrome b6f complex. The accumulation, and protein synthesis; in the case of subunit most well-characterized mutants include the UV-irradiated IV, it was also possible to document an association of this nuclear mutant ac21 (Levine and Goodenough, 1970), the subunit with the thylakoid membrane. The fact that these chloroplast mutants FuD2, FuD4, FuD6, and FuD8 gener- three subunits failed to accumulate to a detectable level in ated by mutagenesis with 5-fluorodeoxyuridine (Bennoun the mutant is most likely a result of their accelerated rate et al., 1978), and the nuclear mutant F18 generated with of turnover in the thylakoids, as documented by a 10-fold 5-fluorouracyl (Girard et al., 1980). These mutants have greater rate of turnover of subunit IV in the mutant as been characterized by Bennoun and coworkers (Bendall compared with the wild type. et ai., 1986; Lemaire et al., 1986, 1987). In general, many A preliminary observation of interest in relation to the of these mutants are similar to those detected with higher turnover of the constituents of the cytochrome complex plants in that they fail to accumulate any of the cytochrome was the finding that the rate of turnover of individual protein b6f subunits. However, in a few alga1 strains, some of the components of the complex was not identical because the chloroplast-encoded subunits were present in the thyla- rate of turnover of cytochrome b6 appeared to be consid- koid membrane even though the nuclear-encodedsubunit, erably lower than that of other protein subunits. This the Rieske Fe-S protein, was absent. These mutants ap- implies that a mechanism must exist for the degradation pear to be substantially different from other cytochrome of specific components within the complex without a deg- b6f mutants, and further study at the molecular genetic radation of the entire complex. This situation is similar in level might prove instructive in terms of the assembly of some respects to observations relating to the turnover of the complex. photosystem II components in the photosynthetic mem- In yeast, a number of studies using respiratory mutants brane, where the turnover of the integral membrane protein have led to the conclusion that there are two different but D1 is considerably greater than the rate of the other protein coexistent pathways for the transport, prosthetic group components in this complex (Mattoo et al., 1981; Ohad et attachment, and assembly of the cytochrome bc, subunits ai., 1985). The mechanism by which a single subunit in a (Crivellone et al., 1988). One pathway that includes only multisubunit complex is specifically degraded, resynthe- the nuclear-encoded subunit is insensitive sized, and reinserted into a complex is not known at the to mutations affecting the accumulation of cytochrome present time, but our findings on the cytochrome b6f b. Cytochrome c, is able to be transported into the mito- complex, taken in conjunction with the previous work on chondria, undergo heme attachment, and to be inserted the photosystem II complex, would suggest that this is a stably into the membrane, independent of the accumula- general property of thylakoid membrane complexes. tion of cytochrome b. In the other assembly pathway, The finding that the nuclear-encoded subunit of the subunits such as the Rieske Fe-S protein, core I, core 11, cytochrome b6f complex is essential for the stable assem- and the low molecular weight subunits are strongly af- bly of this complex in thylakoids is consistent with a similar fected by the status of cytochrome b. When cytochrome concept first presented by Miles (1982) after characteriza- b is missing, these proteins show lower levels of transport tion of maize mutants that lacked the cytochrome complex. and a lack of prosthetic group attachment and fail to insert Biochemical (Miles, 1982; Metz et al., 1983) and molecular stably into the membrane. In a recent study using gene biological (Barkan et al., 1986) studies led to the conclusion disruption mutants in yeast, Crivellone et al. (1988) con- that the entire cytochrome complex failed to assemble in cluded that only the Rieske Fe-S protein had no effect on the mutants and that a “key assembly protein” of nuclear the assembly of the other subunits. A number of studies origin was required for complete assembly of this complex. have shown that it is possible to assemble “partial” cyto- Subsequent work by Herrmann’s group with Oenothera chrome bcl complexes in yeast (Capeillere-Blandin and (Stubbe and Herrmann, 1982) has supported this view, Ohnishi, 1982; Japa and Beattie, 1988; Gatti et al., 1989) and Willey and Gray (1988) extended the proposal to and that the tight assembly of the complex that is observed specify that the Rieske Fe-S protein may function as this in the case of the cytochrome b6 f complex does nOt occur. key nuclear protein. The present results with the L. per- It is not yet clear why organisms that require a functional pusilla mutant identify the as essential for cytochrome complex in either their respiratory or photo- 210 The Plant Cell

synthetic membranes have apparently evolved different lsolation of Polysomal PolyA' RNA mechanisms for the assembly of these complexes, partic- ularly when all of these organisms must assemble this Entire plants were frozen in liquid nitrogen and stored at -80°C. complex from a number of different nuclear-encoded and Polysomal polyA' mRNA was isolated using the procedure of organelle-encodedsubunits. Lincoln et al. (1987).

RNA Gel Blot Analysis METHODS After ethanol precipitation, RNA was resuspended and an aliquot was quantitated by UV absorbance. All lanes were loaded with equal amounts of resuspended RNA. With RNA gel blots using Growth of Lemna gibba and L. perpusilla Wild Type and L. polyA+ RNA, 2 pg of RNA were loaded per lane. RNA gel blots perpusilla No. 1073 using total RNA were loaded with 10 pg of RNA per lane. RNA was denatured with formaldehyde and formamide at 50°C for 30 Lema fronds were grown photoheterotrophically under aseptic min. RNA was separated using a denaturing formaldehyde aga- conditions on 0.5% glucose in E medium (Tobin, 1978) at 25°C rose gel electrophoresis system (Sambrook et al., 1989). RNA using heat- filtered incandescent lights (50 pE/mz/sec to 80 pE/m2/ was transferred to either nitrocellulose or nylon membrane by SeC). Fernbach flasks containing 1O00 mL of medium were inoc- capillary blotting overnight. ulated with one to 10 individual plants and allowed to reach a density of two to three plants deep over a period of 4 weeks to 5 weeks. Preparation and Labeling of cDNA Radiolabeled Probes

lsolated DNA clones were generously donated as follows: a 400- lsolation of Total RNA bp fragment encoding pea cytochrome f and a 420-bp fragment encoding pea subunit IV from Dr. A. Barkan, a 500-bp fragment Whole plants were frozen in liquid nitrogen and pulverized with a encoding spinach Rieske Fe-S protein from Dr. R. Herrmann, an mortar and pestle. This frozen slurry was then mixed and further 850-bp fragment encoding pea Rieske Fe-S protein from Dr. J. pulverized with guanidinium thiocyanate reagent (5 M guanidine Gray, a 300-bp fragment encoding spinach cytochrome b6 from Dr. H. Bohnert, and a 920-bp fragrnent encoding maize LHCP I1 thiocyanate, 25 mM Tris, pH 8.0, 2% Sarkosyl, 10 mM EDTA, 0.1 M 0-mercaptoethanol, and 4 mg/mL diethyldithiocarbamate) at 2 and a 900-bp fragment encoding maize ribulose-l,5-bisphosphate mL/g, wet tissue weight. This suspension was then centrifuged carboxylase/oxygenase small subunit from Dr. W. Taylor. at 8000 rpm for 20 min in an HB4 rotor (Sorvall) to pellet the cell The DNA inserts were cut out of the plasmid by the appropriate debris. After discarding the pellet, the nucleic acids were precipi- restriction enzymes and separated by agarose gel electrophore- sis. These inserts were visualized by ethidium bromide staining tated from the supernatant by addition of 0.03 volumes of 3 M sodium acetate and 0.75 volume of ethanol, and the suspension and excised and recovered as directed in the commercial product was placed at -20°C overnight. The pellet from this precipitation Gene Clean (Bio-101, La Jotla, CA). lsolated insert DNA was quantified by optical absorbance at 260 nrn. was resuspended in 50 mM Tris-HCt, pH 8.0, 1O mM EDTA, 1O0 mM NaCI, 0.2% SDS, and 3 mg/mL diethyldithiocarbamate at the lsolated DNA inserts were radiolabeled with 3zPby using either a commercial nick-translation kit (Amersham) or by hexamer la- level of 1 mL/g of starting tissue. This solution was extracted twice with 2 volumes of phenol/chloroform/isoamyl alcohol beling using a commercial random primer kit. Specific activity for (25:25:1) and then twice with 2 volumes of chloroform/isoamyl the probes was at teast 10' cprnlpg of DNA. alcohol (24:l). The pH of the final aqueous phase was adjusted to 5.0 by the addition of glacial acetic acid. Nucleic acids were precipitated from this solution by addition of NaCl to 200 mM and lmmunoprecipitationfrom Whole Lemna Plants 0.6 volume of isopropyl alcohol, and the suspension was placed at -20°C overnight. The precipitate was collected by centrifuga- Whole plant tissue (about 1O intact Lemna plants containing about tion at 8000 rpm for 20 min in an HB4 rotor. The pellet was 30 pg of chlorophyll) was placed in a small (0.5-mL) hand-held resuspended in 0.1% SDS to the level of 1 mL/g of starting tissue. ground-glass homogenizer and 320 pL of a cold solution contain- The RNA was precipitated selectively from this solution by the ing 0.3 M sucrose, 50 mM Tris-HCI buffer, pH 7.8, 10 mM NaCI, addition of MgCI2 to 1 mM and 0.25 volume of 8 M LiCI. The RNA and 5 mM MgC12 was added and homogenized to completion on was allowed to precipitate overnight at 4°C. The pellet was ice. Protease inhibitors (1 mM EDTA, 1 mM EGTA, 20 pM collected by centrifugation, as above, and resuspended in 1O mM a-aminocaproic acid, 20 pM benzamidine, 1O pg/mL aprotinin, Tris, pH 8.0, 0.1% SDS, 1.0 mM EDTA to the level of 1 mg/mL 200 pg/mL trypsin inhibitor, 10pg/mL pepstatin, and 20 mM of starting tissue. RNA was precipitated as above after addition phenylmethylsulfonyl fluoride) were added before homogeniza- of 0.1 volume of 3 M sodium acetate and 2 volumes of ethanol. tion. This homogenate was transferred to a 1.8-mL microcentri- This final pellet was resuspended in 1O mM Tris, pH 8.0, 1 .O mM fuge tube and 1.3 mL of cold acetone was added; the proteins EDTA to the level of 0.1 mL/g of starting tissue and stored at were allowed to precipitate for 1 hr at -20°C. After the precipi- -7OOC until needed. tation, the microcentrifuge tube was centrifuged at 4°C in a Cytochrome bsf Complex Assembly 21 1

microcentrifuge for 1O min, the supernatant was decanted care- SDS-PAGE fully, and the pellet was allowed to air dry. This pellet was resuspended with a microcentrifuge homoge- Analytical SDS-PAGE was carried out at 23OC with a slightly nizer in 200 pL of 2% SDS and heated to 55OC for 10 min. The modified Laemmli buffer system (Laemmli, 1970) using a 1.5-mm denatured protein was centrifuged for 20 min at top speed in a slab gel with a 4% stacking gel and a 13% to 21% gradient microcentrifuge at room temperature. The supernatant was re- resolving gel. Before electrophoresis, samples were solubilized moved and placed in a 5-mL polypropylene tube. The pellet was with 100 mM DTT, 1% SDS, 10 mM Tris-HCI buffer, pH 8.3, at washed with 1 mL of 4 x immunoprecipitation buffer (4% Triton 25OC for 3 hr or 5OoC for 15 min. Electrophoresis was carried out X-100,0.8% SDS, 20 mM EDTA, 800 mM NaCI, and 50 mM Tris- at 25°C with a constant current of 15 mA for 12 hr to 14 hr. After CI, pH 7.5). The suspension was centrifuged at top speed in a electrophoresis, the gel was either fixed and stained with Coom- microcentrifuge and the supernatant was carefully combined with assie Brilliant Blue or treated with En3Hance and dried on What- the first supernatant. The volume was brought up to 4 mL with man No. 3MM paper using a Hoefer gel dryer for fluorography. distilled water diluting the SDS concentration to about 0.1'/O. This total protein extract (TPE) was first clarified using preimmune serum. Preimmune serum (50 pL) was added to the TPE and the lmmunoblotting solution was rocked overnight at 4°C. Bio-Rad protein A-agarose beads were added to the TPE (twice Protein gel blotting was carried out as described in the Bio-Rad the volume of preimmune serum), and the suspension was incu- alkaline phosphatase instruction manual with the following modi- bated for 1 hr to 2 hr while rocking at 4°C. The beads were fications: 0.1-pm nitrocellulose paper (Schleicher & Schuell) was collected by centrifugation at 6500 rpm for 10 min in an HB4 rotor used, 0.5% SDS was included in the transfer buffer, Carnation (Sorvall). TPE supernatant was removed, leaving behind the pro- nonfat dry milk (3%) was used as a blocking reagent, and lZ5l- tein A-agarose bead pellet. This clarified TPE was transferred to protein A (ICN, Costa Mesa, CA) and autoradiography were used a 5-mL tube, and 25 FL of the specific immunized serum was as the visualization process. added and the solution rocked overnight at 4°C. The specific antigen-antibody complex was precipitated by incubation with Bio-Rad protein A-agarose beads (PAB) and subsequent Other Methods centrifugation. After centrifugation, the TPE supernatant was removed from Chlorophyll determinations were made using 80% acetone with the pellet of protein A-agarose beads, and the pellet was washed the extinction coefficients of Arnon (1949). with 1 mL of immunoprecipitation wash buffer: 0.15% Triton X- 100, 0.03% SDS, 5 mM EDTA, 300 mM NaCI, and 50 mM Tris- CI, pH 7.5. This wash step was repeated in the following order: 3 X 1 mL of wash buffer, 1 x 1 mL of double-distilled H20, 3 x 1 ACKNOWLEDGMENTS mL of wash buffer, and 2 x 1 mL of double-distilled H20. In the final two washes, it was important to remove all liquid from the PAB pellet with a 25-pL Hamilton syringe. This work was supported in pari by a grant from the National This final "dry" PAB pellet was resuspended in twice the pellet lnstitutes of Health. The authors would like to thank Drs. Richard volume of 2 x sample buffer and heated at 55OC for 20 min. This Max Wynn and Toivo Kallas for their helpful comments and Donald denatured sample was then run on standard gel electrophoresis. Carlson for assistance in the growth of Lemna.

Pulse-Chase Labeling of lntact Lema Plants Received October 22, 1990; accepted November 28, 1990 Very high specific activity of labelingwas accomplished by labeling entire plants by floating them directly in a 1 :1 mix of fresh growth media (minus sucrose) and %3-TransLabel(>l O00 Ci/mmol, ICN, Costa Mesa, CA). A small number (30 to 50) of individual plants REFERENCES were washed and placed in a 2-mL beaker. A custom-designed plunger was placed on top of the plants to submerge them Arnon, D.I. (1949). Copper enzymes in isolated chloroplasts. completely in labeling media. The total volume of about 100 pL was just sufficient to submerge the plants completely. The plants Polyphenol oxidase in Beta vulgaris. Plant Physiol. 14, 1-1 5. were illuminatedfrom beneath with heat-filtered incandescent light Barkan, A., Miles, D., and Taylor, W.C. (1986). Chloroplast gene at a leve1 of 50 FE/m2/sec for 1 hr. At the end of labeling, the expression in nuclear, photosynthetic mutants of maize. EM60 plants were washed twice with double-distilled H20 and twice J. 5, 1421-1427. with growth media and returned to a fresh 2-mL beaker containing Bendall, D.S., Sanguansermsri, M., Girard-Bascou, J., and fresh growth media containing 20 mM cysteine and 20 mM Bennoun, P. (1986). Mutations of Chlamydomonas reinhardtii methionine and allowed to grow as described above. At each time affecting the cytochrome bf complex. FEBS Lett. 203, 31-35. point, five to 1O plants were removed and washed severa1 times, Bennoun, P.A., Masson, A., Piccioni, R., and Chua, N.-H. (1978). and the total proteins were extracted as described below in an Uniparental mutants of Chlamydomonas reinhardtii defective in immunoprecipitation procedure. photosynthesis. In Chloroplast Development, G. Akoyunoglou 212 The Plant Cell

and J.H. Argyroudi-Akoyunoglou, eds (New York: Elsevier/ Levine, R.P., and Goodenough, U.W. (1970). The genetics of North-Holland), pp. 721-726. photosynthesis and of the chloroplast in Chlamydomonas rein- Berends, T., Kubicek, Q., and Mullet, J. (1986). Localization of hardtii. Annu. Rev. Genet. 4, 397-408. the genes coding for the 51 kDa PSll chlorophyll apoprotein, Lincoln, J.E., Cordes, S., Read, E., and Fischer, R.L. (1987). apocytochrome b6, the 65-70 kDa PSI chlorophyll apoproteins Regulationof gene expression by ethylene during Lycopersicon and the 44 kDa PSll chlorophyll apoprotein in pea chloroplast esculentum (tomato) fruit development. Proc. Natl. Acad. Sci. DNA. Plant MOI.Biol. 6, 125-134. USA 84, 2793-2797. Capeillere-Blandin, C., and Ohnishi, T. (1982). lnvestigation of Malkin, R. (1988). Structure-function studies of photosynthetic the iron-sulfur clusters in some mitochondrial mutants of cytochrome bc, and b6f complexes. ISI Atlas of Science: Bio- Saccharomyces cerevisiae. Eur. J. Biochem. 122, 403-41 3. chemistry 1,57-64. Crivellone, M.D., Wu, M., and Tzagoloff, A. (1988). Assembly of the mitochondrial membrane system. Analysis of structural Malkin, R., and Posner, H.B. (1978). On the site of function of mutants of the yeast QH,-cytochrome c reductase. J. Biol. the Rieske iron-sulfur center in the chloroplast electron trans- Chem. 263,14323-14333. port chain. Biochim. Biophys. Acta 501, 552-554. Doyle, M.P., Li, L.-B., Yu, L., and Yu, C.-A. (1989). ldentification Mattoo, A.K., Pick, U., Hoffman-Falk, H., and Edelman, M. of a M, = 17,000 protein as the plastoquinone-bindingprotein (1981). The rapidly metabolized 32,000-dalton polypeptide of in the cytochrome bsf complex from spinach chloroplasts. J. the chloroplast is the “proteinaceous shield” regulating photo- , Biol. Chem. 264, 1387-1 392. system II electron transport and mediating diuron herbicide Gatti, D.L., Meinhardt, S.W., Ohnishi, T., and Tzagoloff, A. sensitivity. Proc. Natl. Acad. Sci. USA 78, 1572-1 576. (1989). Structure and function of the mitochondrial bcl com- Metz, J.G., Miles, D., and Rutherford, A.W. (1983). Character- plex-A mutational analysis of the yeast Rieske iron-sulfur ization of nuclear mutants which lack the cytochrome f/b-563 protein. J. MOI. Biol. 205, 421-435. complex. Plant Physiol. 73, 452-459. Girard, J., Chua, N.-H., Bennoun, P., Schmidt, G., and Delosme, Miles, D. (1982). The use of mutations to probe photosynthesis M. (1980). Studies on mutants deficient in the photosystem I in higher plants. In Methods in Chloroplast Molecular Biology, reaction centers in Chlamydomonas reinhardtii. Curr. Genet. 2, M. Edelman, R.B. Hallick, and N.-H. Chua, eds (Amsterdam: 215-221. Elsevier Biomedical Press), pp. 75-1 07. Hauska, G., Nitschke, W., and Herrmann, R.G. (1988). Amino acid identities in the three redox center-carrying polypeptides Ohad, I., Kyle, D.J., and Hirschberg, J. (1985). Light-dependent of the cytochrome bCl/b6 f complexes. J. Bioenerg. Biomembr. degradation of the Qb-protein in isolated pea thylakoids. EMBO 20,211-228. J. 4, 1655-1 659. Heinemeyer, W.J., Alt, J., and Herrmann, R.G. (1984). Nucleo- Rich, P.R. (1986). A perspective on Q-cycles. J. Bioenerg. Bio- tide sequence of the clustered genes for apocytochrome b6 and membr. 18, 145-156. b/f subunit 4 of the cytochrome complex in the spinach plastid Rock, C.D., Barkan, A., and Taylor, W.C. (1987). The maize chromosome. Curr. Genet. 8, 543-549. plastid psbB-psbf-petB-pet D gene cluster: Spliced and un- Japa, S., and Beattie, D.S. (1988). Cytochrome b is necessary spliced pet B and petC RNAs encode alternative products. Curr. for the assembly of subunit VI1 in the cytochrome b-cl complex Genet. 12, 69-77. of yeast mitochondria. Arch. Biochem. Biophys. 264, 231 -237. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Kamienietzky, A., and Nelson, N. (1975). Preparation and prop- Cloning: A Laboratory Manual, 2nd ed. (Cold Spring Harbor, erties of chloroplasts depleted of chloroplast coupling factor 1 NY: Cold Spring Harbor Laboratory). by sodium bromide treatment. Plant Physiol. 55, 282-287. Shahak, Y., Posner, H.B., and Avron, M. (1976). Evidence for a Laemmli, U.K. (1970). Cleavage of structural proteins during the block between plastoquinone and cytochrome f in a photosyn- assembly of bacteriophage T4. Nature 227, 680-685. thetic mutant of Lemna with abnormal flowering behavior. Plant Lam, E., and Malkin, R. (1985). Characterization of a photosyn- Physiol. 57, 577-579. thetic mutant of Lemna lacking the cytochrome b6f complex. Biochim. Biophys. Acta 810, 106-109. Stubbe, W., and Herrmann, R.G. (1982). Selection and mainte- nance of plastome mutants and interspecific genome/plastome Lemaire, C.J., Girard-Bascou, J., Wollman, F.-A., and Bennoun, hybrids in Oenothera. In Methods in Chloroplast Molecular P. (1986). Studies on the cytochrome bsf complex. 1. Charac- Biology, M. Edelman, R.B. Hallick, and N.-H. Chua, eds (Am- terization of the complex subunits in Chlamydomonas reinhard- sterdam: Elsevier Biomedical Press), pp. 149-1 65. tií. Biochim. Biophys. Acta 851, 229-238. Tobin, E.M. (1978). Light regulatiõn of specific mRNA species in Lemaire, C.J., Girard-Bascou, J., and Wollman, F.A. (1987). Characterization of the be/f complex subunits and studies on Lemna gibba L. G-3. Proc. Natl. Acad. Sci. USA 75, the LHC-kinase in Chlamydomonas reinhardtii using mutant 4749-4753. strains altered in the bs/f complex. In Progress in Photosyn- Willey, D.L., and Gray, J.C. (1988). Synthesis and assembly of thesis Research, Vol. IV, J. Biggins, ed (The Hague: Martinus the cytochrome b-fcomplex in higher plants. Photosynth. Res. Nijhoff/Dr. W. Junk Publishers), pp. 655-658. 17, 125-144.