Plant MolecularBiology11:95-107 (1988) © KluwerAcademicPublishers, Dordrecht - Printedin the Netherlands

95

Assembly of the barley light-harvesting chlorophyll a/b proteins in barley etiochloroplasts involves processing of the precursor on thylakoids Parag R. Chitnis, Daryl T. Morishige, Rachel Nechushtai 1 and J. Philip Thornber* Department of Biology and Molecular Biology Institute, University of California, Los Angeles, CA 90024, USA; IPresent address." Department of Botany, The Hebrew University of Jerusalem, Jerusalem 91904, Israel (*author for correspondence) Received29 January 1988; accepted in revised form 11 April 1988

Key words: chloroplasts, light-harvesting chlorophyll proteins, membrane protein assembly, plastid development, processing of precursor protein

Abstract A barley gene encoding the major light-harvesting chlorophyll a/b-binding protein (LHCP) has been sequenced and then expressed in vitro to produce a labelled LHCP precursor (pLHCP). When barley etiochloroplasts are incubated with this pLHCP, both labelled pLHCP and LHCP are found as integral thylakoid membrane proteins, incorporated into the major pigment-protein complex of the thylakoids. The presence of pLHCP in thylakoids and its proportion with respect to labelled LHCP depends on the developmental stage of the plastids used to study the import of pLHCP. The reduced amounts of chlorophyll in a chlorophyll b-less mutant of barley does not affect the proportion of pLHCP to LHCP found in the thylakoids when import of pLHCP into plastids isolated from the mutant plants is examined. Therefore, insufficient chlorophyll during early stages of plastid development does not seem to be responsible for their relative inefficiency in assembling pLHCP. A chase of labelled pLHCP that has been incorporated into the thylakoids of intact plastids, by further incubation of the plastids with unlabelled pLHCP, reveals that the pLHCP incorporated into the thylakoids can be processed to its mature size. Our observations strongly support the hypothesis that after import into plastids, pLHCP is inserted into thylakoids and then processed to its mature size under in vivo conditions.

Introduction The majority of proteins in higher plant chloroplasts are encoded by nuclear genes and are synthesized on cytoplasmic ribosomes as precursor proteins [31]. The precursors are slightly larger than the mature proteins found in the plastids, due to the presence of an N-terminal leader peptide [9]. Import of these precursor proteins by chloroplasts is a posttranslational and energy-dependent event [9]. The precursor of the small subunit of ribulose bisphosphate carboxylase, a nuclear-encoded

stromal protein, binds to the outer envelope of plastids prior to its import [12]. It is then proteolytically processed to its mature size in two steps in the stroma [29]. The precursor of plastocyanin, a thylakoid lumen protein, is also processed in two steps [15]. An exact description of each step involved in the import and assembly of any nuclear-encoded intrinsic thylakoid protein is not available. To further elucidate these steps, we are studying the assembly of the major light-harvesting complex of photosystem II (LHC lib) [6, 7, 20]. Like other nuclear-encoded proteins, the LHC IIb

96 apoproteins (LHCP) are synthesized as a precursor(s) (pLHCP) on cytoplasmic ribosomes [9]. In vitro synthesized pLHCP can be imported by isolated chloroplasts and this process is post-translational and energy-dependent [30]. Binding of pLHCP to the outer chloroplast envelope precedes its translocation across the envelope membranes [12]. It is then processed to the mature size, inserted into thylakoids, incorporated into LHC IIb and bound to chlorophyll and carotenoid molecules. The exact order of these events is not known. In our previous studies [6], when pLHCP, synthesized from a gene of Lemna gibba L., was incubated with barley etiochloroplasts, both labelled pLHCP and the mature protein derived from it were found in the thylakoids as integral membrane proteins. Furthermore, both polypeptides were found in LHC IIb, the pigment-protein complex. Therefore, we concluded that processing of pLHCP is not a prerequisite for its inclusion in the complex or for its insertion into the thylakoid membrane. Our observation raised two major questions: (a) What is the explanation for the unexpected presence of pLHCP in the thylakoids? Was it due to the use of an heterologous import system (L. gibba pLHCP was imported by barley plastids), or was it reflective of a natural step in the assembly of LHC lib that could be seen because the plastids used for import were not fully developed? The developmental stage of plastids used for import has been shown to affect the proportion of pLHCP to LHCP incorporated into the thylakoids [6]; (b) Is the precursor processed on the thylakoids? Our observation was consistent with the hypothesis that pLHCP, after import into chloroplasts, is inserted into thylakoids and then processed to its mature size. In the present paper we have tried to address these issues. We isolated and sequenced a barley gene coding for pLHCP and expressed it in vitro to produce labelled barley pLHCP which was used to study import into isolated barley plastids at different developmental stages. These experiments revealed that the presence of pLHCP in the thylakoids of barley etiochloroplasts that have been incubated with labelled pLHCP, is mainly due to the younger age of plastids used for import. We also chased labelled pLHCP, which had been incorporated into the

thylakoids of intact plastids, by futher incubation of the plastids with unlabelled pLHCP and found that the pLHCP incorporated into the thylakoids of intact plastids can be processed to its mature size.

Materials and methods

Molecular cloning and sequencing Unless otherwise mentioned, the methods used for subcloning and sequencing of the barley cab gene have been described previously [24]. A barley genomic library made in the lambda vector Charon 35, was screened using a barley cDNA probe for LHCP. Both the library and the probe were kind gifts from Prof. K. Apel, University of Kiel, FRG. Two clones containing sequences homologous to a cab cDNA were identified and one of them was further characterized. A 1.2 kb Eco RI-Acc I fragment of this genomic clone was exclusively found to contain a cab gene sequence and hence was cloned in the Eco RI-Acc I sites of the plasmid pGEM-blue (Promega Biotec, Madison, WI, USA), thereby creating pCab-2. Using the nuclease Bal-31, deletions were made from both the ends of this clone and the resultant fragments were cloned in the Sma I site of pGEM-blue. The complete nucleotide sequence of both the strands of the gene was derived by sequencing these deletions from each end using the chain-termination method modified for plasmids (Technical manual, GemSeq K/RT sequenceing system, Promega Biotec, Madison, WI, USA). One of the deletions, pGEM-Cab-2, which contains no ATG sequence in the 30 bp region upstream of the translation initiation site of the Cab-2 gene and has the complete coding region, was used for in vitro expression of Cab-2 gene (see below).

Transcription and translation In vitro expression of the AB30 gene of L. gibba to obtain 35S-methionine-labelled pLHCP was carried out according to methods described previously [20]. The Cab-2 gene product was obtained by transcribing the plasmid pGEM-Cab-2 that had been linea-

97 rized by digestion with Hind III. Its transcription was carried out with SP6 RNA polymerase in the presence of 0.25 mM diguanosine triphosphate (Pharmacia Inc., Piscataway, N J) [20, 21]. When the resulting RNA was electrophoresed on a denaturing agarose gel and stained with ethidium bromide, it was observed as a single band with a size of approximately 1 kb. This RNA was translated in a wheat germ system [28] in the presence of 35S-methionine (ICN Radiochemicals, Irvine, CA) to produce labelled pLHCP with about 40000 cpm per/zl translation product.

Plant material Barley (Hordeum vulgare L. cv, Prato) seedlings were grown in vermiculite at 25 °C in complete darkness. Plastids were isolated from etiolated 7-day-old seedlings that had been illuminated (30 #E (microeinstein) m -2 s -1) for periods indicated in the figure legends. Seedlings were cut about 3 - 4 cm above the ground and leaves were cut into 2 - 3 cm pieces before grinding. The chlorina f2 mutant plants were grown in exactly the same way as the wild type plants. The mutant seeds were obtained originally from Dr H. R. Highkin, California State University, Northridge, CA.

In vitro import of pLHCP by isolated barley plastids Isolated barley plastids were incubated either with labelled barley or L. gibba pLHCP synthesized in vitro (see above), treated subsequently with thermolysin to degrade proteins not imported by plastids, and then thylakoids were obtained from intact plastids as described previously [6]. The thylakoids were denatured and their proteins were separated by polyacrylamide gel electrophoresis. The gels were then dried and autoradiographed.

Other techniques Methods used for chlorophyll and protein determi-

nations, for treatment of thylakoids with 0.1 M NaOH, for protein fractionation and for quantitation of radioactive protein bands have been described previously [6, 8]. Immunoprecipitations of pLHCP's from the labelled translation products were performed using staphylococci cells [19].

Chemicals Glycerol and inorganic chemicals were purchased from Mallincrodt Inc., St. Louis, MO. All molecular biology grade reagents and enzymes were obtained from Bethesda Research Laboratories, Gaithersburg, MD, while other chemicals were from Sigma Chemical Co., St. Louis, MO.

Results

Structure and in vitro expression of barley Cab-2 gene A barley genomic library in a lambda vector, Charon 35, was screened using a barley cDNA clone [13] as a probe and two clones containing cab gene(s) were isolated. One of these clones has been further characterized. A 6 kb Eco RI fragment of this clone contains a sequence coding for pLHCP. We sequenced a 1.1 kb long, LHCP sequence-containing fragment derived from this clone. The nucleotide sequence of this gene and the amino acid sequence deduced from it are shown in Figure 1. The barley Cab-2 gene does not contain an intron and belongs to the type I gene group of cab genes. The sequence has one open reading frame starting with a methionine and containing 264 amino acids up to the translation termination codon, TAA. Within the LHCP coding region, barley Cab-2 gene and L. gibba AB30 gene are highly homologous (90%), but about thirty-four amino acids in the transit sequence and the twenty amino-terminal residues of the mature sequence of Cab-2 show less homology to the corresponding parts of AB30 gene. In fact, in these regions barley Cab-2 gene product is more similar to products of the ab 1.6 gene of wheat [22] or of a gene of maize [25]. Another striking feature of the Cab-2

98 1 AGGATACCAA TTATGGAAAA CATAAGC~rAT A~TCCAA'CAC CGATT'~GAAC 1 CACAATCCGG

61 ATACGAAATA CGATCGATGATACCTATTCGATGACCAAAT CCCGATTAGATCGGCTATTA

121 CGGACGCCGTTGGGTCTGTA"TTACTAGCAT GGTTACGTACCTACCTTGAGTTAAGGGCGA 181 ATGGCCGCCGCGACCATGGCTCTGTCTTCCTCCACCTTCGCCGGGAAGGCGGTGAAAAAC M A

A

A

T

14 A

L

S

S

S

T

F

A

G

K

A

V

K

N

2,',1 CTGTCTTCCTCGTCAGAGGTGCAGGGCGACGCCAGAGTCTCAATGAGCd~ GACCGCCOCC 301 ACCAAGAAGG TCGGCAGCCCCTGGTACGGCCCGGACCGGGTCAAGTACCT GGGCCCCTTT [.

S S

S

S E

V

Q G D

A

R V

S M R K

,L

T

A

A

T

K

K

V G S P

W ¥

G

P D R

V

K Y

t.

G P

F

361 AGCGGGGAGTCCCCGTCCTACCTGACCGGCGAGTTCCCCGGAGACTACGGGTGGGACACG 1.21 GCCGGGCTCAGCGCCGACCCCGAGACCTTCGCCAAGAACC GCGAGCTGGAGGTGATCCAC S

G E

S

P S Y

L

T

G

E

F

P

G 0

Y G

W D

T

A

G L

S A D P

E

T

F

A

K N

R E

L

E

V

!

H

481 GGGCGCTGGGCGATGCTGGGGGCGCTCGGGTGCGTCTTCCCGGAGCTCCTC,C.-CCCGGAAC 541 GC.~GTCAAGTTTGGCGAGCaCCGTGTGGTTCAAGAAGGGTTCCCAGATCTTTAC.-CGII, GGG(: G R W

A M L

G

A

L

G

C V

F

P

E

L

L

A

R N

G V

K

F

G E A

V W F

I¢ K G

S

Q

!

F

S E

G

601 GGGCTGCAGTACCTGGGGAACCCCTCCCTGGTCCACGCGCACd~GCATCCTGC-CCATCTGG 661 GCCTGCCAGGTGGTCCTCATGGGTGCCGTCGkAGGCTACC GCGTGGCCGGCGGTCCCCTG G L

Q

Y

L

G N

P S

L

V

14 A

Q S

]

L

A

!

W

A

C Q

V V

L

M

G A V

E G Y

R V A

G

G P

L

7"21 GGCGAGGTGGTGGACCCCCTGTACCCCGGCGGCAGCTTTGACCCCCTCGGGCTCGCGGAC 781 GACGCGGAAGCCTTCGCCGAGCTGAAGGTCAAGGAGATCAAGAACGGACGCCTGGCCAT~ G E V

V

D P

L

Y P

G

G $

F

D P

L

G

L

A D

D A

E

A

F

A

E

L

K V

g

E

!

K

N G R

L

A M

841 TTCTCCATGTTCGGATTCTTCGTGCAC, GCC ATCGTCACCGGGAAGGGGCCCCTCGAGAAC 901 TTGGCGGACCATCTCGCCGACCCCGTAAACAACAATGCCT GC~CCTTCGCCACCAACTTC F

S M

F

G F

F

V Q A

]

V

T

G g

G P

L

E M

L

A

D

H L

A O

P V N

N

N A

W A

F

A

T

N

F

961 (;TCCCCGGCAAGTA,AA,ATCG GCACCCCTTGCCAAAATTGC TCGCCATGCACATTCCTTCA 1021 TTAATTAATA V

P

G

K

Fig. 1. The nucleotideand derivedamino acid sequenceof Cab-2. The completeamino acid sequencefor the precursorprotein encoded by the gene is given below the DNA strand. The filled triangle indicates the putative processing site while the box enclosesa region of homology in the 5'-upstream region of Cab-2, L. gibba AB genes and petunia cab genes (see [20]). The putative CAAT and TATTA box sequencesare underlined.

mature protein is that it lacks a sequence of three amino acids "SSG" found close to the amino terminus in LHCPs coded by type I genes from other species. This sequence is also largely absent in protein sequences derived from type II cab genes [18, 32]. A monoclonal antibody raised against tobacco L H C P recognizes LHCPs from dicotyledons but not from most of the monocotyledonous plant species examined [34]. The similarity between the amino terminal regions of wheat, maize and barley sequences and their distinctness from the other type I sequences may be significant in deducing the immunological site of this antibody. The transit peptide of Cab-2 protein exhibits the three blocks of homology common to chloroplast transit sequences [17]. The upstream region of this gene contains putative TATA and CAAT boxes (Fig. 1). A region of homology found on the 5' side of CAAT box among cab genes o f L . gibba and petunia, first noted by Kohorn et al. [20], is also present in the barley sequence. The significance o f this sequence is, however, not known. A subclone o f Cab-2 gene (Fig. 1) was obtained

by deleting some nucleotides from the 5' region using Bal 31 nuclease. It was inserted into the Sma I site of pGEM-blue to create the plasmid pGEMCab-2. Transcription of Hind III-digested pGEMCab-2 with SP6 polymerase produced about 1 kb long RNA containing an L H C P transcription unit. This RNA was translated in a wheat germ extract to produce 35S-methionine-labelled p L H C P (Fig. 2, lane b) that had about the same apparent size (29.5 kDa) observed for the p L H C P in barley poly(A)RNA translation products [2]. The predicted molecular weight of p L H C P from the sequence of Cab-2 gene is 28049. This size is lower than the 29.5 kDa predicted from its mobility in denaturing gels. The Cab-2 translation products are immunoprecipitated by a polyclonal antibody against L H C P (Fig. 2, lane d).

Uptake of p L H C P by barley plastids Labelled p L H C P obtained by in vitro translation of

99

Fig. 2. In vRro expression of Cab-2 gene. The autoradiograph shows labelled proteins separated on a 12.5% SDS-polyacrylamide gel. The wheat germ translation products of in vitro transcribed RNA from L. gibba AB30 (lane a) and barley Cab-2 (lane b) genes were immunoprecipitated with a polyclonal antibody raised against LHCPs of L. gibba (lanes c and d, respectively). Pre-immune serum was used to precipitate the Cab-2 translation product (lane e). The Cab-2 translation products were incubated with barley etiochloroplasts isolated from plants that had been illuminated for 12 h and fractionated into envelopes (lane f), stroma (lane g) and thylakoids (h, 1). One batch of thylakoids isolated from plastids used for import, was treated with 0.1 M NaOH (lane i) while another batch was treated with 100 #g/ml trypsin (lane m). Solubilizedthylakoids obtained from plastids that have imported barley pLHCP were fractionated into chlorophyll-proteincomplexesby partially denaturing PAGE. The protein subunits in fractionated LHC IIb (lane j) and photosystem I (lane k) complexeswere resolved by electrophoresis on a completely denaturing polyacrylamidegel. The molecular weight markers (row) in kDa are shown.

Cab-2 RNA, was incubated with plastids isolated

from etiolated barley plants that had been illuminated for 12 h. After a 60 min incubation, thermolysin was added to the reaction mixture to digest any protein that was not located within the plastids; proteins inside intact plastids remain undigested since thermolysin does not cross the envelope [11]. Intact plastids were then separated from those broken during isolation and incubation, and were fractionated into stroma and thylakoids. The results of uptake of p L H C P by barley chloroplasts are shown in Fig. 2. The proteins from different fractions were separated by PAGE and dried gels were subjected to autoradiography as well as to quantitation using the Ambis beta scanning system. The stromal or envelope fractions did not contain any labelled polypeptide (Fig. 2, lanes f, g); however, two labelled polypeptides were observed in the thylakoid fraction (Fig. 2, lane h). The electrophoretic mobility of one of these two polypeptides corresponded with p L H C P while the other corresponded with the slowest migrating of the three L H C P s of barley. When the thylakoids obtained after import o f p L H C P were treated with 0.1 M N a O H , neither of the two labelled polypeptides found in the thylakoids was removed (Fig. 2, lane i). Thus, both p L H C P and L H C P are present as integral thylakoid proteins.

To determine whether the labelled polypeptides have been incorporated into L H C IIb, we fractionated solubilized thylakoids, isolated after an uptake reaction, in a partially dissociating polyacrylamide gel electrophoresis system (G. E Peter and J. P. Thornber, manuscript in preparation). In this system all detectable chlorophyll remains bound to protein. We excised the chlorophyll-protein bands corresponding to L H C IIb and photosystem I, and fractionated their apoproteins by completely denaturing gel electrophoresis. The L H C IIb band contained two radioactive polypeptides with electrophoretic mobilities corresponding to p L H C P and L H C P (Fig. 2, lane j). The photosystem I band, however, did not contain any labelled polypeptide (Fig. 2, lane k). Therefore, both L H C P and p L H C P were present specifically in L H C IIb. Trypsin has been shown to cleave a 2 kDa segment from the amino terminus of the mature L H C P in pea thylakoids [33, 26]. Similar results have been obtained for L. gibba [20] and barley [6]. We wanted to test whether the labelled precursor in the thylakoids is in the same conformation as LHCP. Therefore, thylakoids obtained from plastids that had imported pLHCP, were treated with trypsin (100 /~g/ml) for 10 min at 37 °C. After this treatment the thylakoids contained two labelled polypeptides

100 about 2 and 4 kDa smaller than the native size (26- 27 kDa) of LHCP (Fig. 2, lane m). The former contained 86% of the radioactivity present in the two bands. In the untreated thylakoids (lane l) most of the radioactivity (88%) is contained in LHCP. Therefore, it seems more likely that the major protein band in the trypsin treated thylakoids was derived from LHCP; furthermore, the size of the major digestion product was that expected from the trypsin-treatment of native LHCP. The minor, smaller band was therefore probably derived from pLHCP and the greater reduction in size caused by trypsin digestion would indicate that pLHCP in the membrane was not in precisely the same conformation as the native LHCP. In any case, the protection of labelled polypeptides from complete digestion by trypsin shows that they are largely embedded in the membrane. Thus, when labelled barley pLHCP is imported into barley etiochloroplasts, both labelled pLHCP and LHCP are found as integral thylakoid proteins, incorporated into LHC IIb.

Import of pLHCP by developing barley plastids In our previous studies, it was shown that barley plastids at different developmental stages show differences in their ability to import L. gibba pLHCP and assemble it into LHC IIb [6]. To test if a similar trend is observed when barley plastids import barley pLHCP, we illuminated one-week-old etiolated barley seedlings for various times. The plastids isolated from these plants were then incubated with labelled barley pLHCP or L. gibba pLHCP and the thylakoids were subsequently isolated. Half of the thylakoids from plastids incubated with barley pLHCP were treated with 0.1 M NaOH to remove peripheral membrane proteins. During the time of greening used in our experiment the chlorophyll content of the plastids increased over 100-fold while changes in protein content of plastids were relatively minor (data not shown). Thus, to compare the amounts of pLHCP and LHCP incorporated in the thylakoids of plastids isolated from etiolated plants illuminated for different lengths of time, we loaded equal

amounts of protein on each lane. Marked differences were observed in the relative amounts of pLHCP and LHCP present in thylakoids regardless of whether they had been incubated with barley or L. gibba pLHCP (Fig. 3, a and b). The precursor, but hardly any processed LHCP, was detected in thylakoids isolated either from plastids obtained from etiolated leaves or from those greened for 2 h (Fig. 3a). Labelled, processed LHCP was easily detected in plastids isolated from etiolated plants that had been exposed to light for 5 h. Although import and assembly of both barley pLHCP and L. gibba pLHCP by barley plastids showed similar trends, the ratio of pLHCP to LHCP in thylakoids of plastids that had been incubated with barley pLHCP was lower at all stages of plastid development (Fig. 3b). The NaOH treatment of thylakoids isolated from plastids that had been incubated with barley pLHCP, removed varying amounts of pLHCP from the thylakoids depending on the stage of plastid development (Fig. 3c). The pLHCP in the thylakoids of plastids obtained from completely etiolated plants was not resistant to the NaOH treatment while in the thylakoids of the plastids isolated from plants greened for 12 h, pLHCP was completely resistant to NaOH treatment. The NaOH treatment did not remove any labelled LHCP from the thylakoids of plastids at any stage of development (Fig. 3c). These findings are similar to those reported for the uptake of L. gibba pLHCP by developing barley plastids [8].

Uptake of pLHCP by chlorophyll b-less barley When plastids at different developmental stages were used for import of pLHCP, the proportion of pLHCP to LHCP integrated into thylakoids of intact plastids varied (Fig. 3b). During plastid development, the chlorophyll content also changes dramatically (cf. [6]). Chlorophyll is thought to stabilize LHCP in the thylakoids [3-5], and therefore the variation in the proportion of pLHCP to LHCP may be due to changes in chlorophyll content during light-triggered plastid development. To test this possibility we used the chlorina f2 mutant of barley, completely green tissue of which contains about half

101

Fig. 3. Import of pLHCP by barley plastids at different stages of development. (a) The autoradiograph showing dissociated thylakoid proteins separated on 14°70 polyacrylamide gels. Plastids were isolated from etiolated barley plants illuminated for different periods, as shown in the figure, and were incubated with either L. gibba AB30 (L) or barley Cab-2 (B) translation products. Thylakoids obtained after import were either treated (+) or not treated ( - ) with 0.1 M NaOH. (b) The ratio of total counts in pLHCP to those in LHCP in the untreated thylakoids isolated after uptake of either L. gibba AB30" (filled triangles) or barley Cab-2 (open squares) translation products by barley plastids obtained from etiolated plants greened for different times. The ratios were calculated after quantitation of gels described above using an Ambis beta scanning system. (c) The counts in NaOH-wash-resistant pLHCP (filled circles) or LHCP (filled squares) as a percentage of the total counts in pLHCP or LHCP in the thylakoids isolated after uptake of barley pLHCP by plastids at different stages of development.

102 of the amount of chlorophyll present in wild type barley and almost no chlorophyll b [16]. We illuminated etiolated plants of wild type and the mutant strains of barley for 12 h and isolated their plastids. After incubation of each plastid type with barley pLHCP, the thylakoids isolated from both types of plastids contain pLHCP and LHCP in almost the same proportion (Fig. 4). On the other hand, the wildtype and the mutant plants at the 12 h greening stage have significant differences in their chlorophyll contents (Fig. 4). Thus, the presence of only 63% of the total chlorophyll and very little chlorophyll b in the mutant plastids in comparison to the wild type plastids, does not decrease the proportion of pLHCP to LHCP in the thylakoids. Therefore, it is unlikely that the scarcity of chlorophyll during the early stages of greening results in a higher ratio o f p L H C P to LHCP.

Fig. 4. Import of barley pLHCP by chlorophyll b-less barley plastids. The chlorophyll contents and import studies were done for etiolated plants of wild type and chlorophyll b-less mutant strains of barley illuminated for 12 h. The autoradiograph shows labelled thylakoid proteins of wild type and chlorophyll b-less mutant strains, isolated from plastids used for import of barley pLHCP.

pLHCP integrated into thylakoids is processed on the thylakoids The data presented so far clearly show that pLHCP is integrated into thylakoids even in an homologous uptake system involving import of barley pLHCP by barley plastids. Although this observation is consistent with the hypothesis that pLHCP is processed to its mature size on thylakoids, it is also possible that, under our import conditions, some pLHCP can get into thylakoids of intact plastids as an artifact and the labelled LHCP seen in the thylakoids is a product of the normal assembly process that involves processing of pLHCP prior to its insertion into thylakoids. To test this possibility, pulse-chase experiments were performed using plastids isolated from etiolated seedlings that had been illuminated for 8 h. At this stage of development appreciable amounts of pLHCP are found in thylakoids during import experiments (Fig. 3). The plastids are, however, not fully efficient in integrating pLHCP into the thylakoids [8]. In the pulse-chase experiments import reactions were terminated at different times to study the course of appearance of pLHCP and LHCP in the thylakoids (Fig. 5). The ratio of pLHCP to LHCP decreased with time indicating that the rates of import of pLHCP and its processing are not equal (Fig. 5b, open circles). Both labelled pLHCP and LHCP were present in the thylakoids after 5 min of import (Fig. 5a). The amount of pLHCP in thylakoids initially increased rapidly and then stabilized (Fig. 5c, filled triangles) while the amount of labelled LHCP continued to increase until 90 min of incubation (Fig. 5c, filled squares). In pulse-chase treatments, plastids were incubated with labelled pLHCP for 10 min, then briefly pelleted by centrifugation for 1 min at 4340 g, and the pellet resuspended in a solution containing identical ingredients to the original uptake reaction except that the labelled translation products were substituted by unlabelled ones. The uptake reactions with the unlabelled pLHCP were carried out for different periods and the amounts of pLHCP and LHCP in the thylakoids were quantitated (Fig. 5). The proportion of pLHCP to LHCP decreased and reached zero after 2 h (Fig. 5b, open circles). The amount of pLHCP in thylakoids decreased during the chase time and af-

103

Fig. 5. Chase of labelled barley pLHCP in thylakoids of intact barley plastids by unlabelled barley pLHCP. (a) An automdiograph showing dissociated thylakoid proteins separated by electrophoresis through 12.5°70 polyacrylamide gel. Thylakoids were obtained from plastids of 7-day-old etiolated plants greened for 8 h, and incubated with barley Cab-2 translation products for different times (0, 5, 10, 30, 60, 90, 120 rain). In chase treatments the plastids were incubated with labelled pLHCP for 10 min and then with unlabelled pLHCP for 20 (30'), 50 (60'), 80 (90') or 110 (120') rain. The thylakoids from an equal number of plastids were loaded into each lane. (b) The ratio o f counts in pLHCP to those in LHCP in thylakoids isolated from plastids that were incubated with labelled pLHCP for different periods (filled circles) or with labelled pLHCP for 10 rain followed by chase with unlabelled pLHCP (open circles). (c) The total counts in pLHCP (triangles) or LHCP (squares) in thylakoids isolated from plastids that had been incubated with pLHCP for different times (filled symbols) or with labelled pLHCP for 10 min followed by chase with unlabelled pLHCP (open symbols).

104 ter 80 min there was very little, if any, pLHCP left in the thylakoids (Fig. 5c, open triangles). On the contrary, there was a significant and reproducible increase in the amount of labelled LHCP during the chase (Fig. 5c, open squares). Since the number of labelled methionines in LHCP in considerably less than those in the pLHCP polypeptide, the increase in the amount of labelled LHCP was proportional but not equal to the decrease in the amount of pLHCP, indicating a precursor-product relationship between pLHCP and LHCP in the thylakoids. Therefore, the pLHCP integrated into thylakoids of intact plastids that had been incubated with labelled pLHCP is processed to LHCP while on the thylakoids. There was a decline in the amount of LHCP after ll0 min of chase. At this time there was no pLHCP left in the thylakoids and so the synthesis of new LHCP would cease while the degradation of LHCP would still continue.

Discussion

Presence of pLHCP in thylakoids Import of pLHCP into isolated plastids has been studied by many researchers; however, in most studies pLHCP is not detected inside the plastids, either in the stroma or in the thylakoids [4, 12, 30]. These studies include uptake of poly(A) ÷ RNA translation products by plastids isolated from the same species from which the RNA was isolated (homologous system). Furthermore, the plastids used in such uptake experiments were isolated from plants grown in continuous light. We detected pLHCP integrated into LHC II of the thylakoids [6] in previous studies; however, our import system used L. gibba pLHCP taken up by barley etiochloroplasts (heterologous system). The differences between the import systems in the source of the RNA and/or the developmental stage of plastids may explain the detection of pLHCP by us and not by most other workers. In the present studies, we could still detect the precursor when barley pLHCP was imported by barley plastids (Fig. 2, lane h). It should be noted, however, that barley plastids are more efficient in utilizing barley pLHCP than in utilizing L. gibba pLHCP (Fig. 3b). Therefore, it was not the heterolo-

gous import system we used in our previous studies that was solely responsible for our observing the presence of precursor in thylakoids. Our import system uses etiochloroplasts, incompletely developed plastids, which may explain why we detected pLHCP in the thylakoids. Data in Fig. 3 support this notion; during the early stages of development appreciable amounts of pLHCP and very little LHCP were present in the thylakoids isolated from plastids used for import of either L. gibba or barley pLHCP. As the plastids developed, the proportion of pLHCP to LHCP in the thylakoids decreased and after 24 h of greening, only LHCP was present in the thylakoids of the plastids that had been incubated with barley pLHCP. This latter situation is similar to that observed by other researchers who used completely developed chloroplasts [4, 12, 30].

Import of pLHCP and its assembly during plastid development Chloroplast biogenesis involves many morphological and biochemical changes triggered by light [1, 23]. When etioplasts are exposed to light, not only is the synthesis of pLHCP and chlorophyll triggered but also the machinery for import and processing of pLHCP. In one of our previous studies, we used a heterologous import system and found that the proportion of labelled pLHCP to LHCP in the thylakoids depends on the developmental stage of plastids used for import of pLHCP [6]. In the present paper we used a homologous system and still observed pLHCP in the thylakoids. Its proportion with respect to labelled LHCP was also dependent on the stage of plastid development (Fig. 3). The importance of the developmental stage of plastids for their capability to import and process pLHCP can be explained in several ways. It is possible that factors involved in assembly of LHC II may not be fully functional or even present at the earlier stages of greening. One such factor has been already reported [8]: insertion of pLHCP into isolated thylakoids requires the presence of some stromal protein [7, 10], the appearance of which is dependent on the stage of plastid development [8]. The receptivity of thylakoids for insertion of pLHCP is also

105 developmentally regulated; at early stages of greening plastid membranes are significantly less receptive for the insertion of pLHCP [8]. Another explanation for the influence of plastid development on the import and assembly of pLHCP that we had suspected previously [6] is the availability of chlorophyll, especially of chlorophyll b, which has been proposed to stabilize LHCP in the thylakoids [3-5]. At early stages of greening less chlorophyll is available for binding to LHCP [1] and this could be a reason for finding very little, if any, LHCP in the thylakoids at those stages. However, the greatly reduced quantities of chlorophylls, chlorophyll b in particular, in plastids of a chlorophyll b-less barley mutant strain does not affect the pLHCP to LHCP ratio in the thylakoids of plastids that had been incubated with barley pLHCP (Fig. 4). Therefore, synthesis of chlorophyll does not seem to be responsible for the developmental pattern seen for pLHCP/LHCP ratio in the thylakoids. The possibility that the synthesis or activity of the processing enzyme for pLHCP could be under the control of light and/or plastid development, remains to be tested.

Assembly of LHC lib Like other thylakoid proteins, the targeting of pLHCP and its assembly into LHC II is more complicated than the similar processes for the stromal or envelope proteins, due to the fact that sites of membrane translocation (envelope) and membrane insertion (thylakoids) are different and are separated by an aqueous stroma. Moreover, the precursor is water-soluble while the mature LHCP is an integral membrane protein. These aspects require serious consideration when proposing a mechanism for targeting and assembly of LHC II. We propose the following steps occur during the formation of LHC lib from pLHCP. After its translocation across the chloroplast envelope, pLHCP is modified by a stromal protein factor [7, 10] so that it can now be inserted into thylakoids. The pLHCP integrated into thylakoids is incorporated into LHC IIb, the pigment-protein complex. It is also processed to its mature size on the thylakoids. Processing is not a

prerequisite for its integration into thylakoids or for its incorporation into the complex ([6], Fig. 2). We believe that integration of pLHCP into thylakoids and its subsequent processing there also occur in vivo for the following reasons. First, in our in vitro import system, pLHCP can be detected in thylakoids. The validity of this observation is proven by the detection of pLHCP in LHC lib and also by the developmental regulation of the amount of pLHCP integrated into thylakoids ([6], Figs. 2 and 3). Second, in the plastids, there exists a machinery to insert pLHCP into thylakoids and the appearance of this machinery is dependent on the stage of plastid development [8]. Third, the pLHCP integrated into thylakoids of intact plastids can be processed to LHCP on the thylakoids (Fig. 5). Fourth, mature LHCP is not a water-soluble protein and so it is reasonable to expect it to be processed at or near its final destination if a function of its transit peptide is to make it in some way water-soluble. There is evidence that another thylakoid membrane protein, albeit one synthesized within the chloroplast, is processed after its insertion into the membrane [14]. In summary, our results demonstrate that the precursor of LHCP can be inserted into thylakoid membranes as a member of the LHC II complex when either barley or L. gibba pLHCP is imported by barley plastids. The presence of pLHCP in thylakoids is detected mainly because the import and assembly machinery is incompletely developed in immature plastids. Its amount, but not its presence, is dependent on whether the pLHCP used for uptake into barley plastids is from a barley or L. gibba gene. The proportion of pLHCP to LHCP found in the thylakoids is dependent on the developmental stage of plastids and this dependence is not due to the insufficient chlorophyll at the early stages of greening. The precursor in the thylakoids can be processed to the mature form and this process could normally occur in vivo.

Acknowledgements We thank Vaishali P. Chitnis and Prakash K. Shah for their input during the preparation of this manuscript. We also thank Dr K. Apel for sending us a

106 barley genomic library and a barley cDNA clone for LHCP. We are indebted to Dr E. M. Tobin for providing facilities for some of the early part of studies reported in this paper. This work was supported by grants to J.P.T. from the National Science Foundation and the United States Department of Agriculture. P.R.C. and D.M. were supported by McKnight Foundation graduate fellowships and R.N. by a Weizmann fellowship.

References 1. Akoyunoglou G, Argyroudi-Akoyunoglou J: Posttranslational regulation of chloroplast differentiation. In: Akoyunoglou G, Senger H (eds) Regulation of Chloroplast Differentiation, pp. 571- 582. Alan R Liss, New York (1986). 2. Apel K: Phytochrome-induced appearance of mRNA activity for the apoprotein of the light-harvesting chlorophyll a/bprotein of barley (Hordeum vulgate). Eur J Biochem 97" 183-188 (1979). 3. ApelK, KloppstechK: Theeffect of light on the biosynthesis of the light-harvesting chlorophyll a/b protein. Evidence for the requirement of chlorophyll a for the stabilization of the apoprotein. Planta 150:426-430 (1980). 4. Bellemare G, Bartlett SG, Chua N-H: Biosynthesis of chlorophyll a/b-binding polypeptides in wild type and the chlorina f2 mutant of barley. J Biol Chem 257:7762-7767 (1982). 5. Bennett J: Biosynthesis of the light-harvesting chlorophyll a/b protein. Polypeptide turnover in darkness. Eur J Biochem 118:61-70 (1981). 6. Chitnis PR, Harel E, Kohorn BD, Tobin EM, Thornber JP: Assembly of the precursor and processed light-harvesting chlorophyll a/b protein of Lemna into the light-harvesting complex II of barley etiochloroplasts. J Cell Biol 102: 982-988 (1986). 7. Chitnis PR, Nechushtai R, Harel E, Thornber JP" Some requirements for the insertion of the precursor of apoproteins of Lemna light-harvesting complex II into barley thylakoids. In: Biggins J (ed.) Progress in Photosynthesis Research, Vol. 4, pp. 573-576. Martinus Nijhoff, Dordrecht (1987). 8. Chitnis PR, Nechushtai R, Thornber JP: Insertion of the precursor of the light-harvesting chlorophyll a/b-protein into the thylakoids requires the presence of a developmentally regulated stromal factor. Plant Mol Biol 10:3-12 (1987). 9. Chua N-H, Schmidt GW: Transport of proteins into mitochondria and chloroplasts. J Cell Biol 81:461-483 (1979). 10. Cline K: Import of proteins into chloroplasts. Membrane integration of a thylakoid precursor protein reconstituted in chloroplast lysates. J Biol Chem 261: 14804-14810 (1986). 11. Cline K, Werner-Washburne M, Andrews J, Keegstra K."Thermolysin is a suitable protease for probing the surface of intact pea chloroplasts. Plant Physiol 75:675-678 (1984).

12. Cline K, Werner-Washburne M, Lubben TH, Keegstra K: Precursors to two nuclear-encoded chloroplast proteins bind to the outer envelope membrane before being imported into chloroplasts. J Biol Chem 260: 3691-3696 (1985). 13. Gollmer I, Apel K: The phytochrome-controlled accumulation of mRNA sequences encoding the light-harvesting chlorophyll a/b protein of barley (Hordeum vulgare L.). Eur J Biochem 133: 309- 313 (1983). 14. Grebanier AE, Coen DM, Rich A, Bogorad L: Membrane proteins synthesized but not processed by isolated maize chloroplasts. J Cell Biol 78:734-746 (1978). 15. Hageman J, Robinson C, Smeekens S, Weisbeek P: A thylakoid processing protease is required for complete maturation of the lumen protein plastocyanin. Nature 324:567-569 (1986). 16. Highkin HR: Chlorophyll studies on barley mutants. Plant Physiol 25:294-306 (1950). 17. Karlin-Neumann GA, Tobin EM: Transit peptides of nuclearencoded chloroplast proteins share a common amino-acid framework. EMBO J 5 : 9 - 1 3 (1986). 18. Karlin-Neumann GA, Kohorn BD, Thornber JP, Tobin EM: A chlorophyll a/b-protein encoded by a gene containing an intron with characteristics of a transposable element. J Mol Appl Genet 3 : 4 5 - 6 1 (1985). 19. Kessler SW: Use of protein A-bearing staphylococci for the immunoprecipitation and isolation of antigens from cells. Meth Enzymol 73:442-446 (1981). 20. Kohorn BD, Harel E, Chitnis PR, Thornber JP, Tobin EM: Functional and mutational analysis of the light-harvesting chlorophyll a/b protein of thylakoid membranes. J Cell Biol 102:972-981 (1986). 21. Kohorn BD, Tobin EM: Chloroplast import of lightharvesting chlorophyll a/b-proteins with different amino termini and transit peptides. Plant Physiol 82:1172-1174 (1986). 22. Lamppa GK, Morelli G, Chua N-H: Structure and developmental regulation of a wheat gene encoding the major chlorophyll a/b-binding polypeptide. Mol Cell Biol 5:1370-1378 (1985). 23. Leech RM: Chloroplast development in angiosperms: current knowledge and future prospects. In: Baker NR, Barber J (eds) Chloroplast Biogenesis, pp. 1-21. Elsevier Science Publishers, Amsterdam (1984). 24. Maniatis T, Fritsch EF, Sambrook J: Molecular Cloning A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor; NY (1982). 25. Matsuoka M, Kano-Murakami Y, Yamamoto N" Nucleotide sequence of eDNA encoding the light-harvesting chlorophyll a/b-binding protein from maize. Nucleic Acids Res 15: 6302 (1987). 26. Mullet JE: The amino acid sequence of the polypeptide segment which regulates membrane adhesion (grana stacking) in chloroplasts. J Biol Chem 258:9941-9948 (1983). 27. Mullet JE, Chua N-H: In vitro reconstitution of synthesis, uptake, and assembly of cytoplasmically synthesized chloroplast proteins. Methods Enzymol 97- 502-509 (1983). 28. Roberts BE, Paterson BM: Efficient translation of tobacco

107 mosaic virus RNA and rabbit globin 9S RNA in a cell-free system from commercial wheat germ. Proc Natl Acad Sci USA 70:2330-2334 (1973). 29. Robinson C, Ellis RJ: Transport of proteins into chloroplasts. The precursor of small subunit of ribulose bisphosphate carboxylase is processed to the mature size in two steps. Eur J Biochem 142:343-346 (1984). 30. Schmidt GW, Bartlett SG, Grossman AR, Cashmore AR, Chua N-H: Biosynthetic pathways of two polypeptide subunits of the light-harvesting chlorophyll a/b protein complex. J Cell Biol 91:468-478 (1981). 31. Schmidt GW, Mishkind ML: The transport of proteins into chloroplasts. Ann Rev Biochem 55:879-912 (1986).

32. Stayton MM, Black M, Bedbrook J, Dunsmuir P: A novel chlorophyll a/b binding (Cab) protein gene from petunia which encodes the lower molecular weight Cab precursor protein. Nucl Acids Res 14:9781-9797 (1986). 33. Steinback KE, Burke J J, Arntzen C J: Evidence for the role of surface-exposed segments of the light-harvesting complex in cation-mediated control of chloroplast structure and function. Arch Biochem Biophys 195:546-557 (1979). 35. Thornber JP, Peter GF, Nechushtai R, Chitnis PR, Hunter FA, Tobin EM: Electrophoretic separation of chlorophyllprotein complexes and their apoproteins. In: Akoyunoglou G, Senger H (eds.) Regulation of Chloroplast Differentiation, pp. 249-258. Alan R. Liss, New York (1986).