Eur. J. Biochem. 208, 195 - 202 (1 992)
0FEBS 1992
Integration of early light-inducible proteins into isolated thylakoid membranes Elisabeth KRUSE and Klaus KLOPPSTECH
Institut fur Botanik, Universitat Hannover, FRG
(Received January 29/May 13, 1992) - EJB 920122
An in-vitro system has been established to study the integration of early light-inducible proteins (ELIP) into isolated thylakoid membranes. The in-vitro-expressed ELIP precursor proteins exist in two forms, a high-molecular-mass aggregate which is accessible to trypsin but no longer to the stromal processing protease and a soluble form which is readily cleaved to the mature form by the stromal protease. The mature form of ELIP is integrated into thylakoid membranes; its correct integration can be deduced from the observation that the posttranslationally transported products and the invitro integrated ELIP species are cleaved by trypsin to products of the same apparent molecular mass. Trypsin-resistant fragments of high-molecular-mass and low-molecular-mass ELIP appear to have the same size. The processed ELIP species, as well as an engineered mature form of ELIP, are integrated into isolated thylakoid membranes. Integration of the mature protein occurs in the absence of stroma, into sodium-chloride-washed, and trypsin-treated thylakoid membranes. The process of integration is almost temperature independent over 0 - 30 "C. Analysis of the time course ofintegration leads to the conclusion that, under in-vitro conditions, processing but not integration into membranes is the rate-limiting step. In the absence of stroma, the ELIP precursor is bound to the thylakoid membranes, however, it is no longer accessible to the stromal maturating protease when added after binding has occurred. In conclusion, integration of ELIP differs in many essential details from that of its relatives, the light-harvesting chlorophyll u/b protein family.
Early light-inducible proteins (ELIP) are short lived, thylakoid-membrane-located proteins whose genes are transiently transcribed after illumination of etiolated pea [l] or barley plantlets [ 2 , 31. Whilc, to date, only one protein has been found in pea [l, 41, two small multigene families of different size have been characterized in barley [2]. The expression of ELIP genes is not restricted to etiolated plantlets as their genes are transcribed in green plants early in the morning and under the control of the circadian clock [5]. Both in vivo [6] and after in-vitro import into intact chloroplasts [I, 71, ELIP species are integrated into the thylakoid membranes. By means of detergent fractionation of cross-linked thylakoid membranes, we found that ELIP species are associated with a preparation of photosystem-I1 (PS 11) particles [7, 81 and, in particular, with the D1 protein. D1, as a constituent of PS 11, has been found prevalently in the grana thylakoid membrane fraction [9], while a considerable portion of ELIP is obtained from the stroma thylakoid membranes [6, 71 and the intermediate membrane preparations [7].These findings indicate that the association of E I J P with PS I1 should be confined to the site of insertion of preDl into the PS I1 particles [lo]. The turnover and, consequently, the integration of D1 is accelerated after damage of PS I1 reaction centers due to highlight stress, a process known as photoinhibition [ l l , 121. Currespondence to K. Kloppstech, Institut fur Botanik, Universitat Hannover, Herrenhauser Str. 2, W-3000 Hannover 21, Federal Republic of Germany Fax: 0049 51 1 7623456. Abbreviations. ELIP, Early light-inducible protein; LHCP, lightharvesting chlorophyll a/b protein; PS I, Photosystem I, PS 11, Photosystem 11; LHC, light-harvesting complex.
Newly synthesized D1 has been found incorporated in its precursor form at the transition region between grana and stroma membranes [lo]. The interrelation between ELIP species and D1, as outlined above, prompted the question as to whether ELIP might play a role during the assembly of PS 11 and, therefore, also during photoinhibition. Such an assumption would not necessarily be in conflict with the expression of ELIP early during light-induced gene expression, as PS I1 reaction centers are assembled rather early during development [13]. Furthermore, in a system that lacks protective pigments like chlorophylls and carotenoids, much lower light intensities might be more deleterious to developing chloroplasts than to chloroplasts of fully developed green plants. In an attempt to further explore the possible role of ELIP during photoinhibition, the effect of high-light intensities was studied on ELIP expression. It was found that the levels of both the mRNA and the protein are raised in pea leaves exposed to photoinhibitory light at intensities sufficient to lower the stationary concentrations of D1 within the membrane [14].The decay of D1 was accompanied by a correspondent rise in the expression of ELIP mRNA and protein under all light intensities that were tested [15]. A further indication for a function of ELIP as a light-stress protein stems from studies with the carotenoid-overproducing Dunaliellu bardawil. In these mutants, a protein was cloned which possesses an extensive similarity to ELIP; the level of expression of this protein showed a positive correlation with the level of carotenoids under several experimental conditions [I61 A complementary approach towards the understanding of the function(s) of ELIP species and their topology within 9
196 the thylakoid membranes might be by the investigation of integration of ELIP into the thylakoid membranes using the system originally developed for the study of integration and assembly of the apoprotein of light-harvesting-complex I1 (LHC [I) [17, IS]. This system has been used to determine whether the precursor or the mature protein is integrated into the thylakoid membrane and to study the assembly with chlorophylls and carotenoids during the integration process [19]. Similarly, assembly of the photosystem-I (PS I) [20] or the PS I1 [21] reaction center complexes in the thylakoid membrane have been studied. Since ELIP species are distant relatives of light-harvesting complex (LHC) proteins, this system will allow us to analyze ELIP topology within natural and artificial membranes, its possible association with pigments, the interaction with neighbouring proteins and the effect of in-vitro mutagenesis on these various aspects. In the present study, we analyze the integration of ELIP species into the thylakoid membrane in some detail and compare it with data obtained by others for the LHC protein Family. MATERIALS AND METHODS Construction of clone HV8F6A Clone HVSF6A derived from clone HV8F6 which corresponds to a high-molecular-mass ELIP [3], was produced by digestion of the plasmid with NcoI and EeoRV, filling-in of the Ncol-derived overhang, religation and transformation into Escherichiu coli XL-1 blue. Details of the procedures are described in [22]. In-vitro transcription and translation of the truncated clone gives rise to a protein which lacks all but the first four amino acids of the transit peptidc and the five Nterminal amino acids of the mature protein. Preparation of radioactively labeled preELTP and ELIP In most of the experiments, the low-molecular-mass ELIP clone HV17B10 was used for in-vim expression of preEL1P. The sequence of this clone is identical to HV60 [3], but it differs from the corresponding transcription clone p6022 [3], first by the vector used which was Bluescript KS instead of Bluescribe, second by the site of insertion (EcoRV site, start codon downstream of the T3 Promoter), and third by the lack of CG tails and 70 nucleotides of the untranslated leader. Invitro expression of HV17B10 is about ten-times-more efficient that that of p6022, asjudged from incorporation of [35S]methionine. In some experiments, the high-molecular-mass ELIP clones HV13F3, HVSF6 and HV8F6d were used. The plasmids were linearized with BamHI (HV17B10, HV8F6 and HVSF6A) or HiijdII1 (HV13F3); in-vitro transcription and translation were carried out as described [3,23] with the modification that the concentration of KCI in the translation assays, following in-vitro transcription, was lowered to 70 mM in order to improve the translation efficiency. Post-ribosomal translation products were obtained by 3 - 5-fold dilution of the translation mixtures with SO mM Hepes/KOH, pH 8.0, 330 mM sorbitol and 8 mM methionine (solution A) and centrifugation for 30 min at 70000 rpm at 4°C in a Beckman TLA 100.2 rotor (200000 x 8).Electrophoresis (SDS/PAGE) of the products was according to Neville [24]; the products were detected by fluorography [25]. Preparation of chloroplasts and posttranslational transport Intact chloroplasts were isolated as described [3] from 6 7-day-old pea seedlings (cv. Rosa Krone) grown on vermicu-
lite at 25 -C in a 12-h lighti12-h dark cycle at a light intensity of 20 W/m2. Plants were routinely harvested 1-2 h after the onset of illumination. In-vitro transport followed the method of Grossman et al. [26] with minor modifications [3]. For the preparation of crude lysate, chloroplasts were ruptured by osmotic shock in 10 mM Hepes/KOH, pH 8.0 (solution B) at a chlorophyll concentration of 1 mg/ml and kept on ice for at least 5 min. Membranes were recovered by centrifugation at 10000 x g for 10 min at 4"C, washed twice with buffer A and resuspended in the same buffer at a chlorophyll concentration of 1 mg/ml. The soluble stromal fraction was recentrifuged for 30 min at 200000 x g; the resultant supernatant is referred to as 'stroma'. Assay for membrane integration Integration assays were performed as described [27] for insertion of LHCP with some modifications: 200 pl chloroplast lysate containing 10 mM unlabeled methionine or 100 pl membrane suspension and 100 pl stromal fraction were combined with 30 p1 0.1 M ATP (pH 8.0) in solution A and 70 p1 postribosomal translation products and incubated at 25 "C in dim light for 1 h, unless otherwise stated. The reaction was stopped by addition of 700 p1 ice-cold 1:3-diluted solution A and membranes were recovered by centrifugation at 10000 x g for 10 min at 4°C. Aliquots of the supernatants were saved, recentrifuged at 40000 x g to remove residual membranes and concentrated by trichloroacetic acid precipitation. Membranes were incubated at a chlorophyll concentration of 1 mg/ml with or without 80 pg/ml trypsin in solution A for 30 inin on ice, washed with H 2 0 containing 1 mM phenylmethyl sulponylfuoride, S mM amino caproic acid and 1 mM benzamidine. Proteins were precipitated by 80% acetone at -2O'C, the acetone supernatant used for chlorophyll determination and the proteins solubilized to produce equal chlorophyll concentrations, and analyzed by SDSi PAGE and fluorography. Processing assays Processing reactions were carried out under the same conditions as the integration assays, however, the thylakoid-membrane fraction was replaced by solution B. Trypsin-treatment of thylakoid membranes The membranes were resuspended at a concentration of 1 mg/ml chlorophyll in solution A and incubated in the presence of 2-SO pg/ml trypsin for 30 min on ice. Thereafter, ovomucoid was added to a concentration of 0.5mg/ml and the assay sedimented over a 2.5-ml 10% Percoll cushion at 10000 x g for 10 min. The membrane pellet was washed twice in buffer A in the presence of ovomucoid (0.1 mgiml), resuspended in solution B and used for the integration assay. RESULTS
Posttranslational transport and membrane integration of ELIP constructs, as analyzed by trypsinization As shown in Fig. 1, after transport clone HV17B10 gives rise to a 13.5-kDa product and clone HV13F3 to a product of 18.5 kDa; both products are integrated into the thylakoid membrane. Deletion of the transit sequence of clone HV8F6 leads to a modified clone, HV8F6A, which is deleted for almost
197 gated form of the precursor in the membrane fraction might, therefore, not be an indication for its binding to the membranes, but more likely due to the fact that the aggregate form of the precursor protein is pelletable and cosediments together with the membrane vesicles. This aggregate, however, remains completely accessible to the action of trypsin as indicated by its disappearance after trypsin treatment (Fig. 2,lane: membranes P trypsin). About two thirds of the soluble form of the ELIP precursor which remains in the supernatant is processed to the mature protein and most of this mature ELIP is bound to the thylakoid-membrane fraction. In contrast, the greater part of the unprocessed protein stays in the supernatant after removal of the thylakoid membranes by centrifugation. A considerable part of the membrane-bound mature form of ELIP exists in a partially protease-resistant form; after incubation of tho thylakoids together with the ELIP precursor and preparation of stromal proteins, the integration of a stable product of the size of the mature protein is obtained. The latter can be cleaved by trypsin treatment to a polypeptide of about 10 kDa. This 10-kDa polypeptide will be sensitive to proteases only after treatment of the membranes with detergent concentrations which are above the critical micellar point (data not shown). Depending on the site of action and its true molecular mass, the residual fragments contain 2-4 of the original 10 methionines of the precursor and 6 methionines of the mature protein [3]. The relative weak appearance of the 10-kDa band after trypsinization is therefore partially caused by a proteolytic loss of [35S] label, which was contained in the transit peptide and at the Nterminus of the mature form. Which of the potential cleavage points in the ELIP sequence are cut by trypsin is not known. Consequently, it cannot be decided what percentage of the integrated ELIP molecule remains stable during trypsinization and how many methionines are retained within the trypsinstable fragment.
+
Fig. 1. Trypsin digcstion of ELIP species after integration into thylakoid membranes. In-vitro expressed proteins (TLS) of a low-molecularmass and a high-molecular-mass ELIP were transported into intact chloroplasts. After protcase treatment of chloroplasts, membranes were recovered and kept as controls (-) or treated with trypsin (+) (80 pg/ml trypsin at a chlorophyll concentration of 1 mgiml in solulion A, 30 min on ice). Proteins were pelleted by 80% acetone at -20"C, solubilized to give equal chlorophyll concentrations and analy7ed on a 15% SDSjPAGE gel. Clone HV17B10 is equivalent to clone HV60 [l].
the entire transit peptide; its construction has been described in the Materials and Methods section. In accordance with its missing transit sequence, the translation product of the correspondent transcript can no longer be transported into isolated, intact chloroplasts (data not shown). This clone, therefore, is well suited to test the capability of thylakoids to integrate the mature protein in the absence of a processing event. Trypsin digestion of the thylakoid membranes after transport was examined in order to prove the integration of ELIP into membranes. As can be seen from Fig. 1, both low-molecular-mass and high-molecular-mass ELIP species give rise to trypsin-resistant fragments of about 10 kDa. This observation agrees well with the derived amino acid sequence (not shown) and, in addition, also with the model proposed for integration of the LHC I1 proteins [28]. In accordance with this model, the N-terminal sequences of the small and large ELIP species are supposed to be located at the stromal side of the thylakoid membrane; these would be removed by trypsinization at a lysine residue, close to the N-terminus in both ELIP subfamilies.
Establishing a system for integration of ELIP species into the thylakoid membrane The basic properties of integration into the thylakoid membrane using ELIP precursor and mature proteins are described in Fig. 2. A considerable portion of the in-vitrotranslated ELIP precursor exists in a pelletable form and has to be removed by centrifugation prior to the integration assay. The precursors contained within this protein agglomerate cannot be processed into the mature protein by the processing protease(s) present in the stromal fraction, in contrast to the soluble form of the in-vitro-translated precursor which is readily cleaved by this enzyme. The appearance of the aggre-
Comparison between the different clones and evidence that the processed form of the precursor is integrated into the thylakoid membranes In Fig. 3, the integration of the ELIP transcription-translation product obtained from the different clones, and of a truncated form of clone HV8F6 into thylakoid membranes is compared. The precursor of a small ELIP (clone HV17B10) and of two large ELIP species (clones HV13F3 and HV8F6) are processed by stroma and integrated into the membrane in the correct orientation, as can be concluded from the fact that trypsinization leads to membrane-bound products of the same apparent molecular mass. The truncated 'mature' form of ELIP (clone HV8F6d) is readily integrated into the membrane. It yields a trypsin-resistant form of the same apparent molecular mass, obtained with the precursor proteins of highmolecular-mass and low-molecular-mass clones expressed by transcription-translation. This evidence supports the conclusion that the mature protein and not the precursor is integrated into the membranes while, as shown above, processing occurs in the stroma.
Can the bound precursor be processed within the membranes? To further elaborate on the mechanism of integration of ELIP into the thylakoid membrane, the precursor was presented to the membranes in the presence and absence of stroma during a first incubation. Thereafter, the free, non-bound form of the precursor was removed by treatment of the thylakoid
198
Fig. 2. Tntegration and proteolytic maturation of soluble and pelleted forms of in-vitvo expressed preELIP. Radiolabeled low-molecular-mass prcELIP (clonc HV17BIO) was prcpared by in-vitro translation and diluted fivefold with solution A. Half of the mixture was fractionated into soluble and pelletable components by ultracentrifugation, and the pellet was resuspended in the original volume of the supernatant of a mocktranslation mixture. Aliquots of total (T), soluble (S), and pelletable (P) precursor were analyzed directly (TLS) after incubation with a membrane-free stromal extract (processing) or were presented to isolated thylakoids in the presence of stroma. Membranes recovered from the integration assay were either analyzed without (-) or with (+) trypsin treatment. The non-integrated ELIP species were also analyzed (supernatants).
Fig. 3. Integration of various preELIP and a presequence-deleted ELIP into the thylakoid membrane in a chloroplast lysate. The in-vitro expressed proteins of the indicated clones, including the trdnsit-peptidedeleted HV8F6d were integrated into the rnembrancs (membr.) of a chloroplast lysate. Aliquots of the translation assay (TLS) and membranes + / - trypsin treatment were separated by SDS/PAGE. The 10-kDa trypsin fragment is indicated by an arrow head.
membranes with either buffer or with 0.1 M NaOH. These membranes, containing either the precursor or the mature product, were washed and incubated for a second time in the presence of stroma. The integration of ELIP species was tested by trypsinization of the membranes after the second incubation. A second exposure of membranes loaded with precursors to stroma does not increase the amount of ELIP incorporated in a trypsin-resistant form. The experiments indicate that membrane binding of the precursor can be achieved ; however, this bound form of the precursor cannot be processed to its mature form by the addition of another portion of stroma. It has to be concluded that the transit peptide of membranebound ELIP is no longer accessible to the processing enzyme. Under these conditions, only a small portion of mature ELIP is present in an orientation resistant to trypsin, as can be concluded from the minute amount of the 10-kDa fragment which is recovered. Treatment of membranes with 0.1 M NaOH instead of buffer (Fig. 4), a procedure which is very often used to discriminate between associated proteins and proteins which are integrated into the membranes [29], gives an unexpected result. The amount of the incorporated mature form of ELIP appears to be increased. However, after trypsinization, only part of thc protein is processed to thc trypsin-resistant fragment of 10 kDa which is protected by the membrane vesicles, while a
Fig. 4. Thylakoid-membrane-bound preELIP cannot be processed. Radiolabeled preELIP was incubated with a thylakoid-membrane fraction in the presence (+) or absence (-) of stroma for 30 min. Membranes were recovered from the integration mixture, treated either with buffer (A) or NaOH (B) and subjected to a second incubation for 30 min in the presence (+) and absence (-) of freshly added stroma. Aliquots of the membranes +/ - trypsin treatment were separated by SDS/PAGE. The 10-kDa trypsin fragment is indicated by an arrow head.
considerable part of the radioactivity remains at the position of the mature protein. The fact that the N-terminus of the mature form of the protein is no longer accessible to trypsin digestion indicates that treatment with NaOH simulates correct integration of the processed mature protein into the membranes, at least as shown here for ELIP species.
Is a stromal factor necessary for the correct integration of the mature form of the ELIP? The experiments shown so far are in favor of an interpretation that the stromal factor is necessary for processing of ELIP prior to successful integration. The experiment described in Fig. 5 was designed to prove whether a stromal factor is needed for incorporation of the mature protein into the membrane. This experiment includes two controls. On the lefthand side of Fig. 5, a precursor to a large ELIP and on the right-hand side a precursor to a small ELIP were exposed to membranes in the presence and absence of stroma. Treatment with trypsin shows that, in the absence of stroma, only a minute part of the preELIP is integrated into the membrane in a trypsin-resistant form. This is not the case when the
199
Fig. 5. Integration of preELLP and a presequence-deleted ELIP in the presence or absence of stroma. Radiolabeled precursors of low-molecular-mass (HV17B10) of high-molecular-mass (HV8F6) ELIP or a presequence-deleted protein (HV8F6A) were synthesi~edby transcription translation and assayed for integration into isolated thylakoids in the presence or absence of stroma Ahquots of the translation supernatant (TLS) and the membranes recovered from the integration assay and treated with (+) or without (-) trypsin were analyzed The 10-kDa trypsin fragment is indicated by arrow heads.
mature form of ELIP (clone 8F6d) is offered to the membranes. The mature form of the large ELIP is integrated into the membranes and yields the comparable amount of trypsinresistant form which is produced in the presence of the stromal factor containing the processing activity for cleavage of the transit peptide. This experiment is hghly indicative of the fact that a soluble factor is needed for maturation of the protein, but not for the correct integration of ELIP into the thylakoid membranes. The trypsin resistant fragment derived from clone HV8F6d is indistinguishable from that obtained after trypsinization of membranes following in-vitro transport or after cleavage and integration of the precusor protein in the presence of stroma.
Fig. 6. Effect of NaCl on integration and maturation of ELIP. In-vitro expressed HV17B10 was incubated in a chloroplast lysate in the presence of the indicated concentrations of NaCI. The recovered membranes were treated in the presence (+) or absence (-) of trypsin as shown in (A). The supernatants of the integration assay are presented in the left half of (B). Processing (B, right) was performed in a parallel lysate from which membranes had been removed prior to the addition of translation supcmatant. Untreated translation supernatant (TLS) was included for comparison in both A and B. The 10-kDa trypsin fragment is indicated by an arrow head.
Is a membrane-bound proteinaceous component necessary for ELIP integration? The observation that no stromal factor is necessary for integration of ELIP might result from the fact that thylakoid membranes contain such a factor either as a specific component of an integration machinery, or as an essential stromal factor which might be unspecifically bound in an amount sufficient to mimick the independence of a protein factor. Treatment of membranes with salts might help to unravel unspecific binding. In the presence of NaCl, integration of ELIP into the membranes is abolished at a concentration of about 0.2 M NaCl (Fig. 6A). This result cannot be explained by the effect of salt on the processing step as even in the presence of 1 M NaCl a considerable part of the precursor is processed (Fig. 6B, right). Therefore, this mature form of ELIP is available for integration at the high-salt concentrations. Furthermore, the finding that the processed form of ELIP does not accumulate in the supernatant is a good indication that, in the presence of high-salt concentrations, the precursor protein is bound to the thylakoid membranes more rapidly than processingcan take place. As shown, processing of the membranebound precursor does not occur (Fig.4A). Control experiments were performed which showed that washing of membranes with 1 M NaCl does not interfere with the incorporation process (data not shown).
Fig. 7. Jntegration of ELIP into NaOH-washed thylakoids. Th ylakoid membranes were treated in the presence of 0.1 M NaOH for 30 min on ice or kept in buffer (BUFFER), washed twice with 10 mM Hepes, pH 8.0 and resuspended in the same buffer. Integration of preELIP (A) in the presence of stroma was as in Fig. 7. In B, a Coomassie-bluestained gel of treated membranes (P), trichloroacetic-acid-precipitated NaOH supernatant (S) and control membranes (C) are included for evaluation of the effect of NaOH.
Washing the membranes with 0.1 M NaOH might reduce, but does not completely abolish, the capability of trypsinresistant integration of ELIP into thylakoid membranes (Fig.7). However, in accordance with the above findings (Fig. 4 B), treatment with NaOH leads to an increase in the binding of the mature but entirely trypsin-resistant form of ELIP. From the fact that the amount of the trypsin-resistant 10-kDa fragment is not increased, we conclude that this binding is not followed by correct insertion into the membranes. Electrophoretic separation of thylakoid-membrane proteins after NaOH treatment shows that a considerable fraction of the protein constituents of the thylakoids is removed by the alkaline treatment (Fig. 7B).
200
Fig. 8. Integration of ELIP into trypsin-treated thylakoids. Thylakoid membranes were treated at the indicated trypsin concentrations, included into the integration assay in the presence of stroma and analyzed (A) as outlined in the legend to Fig. 8. The 10-kDa trypsin fragmcnt is indicated by an arrow head. A Coomassie-blue-stained gel of the trypsinized membranes (B) is included.
Fig. 9. Temperature dependency of ELIP maturation and integration. In-vitro-expressed preELIP and thc prcsequencc-deleted ELIP (clones HV8F6 and HV8F6d) were assayed for integration in a chloroplast lysate at the indicated temperatures. Membranes recovered from thc assay mixture were analyzed without (A) or with (B) trypsin treatment. Supernatants are shown in (C). Processing reactions were performed under the same conditions in a membrane-free lysatc (D). T, Supernatant of translation. The 10-kDa trypsin fragment is indicated by an arrow head.
Surprisingly, also the treatment of membranes with up to 50 Fg/ml trypsin does not interfere with the capacity of the thylakoid membranes for correct integration of ELIP (Fig. 8A). It appears that membranes treated with high concentrations of trypsin have even a somewhat increased capability for ELIP integration. In accordance with previously published data, control experiments have shown that treatment of membranes with trypsin at a concentration of 2 pg/ ml is already sufficient to cleave 50% of the stroma-exposed N-terminal domain of LHCP and of other thylakoid membrane proteins (Fig. 8B).
of membranes occurs at temperatures above 35"C, even in vitro [30], this finding indicates that unstacking of membranes might not increase the number of integration sites. For the correct integration of LHCP, enhanced levels of MgATP have been found to be essential [27]. In the presence of the minute amounts of ATP (50 pM) which stem from the translation process, integration of ELIP takes place. The addition of ATP enhances protease-resistant incorporation of ELIP considerably over a concentration of 0.05 - 2 mM (data not shown).
Conditions for ELIP integration
DISCUSSION
Integration of ELIP is dependent on two steps; the processing of the precursor and integration of the mature protein. The precursor-deleted clone (HV8F6A) enabled us to discriminate experimentally the cffcct of temperature on both steps. A temperature optimum for the incorporation of ELIP was obtained at 25 'C when the integration procedure was started from the precursor form of ELIP (Fig. 9A, left half). This finding can be explained by a ternpcrature effect on the processing of the precursor. When the mature form (clone HV8F6d ; Fig. 9A, right) was used, almost no temperature dependency was observed over 0- 30°C. Incorporation of the engineered homologue to the mature protein already occurred at 0°C and yielded the trypsin-resistant 10-kDa fragment. The incorporation of both the precursor and the mature form was impaired at temperatures above 30°C, as judged by the yield of the 10-kDa fragment. These findings indicate that, primarily, the processing step but not the correct integration of the mature protein into the membranes, is temperature dcpendent, and at least at temperatures below 20"C, processing is the rate-limiting step (Fig. 9D). Since it is known that unstacking
ELIP species share sequence similarities with LHC proteins of PS I and PS I1 [3, 81. It is the major aim of the discussion to compare the properties of integration into the thylakoid membranes of I,HCP and ELIP with distantly related members of the light-harvesting protein family. The incorporation studies have been performed with chloroplast membranes of 5 - 7-day-old leaves. Considering aspects of development, the main difference between the two proteins seems to be that LHCP is incorporated into green membranes while ELIP species should be integrated into the membranes of developing chloroplasts. However, ELIP species are also expressed in light/dark-grown plants during the morning [5]. Therefore, it appears that the difference in the development of membranes that are capable of incorporating either ELIP or LHCP species is not as extreme as might be expected; it might not even exist at all. In the barley leaf, the maximum expression of LHCP mRNA and therefore most likely also the maximum rate of translation [31], is in the lower segments 2 and 3 [32]if the leaf is divided into six segments and counting is started from the basis. Maximal expression of ELIP occurs,
20 1 in basal segment 1 [3], that is approximately one-day earlier during development than LHC 11. It follows that under invivo conditions, both proteins are inserted into membranes which are still in the process of being assembled. The question is whether in-vitro integration of ELIP into membranes of green chloroplasts occurs irrespective of the stage of development or, alternatively, only into that particular fraction of the membrancs which had been synthesized de novo immediately prior to isolation of the organelles. This question cannot be answered at present. Even though the folding of ELIP and LHCP within the thylakoid membrane seems rather similar, at least three important differences in the conditions of integration might be either indicative of principal differences due to the protein structure or the result of differences in the experimentation between the different laboratories. As shown in this publication, integration of ELIP species occurs at the level of the mature protein. In previous reports it was mentioned that the incorporation of LHCP occurred at the level of the precursor peptide, followed by the maturation event within the membranes [33]. In more recent work, it was found that maturation preceeds the integration process during the posttranslational transport 134, 351, while in the thylakoid system, integration of the precursor polypeptide has also been observed. A possible explanation for these different findings might be that, in the developing system, the steady-state level of the processing protease is still low and, as shown in this paper also for ELIP species, binding of the precursor protein to membranes occurs rapidly in comparison to the processing step [18]. This should hold true, at least under in-vitro conditions where the envelope is missing as a potential ratelimiting barrier and the so-called stroma fraction is very dilute in comparison to its consistency and structure within the plastid. The major differences observed between ELIP and LHCP integration are as follows. The mature form of ELIP can be incorporated into the thylakoid membranes in the absence of a stromal factor, while a protein factor has been found essential for the incorporation of mature LHCP [36]. Another important distinction between the two systems is that proteaseresistant incorporation of LHCP is possible without cleaving the precursor [27]>while preELIP is bound to thylakoid membranes but the processing can no longer take place. As a consequence, preELIP cannot be integrated into the membranes in the correct form, as can also be concluded from the observation that the bound precursor remains accessible to trypsin. The third difference lies in the fact that LHCP cannot be incorporated into proteinase-K-treated membranes [27], while ELIP incorporation occurs in membranes washed with NaCl or treated with trypsin. This makes it unlikely that a proteinaceous counterpart or anchor in the thylakoid membrane which is exposed to the surface is needed for the integration of ELIP, while such interaction has been proposed for the LHCP [27]. This might be a general difference between the two systems, however, it should be pointed out that different proteases have been used in both experimental systems. The stromal factor required for integration of LHCP might be necessary to bind and incorporate the pigments which are associated with the LHCP and seem to be absent in ELIP. We do not know for sure, however, that ELIP species bind pigments at all. There is no indication for the binding of chlorophylls. Our more recent findings on the inducibility of ELIP by high light [15] and the previous work of others [ 161 indicate the alternate possibility that ELIP species might associate with carotenoids.
Another difference might be that preELIP is processed more efficiently under in-vitro conditions than preLHCP. In experiments of others [27], there is almost no mature form of LHCP found in the thylakoid membranes, but a high amount of the protease-resistant fragment. This observation indicates that integration of the LHCP precursor occurs in the correct orientation. A correspondent finding has not been obtained with ELIP, as shown in this paper. Integration of ELIP species, in contrast to the processing step, is almost temperature independent over 0-30°C. From the fact that a temperature above 35°C does not lead to an increase in the integration of ELIP, we tend to conclude that integration into the thylakoid membrane occurs in the stromal region, as was found in vivo [6] and after posttranslational transport [7]. The observed differences in the integration of ELIP and LHCP might be indicative for their different functions. ELIP species are incorporated at an earlier developmental stage than LHCP and ELIP might function either during assembly or as pigment-(carotenoid) providing proteins. They might be characterized by a simpler and eventually more ancient mode of integration than LHCP. This assumption, which is supported by the observed differences between the two integrations systems, shall be studied in the future. This work was supported by a grant (KL 401/9-1) from the Deutsche Fov.tchungs~emeinschujt,Bonn, FRG.
REFERENCES 1. Meyer, G. & Kloppstcch, K. (1984) Eur. J. Biochem. 138, 201 207. 2. Grimm, B. & Kloppstech, K. (1987) Eur. J . Biochem. 167,493499. 3. Grimm, B., Kruse, E. & Kloppstech, K. (1989) Plant Mol. Bid. 13, 583 - 593. 4. Kolanus, W., Scharnhorst, C., Kiihne, U. & Herzfeld, F. (1987) Mol. Gun. Genet. 209, 234-239. 5. Kloppstcch, K. (1985) Planta (Berlin) 165, 502-506. 6. Cronshagen, U. & Herzfeld, F. (1990) Eur. J . Biochem. IY3,361366. 7. Adarnska, I. & Kloppstech, K. (1991) Plant Mol. Biol. 16, 209223. 8. Green, B., Pichersky, E. & Kloppstech, K. (1991) Trends Biochem. Sci. 16, 181 -186. 9. Mattoo, A. K. & Edelman, M. (1987) Proc. Nut1 Acad. Sci. USA 84, 1497-1501. 10. Adir, N., Shochat, S. & Ohad, I. (1990) J . Biol. Chem. 265, 12563 - 12 568. 11. Mattoo, A. K., Hoffman-Falk, H., Marder, J . & Edelman, M. (1984) Pror. Nut1 Acad. Sci. USA 81, 1380- 1384. 12. Ohad, I., Adir, N., Koike, H., Kyle, D. J. & Inoue, Y. (1990) J . Biol. Chem. 265. 1972- 1979. 13. Baker, N. R. & Leech, R. M. (1977) Plant Physiol. 60, 640-644. 14. Adamska, I., Ohad, I. & Kloppstech, K. (1991) in Regulation of chloroplast hiogenesis (Argyroudi-Akoyunoglou, J., eds) NATO ASI, in the press. 15. Adamska, I., Ohad, I. & Kloppstech, K. (1992) Proc. Nut1 Acad. Sci. USA 89,2610-2613. 16. Lers, A., Levy, H.&Zamir, A. (1991)J. Biol. Chem.266, 1369813705. 17. Cline, K. (1986) J . Biol. Chern. 261, 14804-14810. 18. Chitnis, P. R., Nechushtai, R. & Thornber, P. (1987) Plant ,4401. Bid. 10, 3 - 11. 19. Paulsen, H., Rumler, U. & Riidiger, W. (1990) Planta (Berlin) 181,204-211, 20. Cohen, Y . , Steppuhn, J., Herrmann, R. G., Yalovsky, S. & Nechushtai, R. (1992) EMBO J . 11, 79-85. 21. Friemann, A., Schwarz, H. J. & Hachtel, W. (1991) in Regulation of chloroplast biogenesis (Argyroudi-Akoyunoglou, J., ed.) NATO ASI, in the press.
202 22. Maniatis, T., Frisch, E. F. & Sambrook, J. (1982) Molecular cloning, Cold Spring Harbour Laboratory, New York. 23. Roberts, B. E. & Paterson, B. M. (1973) Proc. Natl Acad. Sci. USA 70,2330-2334. 24. Neville, D. M. (1971) J . Biol Chem. 246,6328-6334. 25. Bonner, W. M. & Laskey, R. A. (1974) Bur. J. Biochem. 46,83 88. 26. Grossman, A. R., Bartlett, S. G., Schmidt, G. W., Mullet, J. E. & Chua, N. H. (1982) J. Biol. Chem. 257,1558-1563. 27. Fulsom, D. R. &Cline, K. (1988) Plant Physiol. 88,1146-1153. 28. Karlin-Neuman, G. A,, Kohorn, B. D., Thornber, J. P. & Tobin, E. M. (1985) J. Mol. Appl. Genet. 3, 45-61. 29. Schmidt, G. W., Bartlett, S. G., Grossman. A. R., Cashmore, R. & Chua, N. H. (1981) J. Cell. Biol. 91,468-478.
30. Gounaris, K., Brain, A. R. R., Quinn. P. J. & Williams, W. P. (1984) Biochim. Biophys. Acta 766, 198-208. 31. Adamska, I., Scheel, B. & Kloppstech, K. (1991) Plant Mol. Biol. 17,1055-1065. 32. Viro, M. & Kloppstech, K. (1980) Pluntu (Berlin) 150, 41 -45. 33. Chitnis, P. R., Harel. E., Kohorn B. D., Tobin, E. M. & Thornber, J. P. (1986) J . Cell Bid. 102, 982-988. 34. Cline, K., Fulsom, D. R. & Viitanen, P. V. (1989) J . Bid. Chem. 264, I4225 - 14232. 35. Reed, J. E., Cline, K., Stephens, L. C., Bacot, K. 0. & Yiitanen, P . V. (1990)Eur. J. Bioc:hem. 194, 33-42. 36. Viitanen, P. V., Doran, E. R. & Dunsmuir, P. (1988) J . Bid. Chem. 263, 15000-15007.
0FEBS 1992
Integration of early light-inducible proteins into isolated thylakoid membranes Elisabeth KRUSE and Klaus KLOPPSTECH
Institut fur Botanik, Universitat Hannover, FRG
(Received January 29/May 13, 1992) - EJB 920122
An in-vitro system has been established to study the integration of early light-inducible proteins (ELIP) into isolated thylakoid membranes. The in-vitro-expressed ELIP precursor proteins exist in two forms, a high-molecular-mass aggregate which is accessible to trypsin but no longer to the stromal processing protease and a soluble form which is readily cleaved to the mature form by the stromal protease. The mature form of ELIP is integrated into thylakoid membranes; its correct integration can be deduced from the observation that the posttranslationally transported products and the invitro integrated ELIP species are cleaved by trypsin to products of the same apparent molecular mass. Trypsin-resistant fragments of high-molecular-mass and low-molecular-mass ELIP appear to have the same size. The processed ELIP species, as well as an engineered mature form of ELIP, are integrated into isolated thylakoid membranes. Integration of the mature protein occurs in the absence of stroma, into sodium-chloride-washed, and trypsin-treated thylakoid membranes. The process of integration is almost temperature independent over 0 - 30 "C. Analysis of the time course ofintegration leads to the conclusion that, under in-vitro conditions, processing but not integration into membranes is the rate-limiting step. In the absence of stroma, the ELIP precursor is bound to the thylakoid membranes, however, it is no longer accessible to the stromal maturating protease when added after binding has occurred. In conclusion, integration of ELIP differs in many essential details from that of its relatives, the light-harvesting chlorophyll u/b protein family.
Early light-inducible proteins (ELIP) are short lived, thylakoid-membrane-located proteins whose genes are transiently transcribed after illumination of etiolated pea [l] or barley plantlets [ 2 , 31. Whilc, to date, only one protein has been found in pea [l, 41, two small multigene families of different size have been characterized in barley [2]. The expression of ELIP genes is not restricted to etiolated plantlets as their genes are transcribed in green plants early in the morning and under the control of the circadian clock [5]. Both in vivo [6] and after in-vitro import into intact chloroplasts [I, 71, ELIP species are integrated into the thylakoid membranes. By means of detergent fractionation of cross-linked thylakoid membranes, we found that ELIP species are associated with a preparation of photosystem-I1 (PS 11) particles [7, 81 and, in particular, with the D1 protein. D1, as a constituent of PS 11, has been found prevalently in the grana thylakoid membrane fraction [9], while a considerable portion of ELIP is obtained from the stroma thylakoid membranes [6, 71 and the intermediate membrane preparations [7].These findings indicate that the association of E I J P with PS I1 should be confined to the site of insertion of preDl into the PS I1 particles [lo]. The turnover and, consequently, the integration of D1 is accelerated after damage of PS I1 reaction centers due to highlight stress, a process known as photoinhibition [ l l , 121. Currespondence to K. Kloppstech, Institut fur Botanik, Universitat Hannover, Herrenhauser Str. 2, W-3000 Hannover 21, Federal Republic of Germany Fax: 0049 51 1 7623456. Abbreviations. ELIP, Early light-inducible protein; LHCP, lightharvesting chlorophyll a/b protein; PS I, Photosystem I, PS 11, Photosystem 11; LHC, light-harvesting complex.
Newly synthesized D1 has been found incorporated in its precursor form at the transition region between grana and stroma membranes [lo]. The interrelation between ELIP species and D1, as outlined above, prompted the question as to whether ELIP might play a role during the assembly of PS 11 and, therefore, also during photoinhibition. Such an assumption would not necessarily be in conflict with the expression of ELIP early during light-induced gene expression, as PS I1 reaction centers are assembled rather early during development [13]. Furthermore, in a system that lacks protective pigments like chlorophylls and carotenoids, much lower light intensities might be more deleterious to developing chloroplasts than to chloroplasts of fully developed green plants. In an attempt to further explore the possible role of ELIP during photoinhibition, the effect of high-light intensities was studied on ELIP expression. It was found that the levels of both the mRNA and the protein are raised in pea leaves exposed to photoinhibitory light at intensities sufficient to lower the stationary concentrations of D1 within the membrane [14].The decay of D1 was accompanied by a correspondent rise in the expression of ELIP mRNA and protein under all light intensities that were tested [15]. A further indication for a function of ELIP as a light-stress protein stems from studies with the carotenoid-overproducing Dunaliellu bardawil. In these mutants, a protein was cloned which possesses an extensive similarity to ELIP; the level of expression of this protein showed a positive correlation with the level of carotenoids under several experimental conditions [I61 A complementary approach towards the understanding of the function(s) of ELIP species and their topology within 9
196 the thylakoid membranes might be by the investigation of integration of ELIP into the thylakoid membranes using the system originally developed for the study of integration and assembly of the apoprotein of light-harvesting-complex I1 (LHC [I) [17, IS]. This system has been used to determine whether the precursor or the mature protein is integrated into the thylakoid membrane and to study the assembly with chlorophylls and carotenoids during the integration process [19]. Similarly, assembly of the photosystem-I (PS I) [20] or the PS I1 [21] reaction center complexes in the thylakoid membrane have been studied. Since ELIP species are distant relatives of light-harvesting complex (LHC) proteins, this system will allow us to analyze ELIP topology within natural and artificial membranes, its possible association with pigments, the interaction with neighbouring proteins and the effect of in-vitro mutagenesis on these various aspects. In the present study, we analyze the integration of ELIP species into the thylakoid membrane in some detail and compare it with data obtained by others for the LHC protein Family. MATERIALS AND METHODS Construction of clone HV8F6A Clone HVSF6A derived from clone HV8F6 which corresponds to a high-molecular-mass ELIP [3], was produced by digestion of the plasmid with NcoI and EeoRV, filling-in of the Ncol-derived overhang, religation and transformation into Escherichiu coli XL-1 blue. Details of the procedures are described in [22]. In-vitro transcription and translation of the truncated clone gives rise to a protein which lacks all but the first four amino acids of the transit peptidc and the five Nterminal amino acids of the mature protein. Preparation of radioactively labeled preELTP and ELIP In most of the experiments, the low-molecular-mass ELIP clone HV17B10 was used for in-vim expression of preEL1P. The sequence of this clone is identical to HV60 [3], but it differs from the corresponding transcription clone p6022 [3], first by the vector used which was Bluescript KS instead of Bluescribe, second by the site of insertion (EcoRV site, start codon downstream of the T3 Promoter), and third by the lack of CG tails and 70 nucleotides of the untranslated leader. Invitro expression of HV17B10 is about ten-times-more efficient that that of p6022, asjudged from incorporation of [35S]methionine. In some experiments, the high-molecular-mass ELIP clones HV13F3, HVSF6 and HV8F6d were used. The plasmids were linearized with BamHI (HV17B10, HV8F6 and HVSF6A) or HiijdII1 (HV13F3); in-vitro transcription and translation were carried out as described [3,23] with the modification that the concentration of KCI in the translation assays, following in-vitro transcription, was lowered to 70 mM in order to improve the translation efficiency. Post-ribosomal translation products were obtained by 3 - 5-fold dilution of the translation mixtures with SO mM Hepes/KOH, pH 8.0, 330 mM sorbitol and 8 mM methionine (solution A) and centrifugation for 30 min at 70000 rpm at 4°C in a Beckman TLA 100.2 rotor (200000 x 8).Electrophoresis (SDS/PAGE) of the products was according to Neville [24]; the products were detected by fluorography [25]. Preparation of chloroplasts and posttranslational transport Intact chloroplasts were isolated as described [3] from 6 7-day-old pea seedlings (cv. Rosa Krone) grown on vermicu-
lite at 25 -C in a 12-h lighti12-h dark cycle at a light intensity of 20 W/m2. Plants were routinely harvested 1-2 h after the onset of illumination. In-vitro transport followed the method of Grossman et al. [26] with minor modifications [3]. For the preparation of crude lysate, chloroplasts were ruptured by osmotic shock in 10 mM Hepes/KOH, pH 8.0 (solution B) at a chlorophyll concentration of 1 mg/ml and kept on ice for at least 5 min. Membranes were recovered by centrifugation at 10000 x g for 10 min at 4"C, washed twice with buffer A and resuspended in the same buffer at a chlorophyll concentration of 1 mg/ml. The soluble stromal fraction was recentrifuged for 30 min at 200000 x g; the resultant supernatant is referred to as 'stroma'. Assay for membrane integration Integration assays were performed as described [27] for insertion of LHCP with some modifications: 200 pl chloroplast lysate containing 10 mM unlabeled methionine or 100 pl membrane suspension and 100 pl stromal fraction were combined with 30 p1 0.1 M ATP (pH 8.0) in solution A and 70 p1 postribosomal translation products and incubated at 25 "C in dim light for 1 h, unless otherwise stated. The reaction was stopped by addition of 700 p1 ice-cold 1:3-diluted solution A and membranes were recovered by centrifugation at 10000 x g for 10 min at 4°C. Aliquots of the supernatants were saved, recentrifuged at 40000 x g to remove residual membranes and concentrated by trichloroacetic acid precipitation. Membranes were incubated at a chlorophyll concentration of 1 mg/ml with or without 80 pg/ml trypsin in solution A for 30 inin on ice, washed with H 2 0 containing 1 mM phenylmethyl sulponylfuoride, S mM amino caproic acid and 1 mM benzamidine. Proteins were precipitated by 80% acetone at -2O'C, the acetone supernatant used for chlorophyll determination and the proteins solubilized to produce equal chlorophyll concentrations, and analyzed by SDSi PAGE and fluorography. Processing assays Processing reactions were carried out under the same conditions as the integration assays, however, the thylakoid-membrane fraction was replaced by solution B. Trypsin-treatment of thylakoid membranes The membranes were resuspended at a concentration of 1 mg/ml chlorophyll in solution A and incubated in the presence of 2-SO pg/ml trypsin for 30 min on ice. Thereafter, ovomucoid was added to a concentration of 0.5mg/ml and the assay sedimented over a 2.5-ml 10% Percoll cushion at 10000 x g for 10 min. The membrane pellet was washed twice in buffer A in the presence of ovomucoid (0.1 mgiml), resuspended in solution B and used for the integration assay. RESULTS
Posttranslational transport and membrane integration of ELIP constructs, as analyzed by trypsinization As shown in Fig. 1, after transport clone HV17B10 gives rise to a 13.5-kDa product and clone HV13F3 to a product of 18.5 kDa; both products are integrated into the thylakoid membrane. Deletion of the transit sequence of clone HV8F6 leads to a modified clone, HV8F6A, which is deleted for almost
197 gated form of the precursor in the membrane fraction might, therefore, not be an indication for its binding to the membranes, but more likely due to the fact that the aggregate form of the precursor protein is pelletable and cosediments together with the membrane vesicles. This aggregate, however, remains completely accessible to the action of trypsin as indicated by its disappearance after trypsin treatment (Fig. 2,lane: membranes P trypsin). About two thirds of the soluble form of the ELIP precursor which remains in the supernatant is processed to the mature protein and most of this mature ELIP is bound to the thylakoid-membrane fraction. In contrast, the greater part of the unprocessed protein stays in the supernatant after removal of the thylakoid membranes by centrifugation. A considerable part of the membrane-bound mature form of ELIP exists in a partially protease-resistant form; after incubation of tho thylakoids together with the ELIP precursor and preparation of stromal proteins, the integration of a stable product of the size of the mature protein is obtained. The latter can be cleaved by trypsin treatment to a polypeptide of about 10 kDa. This 10-kDa polypeptide will be sensitive to proteases only after treatment of the membranes with detergent concentrations which are above the critical micellar point (data not shown). Depending on the site of action and its true molecular mass, the residual fragments contain 2-4 of the original 10 methionines of the precursor and 6 methionines of the mature protein [3]. The relative weak appearance of the 10-kDa band after trypsinization is therefore partially caused by a proteolytic loss of [35S] label, which was contained in the transit peptide and at the Nterminus of the mature form. Which of the potential cleavage points in the ELIP sequence are cut by trypsin is not known. Consequently, it cannot be decided what percentage of the integrated ELIP molecule remains stable during trypsinization and how many methionines are retained within the trypsinstable fragment.
+
Fig. 1. Trypsin digcstion of ELIP species after integration into thylakoid membranes. In-vitro expressed proteins (TLS) of a low-molecularmass and a high-molecular-mass ELIP were transported into intact chloroplasts. After protcase treatment of chloroplasts, membranes were recovered and kept as controls (-) or treated with trypsin (+) (80 pg/ml trypsin at a chlorophyll concentration of 1 mgiml in solulion A, 30 min on ice). Proteins were pelleted by 80% acetone at -20"C, solubilized to give equal chlorophyll concentrations and analy7ed on a 15% SDSjPAGE gel. Clone HV17B10 is equivalent to clone HV60 [l].
the entire transit peptide; its construction has been described in the Materials and Methods section. In accordance with its missing transit sequence, the translation product of the correspondent transcript can no longer be transported into isolated, intact chloroplasts (data not shown). This clone, therefore, is well suited to test the capability of thylakoids to integrate the mature protein in the absence of a processing event. Trypsin digestion of the thylakoid membranes after transport was examined in order to prove the integration of ELIP into membranes. As can be seen from Fig. 1, both low-molecular-mass and high-molecular-mass ELIP species give rise to trypsin-resistant fragments of about 10 kDa. This observation agrees well with the derived amino acid sequence (not shown) and, in addition, also with the model proposed for integration of the LHC I1 proteins [28]. In accordance with this model, the N-terminal sequences of the small and large ELIP species are supposed to be located at the stromal side of the thylakoid membrane; these would be removed by trypsinization at a lysine residue, close to the N-terminus in both ELIP subfamilies.
Establishing a system for integration of ELIP species into the thylakoid membrane The basic properties of integration into the thylakoid membrane using ELIP precursor and mature proteins are described in Fig. 2. A considerable portion of the in-vitrotranslated ELIP precursor exists in a pelletable form and has to be removed by centrifugation prior to the integration assay. The precursors contained within this protein agglomerate cannot be processed into the mature protein by the processing protease(s) present in the stromal fraction, in contrast to the soluble form of the in-vitro-translated precursor which is readily cleaved by this enzyme. The appearance of the aggre-
Comparison between the different clones and evidence that the processed form of the precursor is integrated into the thylakoid membranes In Fig. 3, the integration of the ELIP transcription-translation product obtained from the different clones, and of a truncated form of clone HV8F6 into thylakoid membranes is compared. The precursor of a small ELIP (clone HV17B10) and of two large ELIP species (clones HV13F3 and HV8F6) are processed by stroma and integrated into the membrane in the correct orientation, as can be concluded from the fact that trypsinization leads to membrane-bound products of the same apparent molecular mass. The truncated 'mature' form of ELIP (clone HV8F6d) is readily integrated into the membrane. It yields a trypsin-resistant form of the same apparent molecular mass, obtained with the precursor proteins of highmolecular-mass and low-molecular-mass clones expressed by transcription-translation. This evidence supports the conclusion that the mature protein and not the precursor is integrated into the membranes while, as shown above, processing occurs in the stroma.
Can the bound precursor be processed within the membranes? To further elaborate on the mechanism of integration of ELIP into the thylakoid membrane, the precursor was presented to the membranes in the presence and absence of stroma during a first incubation. Thereafter, the free, non-bound form of the precursor was removed by treatment of the thylakoid
198
Fig. 2. Tntegration and proteolytic maturation of soluble and pelleted forms of in-vitvo expressed preELIP. Radiolabeled low-molecular-mass prcELIP (clonc HV17BIO) was prcpared by in-vitro translation and diluted fivefold with solution A. Half of the mixture was fractionated into soluble and pelletable components by ultracentrifugation, and the pellet was resuspended in the original volume of the supernatant of a mocktranslation mixture. Aliquots of total (T), soluble (S), and pelletable (P) precursor were analyzed directly (TLS) after incubation with a membrane-free stromal extract (processing) or were presented to isolated thylakoids in the presence of stroma. Membranes recovered from the integration assay were either analyzed without (-) or with (+) trypsin treatment. The non-integrated ELIP species were also analyzed (supernatants).
Fig. 3. Integration of various preELIP and a presequence-deleted ELIP into the thylakoid membrane in a chloroplast lysate. The in-vitro expressed proteins of the indicated clones, including the trdnsit-peptidedeleted HV8F6d were integrated into the rnembrancs (membr.) of a chloroplast lysate. Aliquots of the translation assay (TLS) and membranes + / - trypsin treatment were separated by SDS/PAGE. The 10-kDa trypsin fragment is indicated by an arrow head.
membranes with either buffer or with 0.1 M NaOH. These membranes, containing either the precursor or the mature product, were washed and incubated for a second time in the presence of stroma. The integration of ELIP species was tested by trypsinization of the membranes after the second incubation. A second exposure of membranes loaded with precursors to stroma does not increase the amount of ELIP incorporated in a trypsin-resistant form. The experiments indicate that membrane binding of the precursor can be achieved ; however, this bound form of the precursor cannot be processed to its mature form by the addition of another portion of stroma. It has to be concluded that the transit peptide of membranebound ELIP is no longer accessible to the processing enzyme. Under these conditions, only a small portion of mature ELIP is present in an orientation resistant to trypsin, as can be concluded from the minute amount of the 10-kDa fragment which is recovered. Treatment of membranes with 0.1 M NaOH instead of buffer (Fig. 4), a procedure which is very often used to discriminate between associated proteins and proteins which are integrated into the membranes [29], gives an unexpected result. The amount of the incorporated mature form of ELIP appears to be increased. However, after trypsinization, only part of thc protein is processed to thc trypsin-resistant fragment of 10 kDa which is protected by the membrane vesicles, while a
Fig. 4. Thylakoid-membrane-bound preELIP cannot be processed. Radiolabeled preELIP was incubated with a thylakoid-membrane fraction in the presence (+) or absence (-) of stroma for 30 min. Membranes were recovered from the integration mixture, treated either with buffer (A) or NaOH (B) and subjected to a second incubation for 30 min in the presence (+) and absence (-) of freshly added stroma. Aliquots of the membranes +/ - trypsin treatment were separated by SDS/PAGE. The 10-kDa trypsin fragment is indicated by an arrow head.
considerable part of the radioactivity remains at the position of the mature protein. The fact that the N-terminus of the mature form of the protein is no longer accessible to trypsin digestion indicates that treatment with NaOH simulates correct integration of the processed mature protein into the membranes, at least as shown here for ELIP species.
Is a stromal factor necessary for the correct integration of the mature form of the ELIP? The experiments shown so far are in favor of an interpretation that the stromal factor is necessary for processing of ELIP prior to successful integration. The experiment described in Fig. 5 was designed to prove whether a stromal factor is needed for incorporation of the mature protein into the membrane. This experiment includes two controls. On the lefthand side of Fig. 5, a precursor to a large ELIP and on the right-hand side a precursor to a small ELIP were exposed to membranes in the presence and absence of stroma. Treatment with trypsin shows that, in the absence of stroma, only a minute part of the preELIP is integrated into the membrane in a trypsin-resistant form. This is not the case when the
199
Fig. 5. Integration of preELLP and a presequence-deleted ELIP in the presence or absence of stroma. Radiolabeled precursors of low-molecular-mass (HV17B10) of high-molecular-mass (HV8F6) ELIP or a presequence-deleted protein (HV8F6A) were synthesi~edby transcription translation and assayed for integration into isolated thylakoids in the presence or absence of stroma Ahquots of the translation supernatant (TLS) and the membranes recovered from the integration assay and treated with (+) or without (-) trypsin were analyzed The 10-kDa trypsin fragment is indicated by arrow heads.
mature form of ELIP (clone 8F6d) is offered to the membranes. The mature form of the large ELIP is integrated into the membranes and yields the comparable amount of trypsinresistant form which is produced in the presence of the stromal factor containing the processing activity for cleavage of the transit peptide. This experiment is hghly indicative of the fact that a soluble factor is needed for maturation of the protein, but not for the correct integration of ELIP into the thylakoid membranes. The trypsin resistant fragment derived from clone HV8F6d is indistinguishable from that obtained after trypsinization of membranes following in-vitro transport or after cleavage and integration of the precusor protein in the presence of stroma.
Fig. 6. Effect of NaCl on integration and maturation of ELIP. In-vitro expressed HV17B10 was incubated in a chloroplast lysate in the presence of the indicated concentrations of NaCI. The recovered membranes were treated in the presence (+) or absence (-) of trypsin as shown in (A). The supernatants of the integration assay are presented in the left half of (B). Processing (B, right) was performed in a parallel lysate from which membranes had been removed prior to the addition of translation supcmatant. Untreated translation supernatant (TLS) was included for comparison in both A and B. The 10-kDa trypsin fragment is indicated by an arrow head.
Is a membrane-bound proteinaceous component necessary for ELIP integration? The observation that no stromal factor is necessary for integration of ELIP might result from the fact that thylakoid membranes contain such a factor either as a specific component of an integration machinery, or as an essential stromal factor which might be unspecifically bound in an amount sufficient to mimick the independence of a protein factor. Treatment of membranes with salts might help to unravel unspecific binding. In the presence of NaCl, integration of ELIP into the membranes is abolished at a concentration of about 0.2 M NaCl (Fig. 6A). This result cannot be explained by the effect of salt on the processing step as even in the presence of 1 M NaCl a considerable part of the precursor is processed (Fig. 6B, right). Therefore, this mature form of ELIP is available for integration at the high-salt concentrations. Furthermore, the finding that the processed form of ELIP does not accumulate in the supernatant is a good indication that, in the presence of high-salt concentrations, the precursor protein is bound to the thylakoid membranes more rapidly than processingcan take place. As shown, processing of the membranebound precursor does not occur (Fig.4A). Control experiments were performed which showed that washing of membranes with 1 M NaCl does not interfere with the incorporation process (data not shown).
Fig. 7. Jntegration of ELIP into NaOH-washed thylakoids. Th ylakoid membranes were treated in the presence of 0.1 M NaOH for 30 min on ice or kept in buffer (BUFFER), washed twice with 10 mM Hepes, pH 8.0 and resuspended in the same buffer. Integration of preELIP (A) in the presence of stroma was as in Fig. 7. In B, a Coomassie-bluestained gel of treated membranes (P), trichloroacetic-acid-precipitated NaOH supernatant (S) and control membranes (C) are included for evaluation of the effect of NaOH.
Washing the membranes with 0.1 M NaOH might reduce, but does not completely abolish, the capability of trypsinresistant integration of ELIP into thylakoid membranes (Fig.7). However, in accordance with the above findings (Fig. 4 B), treatment with NaOH leads to an increase in the binding of the mature but entirely trypsin-resistant form of ELIP. From the fact that the amount of the trypsin-resistant 10-kDa fragment is not increased, we conclude that this binding is not followed by correct insertion into the membranes. Electrophoretic separation of thylakoid-membrane proteins after NaOH treatment shows that a considerable fraction of the protein constituents of the thylakoids is removed by the alkaline treatment (Fig. 7B).
200
Fig. 8. Integration of ELIP into trypsin-treated thylakoids. Thylakoid membranes were treated at the indicated trypsin concentrations, included into the integration assay in the presence of stroma and analyzed (A) as outlined in the legend to Fig. 8. The 10-kDa trypsin fragmcnt is indicated by an arrow head. A Coomassie-blue-stained gel of the trypsinized membranes (B) is included.
Fig. 9. Temperature dependency of ELIP maturation and integration. In-vitro-expressed preELIP and thc prcsequencc-deleted ELIP (clones HV8F6 and HV8F6d) were assayed for integration in a chloroplast lysate at the indicated temperatures. Membranes recovered from thc assay mixture were analyzed without (A) or with (B) trypsin treatment. Supernatants are shown in (C). Processing reactions were performed under the same conditions in a membrane-free lysatc (D). T, Supernatant of translation. The 10-kDa trypsin fragment is indicated by an arrow head.
Surprisingly, also the treatment of membranes with up to 50 Fg/ml trypsin does not interfere with the capacity of the thylakoid membranes for correct integration of ELIP (Fig. 8A). It appears that membranes treated with high concentrations of trypsin have even a somewhat increased capability for ELIP integration. In accordance with previously published data, control experiments have shown that treatment of membranes with trypsin at a concentration of 2 pg/ ml is already sufficient to cleave 50% of the stroma-exposed N-terminal domain of LHCP and of other thylakoid membrane proteins (Fig. 8B).
of membranes occurs at temperatures above 35"C, even in vitro [30], this finding indicates that unstacking of membranes might not increase the number of integration sites. For the correct integration of LHCP, enhanced levels of MgATP have been found to be essential [27]. In the presence of the minute amounts of ATP (50 pM) which stem from the translation process, integration of ELIP takes place. The addition of ATP enhances protease-resistant incorporation of ELIP considerably over a concentration of 0.05 - 2 mM (data not shown).
Conditions for ELIP integration
DISCUSSION
Integration of ELIP is dependent on two steps; the processing of the precursor and integration of the mature protein. The precursor-deleted clone (HV8F6A) enabled us to discriminate experimentally the cffcct of temperature on both steps. A temperature optimum for the incorporation of ELIP was obtained at 25 'C when the integration procedure was started from the precursor form of ELIP (Fig. 9A, left half). This finding can be explained by a ternpcrature effect on the processing of the precursor. When the mature form (clone HV8F6d ; Fig. 9A, right) was used, almost no temperature dependency was observed over 0- 30°C. Incorporation of the engineered homologue to the mature protein already occurred at 0°C and yielded the trypsin-resistant 10-kDa fragment. The incorporation of both the precursor and the mature form was impaired at temperatures above 30°C, as judged by the yield of the 10-kDa fragment. These findings indicate that, primarily, the processing step but not the correct integration of the mature protein into the membranes, is temperature dcpendent, and at least at temperatures below 20"C, processing is the rate-limiting step (Fig. 9D). Since it is known that unstacking
ELIP species share sequence similarities with LHC proteins of PS I and PS I1 [3, 81. It is the major aim of the discussion to compare the properties of integration into the thylakoid membranes of I,HCP and ELIP with distantly related members of the light-harvesting protein family. The incorporation studies have been performed with chloroplast membranes of 5 - 7-day-old leaves. Considering aspects of development, the main difference between the two proteins seems to be that LHCP is incorporated into green membranes while ELIP species should be integrated into the membranes of developing chloroplasts. However, ELIP species are also expressed in light/dark-grown plants during the morning [5]. Therefore, it appears that the difference in the development of membranes that are capable of incorporating either ELIP or LHCP species is not as extreme as might be expected; it might not even exist at all. In the barley leaf, the maximum expression of LHCP mRNA and therefore most likely also the maximum rate of translation [31], is in the lower segments 2 and 3 [32]if the leaf is divided into six segments and counting is started from the basis. Maximal expression of ELIP occurs,
20 1 in basal segment 1 [3], that is approximately one-day earlier during development than LHC 11. It follows that under invivo conditions, both proteins are inserted into membranes which are still in the process of being assembled. The question is whether in-vitro integration of ELIP into membranes of green chloroplasts occurs irrespective of the stage of development or, alternatively, only into that particular fraction of the membrancs which had been synthesized de novo immediately prior to isolation of the organelles. This question cannot be answered at present. Even though the folding of ELIP and LHCP within the thylakoid membrane seems rather similar, at least three important differences in the conditions of integration might be either indicative of principal differences due to the protein structure or the result of differences in the experimentation between the different laboratories. As shown in this publication, integration of ELIP species occurs at the level of the mature protein. In previous reports it was mentioned that the incorporation of LHCP occurred at the level of the precursor peptide, followed by the maturation event within the membranes [33]. In more recent work, it was found that maturation preceeds the integration process during the posttranslational transport 134, 351, while in the thylakoid system, integration of the precursor polypeptide has also been observed. A possible explanation for these different findings might be that, in the developing system, the steady-state level of the processing protease is still low and, as shown in this paper also for ELIP species, binding of the precursor protein to membranes occurs rapidly in comparison to the processing step [18]. This should hold true, at least under in-vitro conditions where the envelope is missing as a potential ratelimiting barrier and the so-called stroma fraction is very dilute in comparison to its consistency and structure within the plastid. The major differences observed between ELIP and LHCP integration are as follows. The mature form of ELIP can be incorporated into the thylakoid membranes in the absence of a stromal factor, while a protein factor has been found essential for the incorporation of mature LHCP [36]. Another important distinction between the two systems is that proteaseresistant incorporation of LHCP is possible without cleaving the precursor [27]>while preELIP is bound to thylakoid membranes but the processing can no longer take place. As a consequence, preELIP cannot be integrated into the membranes in the correct form, as can also be concluded from the observation that the bound precursor remains accessible to trypsin. The third difference lies in the fact that LHCP cannot be incorporated into proteinase-K-treated membranes [27], while ELIP incorporation occurs in membranes washed with NaCl or treated with trypsin. This makes it unlikely that a proteinaceous counterpart or anchor in the thylakoid membrane which is exposed to the surface is needed for the integration of ELIP, while such interaction has been proposed for the LHCP [27]. This might be a general difference between the two systems, however, it should be pointed out that different proteases have been used in both experimental systems. The stromal factor required for integration of LHCP might be necessary to bind and incorporate the pigments which are associated with the LHCP and seem to be absent in ELIP. We do not know for sure, however, that ELIP species bind pigments at all. There is no indication for the binding of chlorophylls. Our more recent findings on the inducibility of ELIP by high light [15] and the previous work of others [ 161 indicate the alternate possibility that ELIP species might associate with carotenoids.
Another difference might be that preELIP is processed more efficiently under in-vitro conditions than preLHCP. In experiments of others [27], there is almost no mature form of LHCP found in the thylakoid membranes, but a high amount of the protease-resistant fragment. This observation indicates that integration of the LHCP precursor occurs in the correct orientation. A correspondent finding has not been obtained with ELIP, as shown in this paper. Integration of ELIP species, in contrast to the processing step, is almost temperature independent over 0-30°C. From the fact that a temperature above 35°C does not lead to an increase in the integration of ELIP, we tend to conclude that integration into the thylakoid membrane occurs in the stromal region, as was found in vivo [6] and after posttranslational transport [7]. The observed differences in the integration of ELIP and LHCP might be indicative for their different functions. ELIP species are incorporated at an earlier developmental stage than LHCP and ELIP might function either during assembly or as pigment-(carotenoid) providing proteins. They might be characterized by a simpler and eventually more ancient mode of integration than LHCP. This assumption, which is supported by the observed differences between the two integrations systems, shall be studied in the future. This work was supported by a grant (KL 401/9-1) from the Deutsche Fov.tchungs~emeinschujt,Bonn, FRG.
REFERENCES 1. Meyer, G. & Kloppstcch, K. (1984) Eur. J. Biochem. 138, 201 207. 2. Grimm, B. & Kloppstech, K. (1987) Eur. J . Biochem. 167,493499. 3. Grimm, B., Kruse, E. & Kloppstech, K. (1989) Plant Mol. Bid. 13, 583 - 593. 4. Kolanus, W., Scharnhorst, C., Kiihne, U. & Herzfeld, F. (1987) Mol. Gun. Genet. 209, 234-239. 5. Kloppstcch, K. (1985) Planta (Berlin) 165, 502-506. 6. Cronshagen, U. & Herzfeld, F. (1990) Eur. J . Biochem. IY3,361366. 7. Adarnska, I. & Kloppstech, K. (1991) Plant Mol. Biol. 16, 209223. 8. Green, B., Pichersky, E. & Kloppstech, K. (1991) Trends Biochem. Sci. 16, 181 -186. 9. Mattoo, A. K. & Edelman, M. (1987) Proc. Nut1 Acad. Sci. USA 84, 1497-1501. 10. Adir, N., Shochat, S. & Ohad, I. (1990) J . Biol. Chem. 265, 12563 - 12 568. 11. Mattoo, A. K., Hoffman-Falk, H., Marder, J . & Edelman, M. (1984) Pror. Nut1 Acad. Sci. USA 81, 1380- 1384. 12. Ohad, I., Adir, N., Koike, H., Kyle, D. J. & Inoue, Y. (1990) J . Biol. Chem. 265. 1972- 1979. 13. Baker, N. R. & Leech, R. M. (1977) Plant Physiol. 60, 640-644. 14. Adamska, I., Ohad, I. & Kloppstech, K. (1991) in Regulation of chloroplast hiogenesis (Argyroudi-Akoyunoglou, J., eds) NATO ASI, in the press. 15. Adamska, I., Ohad, I. & Kloppstech, K. (1992) Proc. Nut1 Acad. Sci. USA 89,2610-2613. 16. Lers, A., Levy, H.&Zamir, A. (1991)J. Biol. Chem.266, 1369813705. 17. Cline, K. (1986) J . Biol. Chern. 261, 14804-14810. 18. Chitnis, P. R., Nechushtai, R. & Thornber, P. (1987) Plant ,4401. Bid. 10, 3 - 11. 19. Paulsen, H., Rumler, U. & Riidiger, W. (1990) Planta (Berlin) 181,204-211, 20. Cohen, Y . , Steppuhn, J., Herrmann, R. G., Yalovsky, S. & Nechushtai, R. (1992) EMBO J . 11, 79-85. 21. Friemann, A., Schwarz, H. J. & Hachtel, W. (1991) in Regulation of chloroplast biogenesis (Argyroudi-Akoyunoglou, J., ed.) NATO ASI, in the press.
202 22. Maniatis, T., Frisch, E. F. & Sambrook, J. (1982) Molecular cloning, Cold Spring Harbour Laboratory, New York. 23. Roberts, B. E. & Paterson, B. M. (1973) Proc. Natl Acad. Sci. USA 70,2330-2334. 24. Neville, D. M. (1971) J . Biol Chem. 246,6328-6334. 25. Bonner, W. M. & Laskey, R. A. (1974) Bur. J. Biochem. 46,83 88. 26. Grossman, A. R., Bartlett, S. G., Schmidt, G. W., Mullet, J. E. & Chua, N. H. (1982) J. Biol. Chem. 257,1558-1563. 27. Fulsom, D. R. &Cline, K. (1988) Plant Physiol. 88,1146-1153. 28. Karlin-Neuman, G. A,, Kohorn, B. D., Thornber, J. P. & Tobin, E. M. (1985) J. Mol. Appl. Genet. 3, 45-61. 29. Schmidt, G. W., Bartlett, S. G., Grossman. A. R., Cashmore, R. & Chua, N. H. (1981) J. Cell. Biol. 91,468-478.
30. Gounaris, K., Brain, A. R. R., Quinn. P. J. & Williams, W. P. (1984) Biochim. Biophys. Acta 766, 198-208. 31. Adamska, I., Scheel, B. & Kloppstech, K. (1991) Plant Mol. Biol. 17,1055-1065. 32. Viro, M. & Kloppstech, K. (1980) Pluntu (Berlin) 150, 41 -45. 33. Chitnis, P. R., Harel. E., Kohorn B. D., Tobin, E. M. & Thornber, J. P. (1986) J . Cell Bid. 102, 982-988. 34. Cline, K., Fulsom, D. R. & Viitanen, P. V. (1989) J . Bid. Chem. 264, I4225 - 14232. 35. Reed, J. E., Cline, K., Stephens, L. C., Bacot, K. 0. & Yiitanen, P . V. (1990)Eur. J. Bioc:hem. 194, 33-42. 36. Viitanen, P. V., Doran, E. R. & Dunsmuir, P. (1988) J . Bid. Chem. 263, 15000-15007.
