The Transport of Proteins into Chloroplasts

0 downloads 0 Views 1MB Size Report
envelope membranes, the inner of which maintains a permeability barrier to .... and their posttranslational transport into their respective organelles was documented .... (0), negatively charged (-), and positively charged (+) chains are indicated. ..... phyll a, chlorophyll b, lutein, neoxanthin, and violaxanthin, in approximate.
ANNUAL REVIEWS

Further

Quick links to online content Ann. Rev. Biochem. 1986. 55:879-912

Copyright © 1986 by Annual Reviews Inc. All rights reserved

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

THE TRANSPORT OF PROTEINS INTO CHLOROPLASTS l Gregory W. Schmidt and Michael L. Mishkind Botany Department, University of Georgia, Athens, Georgia 30602 IPresent

address:

Department of Biochemistry and Microbiology, Cook College, Rutgers

University, New Brunswick, New Jersey 08903

CONTENTS PERSPECTIVES AND SUMMARy................... . . . . . . .... .......... . . .. ...................

879

OVERVIEW OF TRANSPORT OF PROTEINS THROUGH MEMBRANES ............ Cotranslational Transport....................................................................... Postt ranslational Transport.....................................................................

880 880 881

CHLOROPLAST TOPOLOGY: DIVERGENT PATHWAYS FOR PROTEIN IMPORT AND SORTING.......................................................................... Proteins of the Stromal Compartment......................................................... Transit Sequences of Stromal Protein Precursors........................................... Proteins of the Thylakoid Compartment ...................................................... Transit Sequences of Thylakoid Membrane Proteins .................................. . .... Envelope Membrane Proteins....... ...........................................................

882 882 886 889

893 894

ENVELOPE RECEPTORS ......... .. . . ............ . ......... . . . . . . .... . . . ......... . . . .. . . . . . . . .....

895

ENERGETICS OF TRANSPORT . . . . . . . . . . . . . . . . .. . .. . . . . . ...... . . . . . . . . . . . . . . ....... . . . .. . . . . . . . .

897

TRANSIT PEPTIDASES .... ..... . ..... ........... .... ....... ... ........ ........... ... ........ ......

899

DIRECTIONS AND PROSPECTS . . . .. . . . . . . . . . . . . . .................... ........................... Translocation to Inner Compartments......................................................... Transit Sequences........................... ........................... .......... . ................ Import Factors..................................................................................... Transport Mutants . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

903 903 905 906 908

PERSPECTIVES AND SUMMARY The means by which protein s become compartmentalized is an interest of those concerned with organelle biogenesis and presents a formidable area for research on membrane-protein interactions. Chloroplast proteins that are synthesized by 879

0066-4 I 54/86/0701-0879$02. 00

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

880

SCHMIDT & MISHKIND

cytoplasmic ribosomes typically are synthesized as higher-molecular-weight precursors. They contain NHz-terminal extensions, termed transit sequences, that are both necessary and sufficient for transport. The structural properties of the precursors have not been resolved completely, but they possess a number of functional characteristics that distinguish them from mature chloroplast pro­ teins. The precursors uniquely are able to interact with the outer envelope membranes of chloroplasts, apparently by associating with specific receptors. Next, by an ATP-dependent mechanism, the precursors pass through the two envelope membranes, the inner of which maintains a permeability barrier to low-molecular-weight solutes during this process. Upon transport into the chloroplast, the precursors undergo endoproteolytic maturation in a pathway that in some instances involves at least two steps. The maturation enzymes have been partially characterized but it is not established whether all precursors are processed by identical proteases. Many of the major proteins of chloroplasts reside in thylakoids, membranes that harbor the photosynthetic electron trans­ port components. Since thylakoids are not connected with the envelope, the proteins of this compartment that are synthesized by cytoplasmic ribosomes penetrate the envelope membranes, traverse the soluble stromal compartment,

and then integrate into the thylakoid or, in some cases, become localized in its lumen. This last process is unique to chloroplasts since in other organelles the site for stable membrane integration of proteins is not separated from their site of membrane transport. It will become clear that there are many outstanding problems to be solved with regard to the transport of proteins into chloroplasts. Because these are also areas of current research in other membrane and organelle transport systems, we occasionally draw attention to relevant work on protein secretion and import into mitochondria. Our emphasis, however, will be to review what is currently understood concerning the characteristics of precursors, the mechanisms for their transport through membrane bilayers, the maturation pathways and, finally, translocation of the chloroplast proteins to their ultimate organellar compartment.

OVERVIEW OF TRANSPORT OF PROTEINS THROUGH MEMBRANES Cotranslational Transport Among the most important breakthroughs toward understanding the mech­ anisms by which proteins are directed to particular subcellular compartments were the findings that proteins that are transported through membranes are synthesized as higher-molecular-weight precursors. This was extensively documented for secretory proteins of animals (1-3) and bacteria (4) and, later, for the storage proteins of plants (5). In these systems, precursors usually could

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

881

be detected only through the use of in vitro mRNA translation systems. In vivo, the precursor extensions, termed signal sequences, are generally removed cotranslationally in concert with the movement of nascent polypeptide chains through the membrane (6). Cotranslational transport has been demonstrated to be obligatory for many of these proteins (6, 7). As a means to prevent the wasteful production of precursors in a completely synthesized, nontransport­ able form, the signal sequence, emerging from ribosomes, serves as a prompt for the temporary arrest of translation (8). In eukaryotes, this is achieved through signal sequence binding by a complex consisting of 7S RNA and cytosolic proteins termed the signal recognition particle (SRP). The SRP­ ribosome-mRNA-nascent polypeptide complex remains translationally dor­ mant until bound to membrane "docking proteins" (also termed SRP receptors) on the surface of the endoplasmic reticulum (E.R.). After ribosome-binding components of the membranes stabilize the E.R. association of the translational apparatus, the SRP is released and translation resumes. The 15-30 amino acids of signal sequences, consisting of mostly hydrophobic amino acids, are pre­ sumed to interact with the membrane, initiate the transfer of the elongating protein through the membrane, and then are cleaved by membrane-bound "signal peptidases" located on the lumenal side of the E. R. The remainder of the secreted protein is injected into and through the membrane by a mechanism that is dependent upon polypeptide chain elongation. As evidenced from the num­ ber of gene loci in Escherichia coli that affect protein export, the machinery to regulate and facilitate transport of proteins through membranes is highly com­ plex (9). Posttranslational Transport In the late 1970s, the question of how proteins become associated with chloro­ plasts, mitochondria, peroxisomes, and glyoxysomes began to be studied. Initial approaches involved the use of cell-free protein synthesis systems for characterizing the primary translation products of mRNAs that direct cytoplasmic synthesis of organelle polypeptides. This first was performed with mRNA purified from the unicellular green alga, Chlamydomonas reinhardtii. Dobberstein et al (10) discovered that the small subunit of ribulose 1,5bisphosphate carboxylase/oxygenase (RuBPCase) is synthesized as a higher­ molecular-weight precursor. As compared to the 2-3-kilodalton signal se­ quences of secretory proteins, the extension of the small subunit precursor appeared to be longer with an apparent molecular mass of 3500 daltons. When incubated with algal extracts, the small subunit precursor could be converted to its mature, 14-kilodalton form. Another important discovery was that small subunit precursor mRNA was localized in preparations of free polysomes, contrasting markedly with the mRNAs for secretory proteins which are en­ riched in membrane-associated polysome fractions. These findings indicated

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

882

SCHMIDT & MISHKIND

that chloroplast proteins are transported through the two envelope membranes of chloroplasts by a posttranslational mechanism, a hypothesis consistent with the virtual absence of cytoplasmic ribosomes intimately associated with the organelles as determined by transmission electron microscopy (10). To test the hypothesis for posttranslational transport of proteins, methods to reconstitute the process in vitro were developed. This was first accomplished by simply adding in vitro translation products to chloroplasts isolated from vascu­ lar plant leaves. Although the initial studies differed in incubation buffers and method of chloroplast isolation, the results were in general agreement (11, 12). The precursors to the small subunits of pea and spinach were transported into isolated chloroplasts and processed to their mature forms. The use of postribo­ somal supernatants ( 1 1), measurements of isotope incorporation (12), and the use of protein synthesis inhibitors by both groups demonstrated unequivocally that import occurs posttranslationally. Evidence that transport across both plastid membranes had occurred included resistance of the imported small subunit to digestion by exogenously added proteases. Also, the matured protein was recovered from the organelle's soluble fraction as part of the RuBPCase holoenzyme; therefore, it had assembled with chloroplast-synthesized large subunit upon its localization in the stroma (11). I n subsequent years, higher-molecular-weight precursors of proteins of mitochondria and microbodies (peroxisomes and glyoxysomes) were identified and their posttranslational transport into their respective organelles was documented ( 13- 15). Overall, the mechanisms for transport of proteins into organelles are similar. However, it is becoming apparent that chloroplasts, mitochondria, and microbodies utilize rather different modes of protein import due to the number of membranes that are traversed and the energetics of this process.

CHLOROPLAST TOPOLOGY: DIVERGENT PATHWAYS FOR PROTEIN IMPORT AND SORTING Chloroplast proteins that are nuclear-encoded can reside in one of six sub­ organellar compartments. These are the outer and inner membranes of the envelope, the envelope intermembrane space, the soluble or stroma fraction, the thylakoid membrane and, finally, the lumen compartment of thylakoids. It has been proposed that translocation of cytoplasmic products to these specific sites is achieved by a variety of receptors which interact with specific determi­ nants of the imported precursors (16). Whether this is the case has not been proven but this hypothesis is supported by circumstantial evidence. Proteins of the Stromal Compartment Plastids are responsible for many of the major biosynthetic activities of plant and algal cells. The principal pathways found, at least in part, in the stroma of

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

883

plastids in all plant cell types include those for nitrogen assimilation (17) and synthesis of fatty acids (18), many amino acids (19), terpenoids (20), and porphyrins (21). As far as is known, all of the above processes are accom­ plished by enzymes that are nuclear gene products synthesized by cytoplasmic ribosomes. Other plastid functions require at least some products of the organelle's genome as well as a host of nuclear gene products (22). Prominent among these are components for protein synthesis: the organelle genome encodes ribosomal and tRNAs, a few ribosomal proteins, and elongation factors (23). The replication and transcription machinery, however, is derived from nuclear genes in mOst systems (but see 24). In the case of the highly specialized photosynthetic organelles, chloroplasts, the major soluble enzyme, RuBPCase, contains a polypeptide, the large subunit, encoded by the organelle. All other chloroplast enzymes of the reductive pentose phosphate cycle for carbon dioxide assimilation and starch biosynthesis are synthesized in the cytoplasm (22). Thus, the vast majority of stromal proteins must be transported into this compartment after they are synthesized by cytoplasmic ribosomes. The small subunit of RuBPCase is one of the most abundantly synthesized proteins in photosynthetic cells. High levels of small subunit mRNA enabled the early protein transport studies, summarized above, to be accomplished in spite of suboptimal conditions. Also, the highly expressed small subunit genes have been especially amenable to isolation via cDNA cloning (25). Con­ sequently, small subunit precursors from several vascular plant species have been characterized by transport and nucleotide sequence analyses (26-34). Although the amino acid sequences of the mature polypeptides are highly homologous, the NHrterminal transit sequences of small subunit precursors have only a few regions of homology (Figure 1). In photosynthetic prokaryotes, where membrane transport of the small subunit does not occur, the small subunit is not made as a precursor (35). Likewise, in some eukaryotic algae, such as the chromophyte Olisthodiscus, the small subunit is a chloroplast rather than a nuclear gene product and also is synthesized in its mature form (36). Thus, the occurrence of a precursor of the small subunit correlates with the necessity for posttranslational transport through envelope membranes. A vast number of other stromal proteins are posttranslationally imported into chloroplasts in vitro, especially when more optimal incubation conditions are employed (37). However, partly because their mRNA levels are low, few of these proteins have been characterized extensively at the molecular level. As an example, attempts to characterize the primary translation product for fructose1 ,6-bisphosphatase failed because of low mRNA abundance or low incorpora­ tion of the radioactive amino acids used for in vitro translation (37). In spite of hindrances of this kind, information concerning other nuclear-encoded chloro­ plast proteins is beginning to emerge.

884

SCHMIDT & MISHKIND

RuBPCase Small Subunits 00

Lemna

00

Ref .

+

0

+00

- 00

00

+ 0

H2N-MASSMMA STAAVARVGP AQTNMVGPFN GLRSSVPFPA TRKANNDLST LPSSGGRVSC M A R A e V

29

00 0 0 0+0 a +0 a 0++ - 00 a + MASSVLS SAAVATRSNV AQANMVAPFT GLKSAASFPV SRKQNLDITS IASNGGRVQC M

Tobacco

M

R

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

0++

o

00 0

00 0+ a +

a

+0 0

++

0- 00

00

MASMISS SAVTTVSRAS RGQSAAVAPF GGLKSMTGFP VKKVNTDITS ITSNGGRVKC M

Pea

00

00

00

+

a

+0

0++0

- 00

a

+

MASSM ISSPAVTTVN RAGAGMVAPF TGLKSMAGFP TRKTNNDITS IASNGGRVQC M

Soybean

00 00

+00

0 +0 00

0 0 0

+00 a

+

t

+00 ""

+

+

MAV! AKSSVSAAVA RPARSSVRPM AALKPAVKAA PVVAPAEAND M --

Chlamydomonas

S

28

+ +

MAPAVMA SSATTVAPFQ GLKSTAGLPI SCRSGSTGLS SVSNGGRIRC M

Wheat

27

34

63

A

Ferredoxin 00 00 0

0 0

+

00

0

+

0+ + 0

MASTLSTLS VSASLLPKQQ PMVASSLPTN MGQALFGLKA GSRGRVTA M A

Silene

75

LHCP II Polypeptides 0000 0

00 0

+

+

00

-

.... 0

+

MAASSSS SMALSSPTLA GKQLKLNPSS QELGAARFTM R

Pea

00

00

0

+ +--

++

0- + 0

MAASSAI QSSAFAGQTA LKQRDELVRK VGVSDGRFSM R

Lemna

o

0000

+ 00000 - 0

+

+

0

NY

P

T

a a a 0000

+

+

Q

R

A

- +

00

130

....

MAAAT MALSSSSFAG KAVKLSSSSS EITGNGKVTM R

Petunia

128

+

93

+

MAAT TMSLSSSSFA GKAVKNLPSS ALlGDARVNM R

129

o 00 0 0 + 0000+ 0 + 0 + 0 + 0 +0 MATVTS SAAVAIPSFA GLKASSTTRA ATVKVAVATP RMSIKASLKD VGVVVAATAA AGILAGNAMA A

133

Wheat

Plastocyanin Silene

I

-60

Figure 1

I

-50

I

-40

I

-30

I

-20

I

-10

I

+1

Transit sequences of proteins transported into chloroplasts. Amino acids with hydroxyl

(0), negatively charged (-), and positively charged (+) chains are indicated. Underlined regions are conserved sequences that may serve for receptor and/or transit peptidase recognition. One letter code is: A alanine

C cysteine

D aspartate

E glutamate

F phenylalanine

G glycine

H histidine

I isoleucine

K lysine

L leucine

M methionine

N asparagine

P proline

Q glutamine

R arginine

S serine

T threonine

V valine

W tryptophan

Y tyrosine

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

885

Pyruvate orthophosphate dikinase (PPDK) is responsible for regeneration of phosphoenolpyruvate for CO2 fixation in the mesophyll chloroplasts of C4 plants, but is also found in the cytoplasm of other tissues and in C3 plants. In maize, a C4 plant, the chloroplast-localized form is synthesized as a precursor with an apparent molecular mass of 110 kilodaltons whereas the mature protein is 94-95 kilodaltons (38, 39). The size of the PPDK transit sequence therefore appears to be much larger (> 100 amino acids) than that of the small subunit precursors even though both proteins are localized in the stroma. In nonphotosynthetic tissues, cytoplasmic PPDK, though related im­ munologically to the plastid form, is not synthesized as a precursor since the translation product of its mRNA is 94 kilodaltons (39). As suggested by Aoyagi & Bassham (39), it is possible that the only difference between the two mRNAs is the occurrence of coding sequences for the transit sequence. Thus, the primary transcript of a single structural gene, containing an exon encoding the transit sequence, could be processed in alternative pathways to give rise to

mRNAs encoding for either the chloroplastic or cytosolic forms of the enzyme. An analogous process occurs in yeast for formation of secretory and cytosolic forms of invertase (eg. Ref 40) and for a tRNA methylase for both the cytosol and mitochondria (41). Alternative transcript maturation pathways could also be a means of generat­ ing chloroplast and cytosolic forms of enzymes in the same cell. Plastid and cytoplasmic aldolases of wheat, except for their compartmentation, are in­ distinguishable (42). In contrast, different genes must encode cytosolic and chloroplast aldolases in spinach and com because antibodies against these enzymes recognize completely different antigenic determinants (43, 44). Pro­ plastids of castor bean endosperm appear to contain 6-phosphogluconate de­ hydrogenase identical in size and enzymatic properties to that of the cytosol although these differ slightly in net charge (45). Similarly, 1-2% of leaf calmodulin could become chloroplast localized (46, 47); however, chloroplast association of calmodulin is an unsettled matter and there is evidence that it functions at the level of the chloroplast (and mitochondrial) envelope mem­ branes as opposed to having physiological functions inside the organelles (48-50). Studies on the structures of genes encoding bimodally localized proteins will determine if retention or removal of exons encoding transit sequence is a causative mechanism for differential sorting of some of the above enzymes, particularly those that are immunologically related. It should be noted that cytosolic and mitochondrial forms of rat liver fumarase share anti­ genic determinants and are enzymatically indistinguishable but, unexpectedly, differ in sequence at their amino termini (51). Thus, posttranscriptional pro­ cesses may not necessarily account for the occurrence of immunologically related proteins found in different cellular compartments.

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

886

SCHMIDT & MISHKIND

In most instances, cytosolic and chloroplastic forms of similar enzymes are indisputably products of different genes as has been shown for triose phosphate isomerases (52,53) and glyceraldehyde-3-phosphate dehydrogenases (54). In the latter case, the primary translation products of the mRNAs for two subunits of the chloroplast forms were 4-12 kilodaltons larger than the mature (36-43kilodalton) subunits, depending on the plant species. The great variability in the apparent length of the putative transit sequences, 36-110 amino acids, led to consideration that the precursor extension does not contain the functional determinant for chloroplast import. It was suggested that, instead, transit sequences merely serve as effectors for folding of the precursors in order to expose "true" recognition sites; precursor maturation would cause con­ formational changes that ensure that the proteins remain in the chloroplasts when the enzymes acquire their functional form (54) . This notion, espoused in earlier reviews of protein transport through membranes (55, 56), is not sup­ ported by recent data. In recent years, precursors of many other stromal proteins that are nuclear gene products have been identified. These include nitrite reductase (57), proteins of chloroplast ribosomes (58), phytoferritin (59), ferredoxin-NADP oxidoreductase (37), UDP-glucosyl transferase (waxy protein) of maize endo­ sperm starch granules (60), the acyl carrier proteins for fatty acid synthesis (61), and a subset of stress proteins that become chloroplast localized (61a, 62). In some of these studies, import of the precursors into isolated chloroplasts has been accomplished, resulting in their maturation. Transit Sequences of Stromal Protein Precursors Until sequence analysis was available, it was a matter of speculation whether the amino acid extension of the small subunit precursor was at the amino or carboxy terminus. The early studies of Dobberstein et al indicated that there was a single entity for the precursor chain because a rather large fragment could be released from the Chlamydom onas small subunit precursor upon its matura­ tion with algal extracts (10). Microsequencing techniques,which had been of enormous utility in the characterization of the precursors of secretory proteins (6), were equally useful in determining that small subunit precursors of Chlamydomonas contain 44 amino acids at their NH2 termini that are not present in the mature protein (63). This extension was termed a "transit sequence" since its presumed function (posttranslational transport) as well as composition (mostly apolar and polar residues) differed significantly from the mostly hydrophobic amino acids characteristic of the signal sequences of secretory proteins. The transit sequence was precisely removed upon incuba­ tion with extracts from Chlamydomonas, yielding a correctly matured small subunit. Moreover, the 44 amino acids of the transit sequence account fully for the molecular weight differences observed upon gel electrophoresis of pre-

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

887

cursor and mature forms of the small subunit (63). Therefore, it seemed unlikely that any of the carboxy termini are responsible for the larger size of the precursors. The precise roles of transit sequences in the import and maturation of precursors into organelles have not been resolved. Transit sequences of quite different lengths and amino acid sequence share functional properties: small subunit precursors of monocots and algae can be imported in vitro into chloro­ plasts from dicots (64, 65). Likewise, chloroplasts in plants transformed with plasmids containing small subunit genes from other species will import the heterologous proteins, process them to their mature size, and utilize them for formation of multisubunit complexes (66). Transport is dependent upon the presence of transit sequences because if they are partially or completely re­ moved from precursors, the proteins cannot be transported into chloroplasts in vitro (65). Recombinant plasmids containing sequences for transit sequences have been constructed with foreign genes placed in frame for their in vivo expression: the resulting products also undergo chloroplast transport with concomitant maturation of the chimeric precursors (67, 68). This elegantly demonstrates that transit sequences are all that are required for the import of proteins into chloroplasts. Thus, genetic engineering techniques now can be used to modify transit sequences as a means of identifying their functional determinants. Also, high resolution analysis of protein transport has been achieved with homogeneous preparations of mRNA (69). Therefore, it is feasible to use plasmids containing mutated genes for their transcription and translation in vitro followed by assays of the transport activity of altered transit sequences. At this time, however, information concerning the effects of transit sequence mutations have not been published. Consequently, we can discuss their roles only on the basis that evolutionary preservation of transport activity results from conservation of at least some of the structural features of the transit sequence. The lengths of small subunit precursor transit sequences vary from 44 amino acids (Chlamydom onas) to 57 amino acids in tobacco and Lemna. Figure 1 displays the published transit sequences of the precursors of chloroplast pro­ teins, which, except for small subunit precursor of Chlamydom onas, have been deduced from nucleotide sequences of cDNA or genomic clones. The most striking features of small subunit precursor transit sequences are their net positive charge and, for the vascular plants, highly homologous primary struc­ tures near the maturation site. The net positive charge, concentrated in the maturation site, has been proposed to be important in interaction of the cytosol­ ic forms of the precursors with the negatively charged envelope membranes of chloroplasts (16). A net positive charge appears to be essential also in the import and maturation of precursors of mitochondrial proteins; when the arginine analogue canavanine, which has a pKa of 6.6, is used during labeling

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

888

SCHMIDT

&

MISHKIND

studies of HeLa cells transfected with high copy plasmids, precursor and, "intermediate" processing forms of ornithine transcarbamylase accumulate (70). Von Heijne has used comparative sequence analyses to identify functional determinants in the signal sequences of secretory proteins (71). Following his lead, we have identified three domains in small subunit transit,sequences between which correlations of structure and function can be inferred' (65). S tructurally, Domain I consists of the first 12-25 NHrterminal amino acids of which 24-40% are threonine and serine and approximately 5% are arginine or lysine. The great variability in length and primary structure of Domain I among different sp�cies'appears to preclude a role for it in specific interactions with chloroplast receptors. However, the hydrophilic character of Domain I may be e ssential for keeping transit sequences on the protein surface during synthesis of the precursors. Domain II varies in length from 1 1 to 18 residues in the small subunit precursors and is underlined in the middle segments of transit sequences in Figure 1. This region exhibits evolutionary conservation with regard to ,the series of residues surrounding glycine-leucine-lysine. This site is where the

small subunit precursors evidently experience the first of at least two maturation e vents upon chloroplast import (65), discussed in more detail below. Because this region is the only part of small subunit transit sequences that is well conserved in plants and algae, we have proposed it constitutes the binding site for envelope receptors (65). Domain II is also positively charged and contains one or two prolines which may allow for distinctive secondary structure characteristics in the middle of the transit sequence. Proline may also facilitate small subunit precursor conformational changes as it traverses the envelope membranes. This amino acid reversibly undergoes cis/trans isomerization at its nitrogen atom during protein folding (72, 73). Pro line is also an abundant residue in the hinge region of immunoglobulins, presumably facilitating protein flexibility (74). Except in the case of the Chlamydom onas small subunit precursor, Domain III is serine/threonine-rich, has a net positive charge of 2-3, and possesses cysteine at the maturation site. We have suggested that the conserved primary sequence in this region of vascular plant transit sequences signifies its im­ portance for precursor maturation but is of minor importance for chloroplast transport (65). The small subunit precursor ofChlamydomonas can be imported by chloroplasts from spinach and pea even though it lacks the Domain III sequence of vascular plant small subunit precursors (65). On the basis that there i s no structural-functional correlation for chloroplast import in this region of transit sequences, we conclude that Domain III does not contain a binding site for chloroplast receptors. Ferredoxin, a polypeptide that is associated loosely with the stromal surface of thylakoid membranes, may be imported into chloroplasts by a similar

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

889

mechanism as that for the small subunit precursor. The ferredoxin precursor is about 5 kilodaltons larger than the mature (22-kilodalton) protein (59, 75). Its transit sequence (75), 48 residues in length, contains a sequence, underlined in Figure I, which resembles Domain II of the small subunit precursor transit sequence. Hence, pre-ferredoxin might use the same receptors as the small subunit precursors. As discussed below, it is probable that maturation of pre-ferredoxin involves the same enzyme as the small subunit because of homologies at the processing site. Aside from these limited similarities in primary structure, there are no other homologies in primary sequence or, as shown by Smeekens et ai, in the hydropathicity profiles of the small subunit and ferredoxin transit sequences (75). It is likely that the tertiary structure of precursors of chloroplast proteins is of major importance in their interactions with envelope membrane receptors. At present, only predictions of their secondary structures can be made and this has not been pursued extensively. In the analysis of the small subunit precursor transit sequence from Chlamydomonas, Chou-Fasman paradigms predict that the first 1 5 residues of Domain I have a propensity for a-helix formation (63). The remainder of the transit sequence appears to be a random coil bordered by 13 turns. Recently, Pongor & Szalay developed a quantitative method for assess­ ing secondary structure homologies of peptides and applied this to analyses of the RuBPCase small subunit precursors from several vascular plant species (76). As compared to the NHz terminus of the small subunit of the cyanobacteri­ um Synechococcus (35), which is not made as a precursor, transit sequences of vascular plants are disposed toward J3-structure formation. Pongor & Szalay consider their calculations to be consistent with the general tendency of proteins that undergo membrane translocation to form NHz-terminaIJ3-structures (77). However, this rule is not universal because the signal sequence of the E. coli LamB protein (J3-lactamase) must form a-helixes (78): some LamB mutations that block cotranslational transport concordantly affect the ability of signal sequences to assume a-helixes in hydrophobic environments (78, 79). Proteins of the Thylakoid Compartment Formation of thylakoids is a light-regulated process in vascular plants . Photoactivation of phytochrome leads to increased transcription of nuclear genes e n co d ing chloroplast prote ins , and p hotored uction of pro­ tochlorophyllide'-initiates chlorophyll synthesis (80). Except for NADPH­ protochlorophyllide reductase (NPCR) , little is known about the synthesis and function of proteins in the prolamellar membranes of etioplasts in dark-grown plants. NPCR is a major product of cytoplasmic protein synthesis in plants kept in darkness for prolonged periods and appears to be responsible for the elabo­ rate morphology of prolamellar bodies (8 1). The mRNA for NPCR yields a 44-kilodalton precursor of the 36-kilodalton polypeptide when translated in the wheat germ system (8 1 -83). As for many photosynthetic proteins, accumula-

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

890

SCHMIDT & MISHKIND

tion of NPCR is light-regulated via phytochrome, but in this case illumination is followed by dramatic decreases of its mRNA levels, especially in monocots (82, 83). Shortly after the onset of illumination, the enzyme begins to be selectively degraded, a process that is enhanced in vitro when its substrates become depleted (84, 85). A membrane-associated protease appears to be the cause for NPCR degradation and disorganization of prolamellar bodies which occurs during the early stages of greening. Meyer & Kloppstech (86) have suggested that NPCR proteolysis may be due to a nuclear gene product whose mRNA rises and falls during the first 5 hours of illumination of etiolated pea seedlings . In vitro translation of the mRNA produces a 24-kilodalton precursor which is processed to a 1 7-kilodalton thylakoid membrane polypeptide upon in vitro transport into chloroplasts. There is no direct evidence about this protein's role in developing chloroplasts and other suggestions were made concerning its possible function in mediating light-dependent formation of photosynthetic membranes (86) . In fully developed chloroplasts of light-grown plants , thylakoids contain several chloroplast-synthesized integral membrane proteins of the photosyn­ thetic electron transport pathway (22, 23). These include the chlorophyll a-binding polypeptides of Photosystem 1 and II reaction centers , at least two quinone-binding polypeptides, cytochromes b6, b559, andj, and two ofthe three subunits (I and III) of the CFo moiety of the ATP synthase . The few extrinsic proteins of the thylakoid that are made in the organelle are the a, 13 , and e subunits of the CFt ATP synthase. Thus, most of the extrinsic proteins of thylakoids are imported from the cytoplasm. The light-harvesting chlorophyll alb-binding proteins that associate either primarily with Photosystem II (LHCP II) or Photosystem I (LHCP I) and subunit II of CFo ATP synthase are notable in that they are integral membrane proteins that are products of cytoplasmic ribosomes (22, 87). When LHCP II complexes are purified from thylakoid membranes and subjected to electrophoresis under denaturing conditions, several discrete polypeptides with apparent molecular masses of 25 -30 kilodaltons are detected ( 88-90). These are the most abundant proteins of photosynthetic membranes, are immunologically related (9 1 , 92), and are encoded by highly homologous nuclear genes; in Petunia, the LHCP multigene family contains at least 16 loci (93). The LHCP II polypeptides noncovalently bind antenna pigments , chloro­ phyll a, chlorophyll b, lutein, neoxanthin, and violaxanthin, in approximate molar ratios of 6; 6; 3; I; 1 per mole of apoprotein (90, 94). In addition, the LHCP II proteins assemble with each other (95), probably to form hexameric complexes in thylakoids (96, 97). Finally, they function to transfer light energy to Photosystem II reaction centers primarily by interactions that appear to be dependent upon monogalactosyl diacylglyceride (98), one of the major thyla­ koid lipids. The LHCP II polypeptides are highly hydrophobic (99, 100), and are integrated into thylakoids with 1-2 kilodaltons of their NH2 termini

CHLOROPLAST PROTEIN TRANSPORT

89 1

oriented toward the stroma ( 1 0 1 ). Phosphorylation of the exposed domains by a thylakoid-associated protein kinase causes dissociation of LHCP II complexes

from Photosystem II reaction centers (1 02). The surface moieties of the LHCP II polypeptides

are

implicated also in thylakoid adhesion to form grana stacks

(103). This summary is incomplete but is intended only to emphasize that the LHCP II apoproteins become able to interact with many membrane components

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

as well as with each other upon their import into chloroplasts.

Genetic evidence that genes encoding the LHCP II polypeptides are localized in the nucleus was provided by Kung et aI, who analyzed Nicotiana tabacum,

Nicotiana glauca, and their hybrids. In these species chloroplast DNA

is

inherited maternally but inheritance of tryptic peptide patterns of the LHCP II polypeptides is biparental ( 1 04). Their synthesis by cytoplasmic ribosomes was established by in vivo labeling studies with inhibitors of cytoplasmic and chloroplast protein synthesis (89, 1 05-1 07), and their absence among the proteins made by isolated chloroplasts (108). Chloroplast-synthesized proteins are

not involved in the import, maturation, or thylakoid insertion of the LHCP II

polypeptides: long-term growth of Chlamydomonas in the presence of in­ hibitors of chloroplast protein synthesis has no effect on their biogenesis (87). Accumulation of the LHCP

II polypeptides is light-dependent and is subject

to transcriptional control by a phytochrome-dependent process (109-1 1 3). This was determined initially by Apel & Kloppstech, who also provided the first data

concerning synthesis of the LHCP polypeptides as higher-molecular-weight precursors (109). They found that antibodies against the 25-kilodalton LHCP II polypeptide of barley immunoprecipitate a 29.5-kilodalton polypeptide from in vitro translation mixtures. The precursor-product relationship was sub­ stantiated by partial peptide mapping. Clarification of whether the LHCP II precursors are imported into chloroplasts by a direct, cotranslational mode, or

via a posttranslational pathway similar to that for the small subunit precursor, depended on improving the sensitivity of the in vitro transport system (37, 1 1 4). Enhanced resolution of LHCP II transport was also achieved by using plastids deficient in endogenous LHCP II, obtained from the inner leaves of Romaine lettuce (89).

In the case of pea leaf mRNA, in vitro translation products contain

two polypeptides of 33 and 32 kilodaltons that are immunoprecipitated with antibodies to the major LHCP II polypeptide of

Chlamydomonas (89). The II

translation products were shown to be structurally related to LHCP

polypeptides of 27-28 kilodaltons by peptide mapping analysis. When recov­ ered from postribosomal supernatants of in vitro translation mixtures and incubated with trypsin, the LHCP II precursors are completely degraded. This demonstrated that the precursors are soluble in aqueous solutions and do not become sequestered in membrane vesicles that conceivably occur in the wheat germ translation system. Therefore, one function of the precursors' amino acid extension is to affect the overall hydrophilic character of these polypeptides.

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

892

SCHMIDT & MISHKIND

The amino acid extension of the LHCP II precursors also appears to function as a transit sequence as shown by import of the precursors into isolated lettuce or pea chloroplasts (89). During in vitro transport, the precursors are converted to their mature size, are integrated fully and correctly into the thylakoid membrane, and bind chlorophylls to form pigment-protein complexes. Pre­ existing pools of chlorophylls support formation of LHCP II complexes since these were formed from precursors imported into chloroplasts in darkness (89). No precursors, processing intermediates, or mature forms of the in vitro synthesized LHCPs were recovered with either the chloroplast envelope or stroma fractions. Since lettuce chloroplasts mature and integrate the pea LHCP I I polypeptides in the same manner as pea chloroplasts, it was apparent that the import and maturation pathways for these polypeptides have been evolutionari­ ly conserved. Moreover, how they become topologically organized in thyla­ koids is an inherent property of these proteins (89). The other chlorophyll binding proteins that are nuclear encoded are the LHCP I apoproteins (D. L. Herrin, F. G. Plumley, G. W . Schmidt, submitted). Although their p recursors have not been identified, it has been shown that five polypeptides that associate with Photosystem I complexes are also transported into chloroplasts by a posttranslational process and are recovered in high­ molecular-weight complexes with the Photosystem I reaction centers (115). The intermediate steps in the translocation of the LHCP polypeptides to the thylakoid have been difficult to resolve. As indicated above, concomitant chlorophyll synthesis is not required for their import and thylakoid integration since these steps can be performed in darkness in vitro (89). One question that appears to be settled concerns the role of chlorophyll b in assembly of the proteins into thylakoids. In chlorophyll �eficient mutants of vascular plants, . and in phenocopies produced by growing seedlings under intermittent illumina­ tion, thylakoids are severely deficient in both LHCP I and LHCP I I (116). However, these plants contain normal levels of mRNAs for the LHCP II polypeptides in association with cytoplasmic polysomes (109, 117-119). From pulse-labeling and in vitro transport studies, the LHCP II polypeptides can be synthesized in these plants whereupon they are transported into chloroplasts and associate with thylakoids (118, 120). Thus, chlorophyll b does not partici­ pate in protein translocation to the thylakoids, but it is necessary to confer protection of the LHCP I I apoproteins from degradation by a highly specific, chloroplast-localized protease. The lack of a chlorophyll b requirement for integration of LHCP apoproteins is also evident in studies of a chlorophyll b-Iess mutant of Chlamydomonas which possesses thylakoids with normal levels of LHCP II apoproteins ( 1 2 1 ). The final class of thylakoid membrane proteins are those that function in the thylakoid lumen. These proteins are extrinsic components and include three polypeptides of the water oxidation complex associated with Photosystem II (122) and plastocyanin (123), whose copper atom carries electrons from the

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

893

cytochrome Jlb6 complex to Photosystem I reaction centers . Each of these proteins is separated from its cytoplasmic site of synthesis by three membranes. In the initial studies of the primary translation products of plastocyanin mRNA from pea leaves (37), the precursor exhibited an electrophoretic mobil­ ity of 25 kilodaltons. This is more than twice the size of the mature protein which, on the basis of amino acid sequences ( 1 24), is 10 kilodaltons in a wide range of photosynthetic organisms. It was proposed that the transit sequence of this protein was considerably larger ( 1 30-1 40 amino acids) than that of other precursors because it must be transported through thylakoid as well as envelope membranes (37). It still is unclear how plastocyanin translocation into the thylakoid space is achieved, but the work of Bohner et al indicates that copper plays a role (1 25). They found that growth of the alga Scenedesm us acutus in copper-free media leads to accumulation of a 14-kilodalton polypeptide that is immunoprecipitab1e with plastocyanin antibody. In contrast, copper dcficiency in Chlamydomonas results in the absence of detectable plastocyanin even though these cells have high levels of its mRNA (126). When immunoprecipi­ tated from in vitro translation mixtures, the plastocyanin precursor from Chlamydomonas has an electrophoretic mobility of 1 7 kilodaltons while, in the same gels, the mature form migrates as a 6-kilodalton polypeptide ( 1 26). Presumably, copper is necessary for protection of either the precursor or mature forms of plastocyanin from degradation, assuming the mRNA is translated in vivo. Copper-deficient Chlamydomonas cells synthesize cytochrome C553 «6 kilodaltons), a plastocyanin substitute in algae. In vitro translation of cyto­ chrome CSS3 mRNA yields a product with an apparent size of 14 kilodaltons ( 1 26). The work on plastocyanin indicates that transit sequences of nearly the size of the mature polypeptides may be required for their transport to the thylakoid lumen. However, studies of the water oxidation polypeptides indicate that large transit sequences are not obligatory for this process. In spinach, these polypeptides have molecular masses of approximately 34, 23 , and 1 6 kilo­ daltons, while their precursors are, respectively, 40, 33, and 26 kilodaltons ( 1 27). Thus, their transit sequences vary from the size of RuBPCase small s ubunit precursors to that of NADPH-dependent glyceraldehyde 3 -phosphate dehydrogenase of the stroma. Westhoff et al have followed the import of precursors to the water oxidation polypeptides into spinach chloroplasts; they become localized inside thylakoid membranes as mature proteins ( 1 27). These workers suggested that precursor maturation occurs at or in the thylakoids but were unable to demonstrate this directly. Transit Sequences of Thylakoid Membrane Proteins Genomic and cDNA clones for LHCP II polypeptides of pea (1 28), wheat ( 129), Petunia (93), and Lemna ( 130) have been isolated and sequenced. In the case of the RuBPCase small subunit, the transit sequences and two amino acids

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

894

SCHMIDT & MISHKIND

of the mature polypeptide generally are coded by exons (e.g. 28), consistent with hypotheses of the role of exons in encoding for different functional domains in eukaryotes ( 13 1 , 132). However, the genes for preLHCPs, except for Lemna, contain no intervening sequences . The transit sequences of the LHCP II precursors are NHrterminal as determined by comparisons of the deduced sequences with that of a partial NHrterminal sequence of one of the major LHCP components of pea (10 1 ). The transit sequences of these proteins are 37-38 residues in length and, for the species that have been examined so far, exhibit a high degree of homology (Figure 1). Although they are serine/ threonine-rich and possess an overall net positive charge, there is minimal homology with transit sequences of small subunit precursors. The main excep­ tion is the three-amino-acid segment near the processing site, indicating the LHCP II precursors and soluble proteins are matured by similar, if not identical, endoproteases. Since the LHCP II precursor transit sequences are well­ conserved evolutionarily, it is difficult to identify functional determinants such as those for receptor binding. However, their hydrophilicity undoubtedly promotes the solubility of the membrane precursors upon their synthesis by free cytoplasmic polysomes. A cDNA clone to plastocyanin mRNA has been obtained and sequenced ( 1 33). The precursor contains a smaller NH2-terminal transit sequence than was estimated from gel electrophoresis analysis of the translation product of pea mRNA (37). Pre-plastocyanin is actually 16. 6 kilodaltons due to a transit sequence of 66 amino acids (Figure 1). There are no homologies with any of the transit sequences of other precursors, although serine and threonine are prev­ alent in the pre-plastocyanin's NH2 terminus. Smeekens et al observed that the processing site is preceded by a region of 20 uncharged amino acids which could form a membrane-spanning region and proposed this contributes to translocation of the precursor through the thylakoid membrane (133). Envelope Membrane Proteins Chloroplast envelope membranes serve many roles in plant cells besides main­ tenance of the photosynthetic compartment. Comprehensive reviews of their roles in metabolite and ion transport ( 1 34) and their biosynthetic activities (lipid and terpenoid synthesis) ( 1 3 5 , 1 36) have been published recently . Like the outer membranes of gram-negative bacteria ( 1 37) and mitochondria ( 1 38), the outer membrane of chloroplast envelopes is permeable and allows unhindered passage of small solutes ( 1 34). Proteins smaller than 9-10 kilodaltons , includ­ ing trypsin, also can diffuse easily into the intermembrane space; this is due to pores with a calculated diameter of 2 . 5-3 nm ( 1 39). Although both the inner and outer membranes have been characterized in terms of their composition ( 140-1 45), the pore polypeptides have not been identified. The only proteins of either envelope membrane that have assigned functions are the 30-kilodalton

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

895

phosphate translocator ( 146) and, possibly, the dicarboxylic acid transporter . ( 1 47) . Another similarity between the chloroplast envelopes, mitochondrial mem­ branes, and the outer membranes of bacteria is the occurrence of zones of adhesion at intervals between the proximal membranes . So far, the characterization of adhesion zones has been mostly structural ( 1 6 , 148- 1 5 1 ) , although in bacteria they have been implicated in the translocation of proteins from the inner to the outer membranes ( 149, 1 52), and in the case of chloro­ plasts the transport of proteins from the outer to inner membrane compartments ( 1 6) . The physical and biochemical nature of the membrane connections is unclear but they do not allow for continuity of the two lipid bilayers: in chloroplasts , the inner and outer membranes have quite distinct lipid, enzyme, and pigment compositions ( 1 39, 142-145). Most, if not all, chloroplast envelope membrane polypeptides are nuclear gene products (136), but only one study has appeared concerning their bioge­ netic pathways ( 1 53). Two inner envelope polypeptides, the 30-kilodalton phosphate translocator and a 36-kilodalton protein, are synthesized in wheat germ extracts as precursors with molecular masses I I and 2 kilodaltons greater, respectively, than the mature forms (153). A major outer envelope protein of 22 kilodaltons is synthesized in vitro as a 32-kilodalton precursor ( 153). It is not yet known whether these higher-molecular-mass forms possess amino terminal extensions or if they are competent for cotranslational or posttranslational integration into chloroplasts. The occurrence of a higher-molecular-mass pre­ cursor for an outer envelope membrane polypeptide contrasts with the outer membrane polypeptides of mitochondria; these apparently do not experience proteolytic maturation during membrane assembly (154-156 but see 1 57).

ENVELOPE RECEPTORS Studies with isolated intact chloroplasts have shown that precursors bind receptors exposed at the cytoplasmic surface ofthe outer envelope membrane as a first step in their import. Proteins bound to the outer surface of repurified plastids can be distinguished from those in internal compartments by their sensitivity to nonpenetrating proteases. Although trypsin and chymotrypsin have been employed for this purpose, thermolysin is more selective because it cleaves a subset of outer envelope membrane proteins without affecting those of the inner membrane ( 142, 158). Grossman et al ( 1 14) found that transport, but not binding, of precursors, is greatly reduced when incubations are performed at 4°C. At least some of the precursors that bind in the cold are transferred into the chloroplast when the temperature is raised. As another means of uncoupling binding from transport, Cline et al employed the K+ IH + ionophore nigericin in studies of small subunit and LHCP II polypeptide binding and import (69);

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

896

SCHMIDT

&

MISHKIND

for this work they used translation products of mRNAs purified by hybridiza­ tion to cDNA clones. As compared to untreated chloroplasts, the number of precursor polypeptides that accumulate at the organelle surface in the presence of nigericin increases two to fourfold for the small subunit precursor and up to twofold for the LHCP II precursors . When the ionophore is subsequently removed, 80% of the small subunit precursors and 25% ofthose for the LHCP II polypeptides are transported into the reisolated chloroplasts. Thus, at least some of the binding in the presence of the uncoupler is of functional signifi­ cance to the transport process. From independent quantitative assays, it has been estimated that a chloroplast can bind between 3000 and 5000 molecules (69, 1 59). Pfisterer et al calculate that this binding capacity would enable rapid import of the large amounts of chloroplast proteins that are synthesized during light-induced chloroplast development ( 1 59). Isolated envelopes also selectively bind precursors when incubated with in vitro translation products from total cytoplasmic mRNAs (22, 1 59). Pfisterer et al found that purified thylakoid membranes do not selectively bind organelle precursors ( 1 59). In contrast, Ellis reported that both envelopes and thylakoids were equal in their avidity for precursors while red blood cell membranes were without specificity (22). Cline et al point out that preferential binding of translation products by isolated envelopes is not due solely to receptors since the highly basic precursors are expected to interact ionically with the highly negatively charged outer envelope membrane (69); intact chloroplasts have a pI of 4 . 5 ( 1 60) . Neither precursor binding nor import occurs if isolated chloroplasts are pretreated with proteases , presumably because binding sites for transit se­ quences are destroyed ( 1 6 , 69). Although further studies are in order, the protease studies provide circumstantial evidence that the LHCP II and small subunit precursors have different receptors . We note from the data of Cline et al that import of LHCP II polypeptides is inhibited about 90% when chloroplasts are pretreated with 200 f..Lg/ml of thermolysin whereas a similar pretreatment reduces small subunit import by less than 50% (Figure 4 of 69). Since the components of the outer envelope membrane critical for LHCP II precursor import exhibit hypersensitivity to proteolysis, they may be distinct from the envelope determinants required by the small subunit. Even with untreated chloroplasts, import of the LHCP precursors is less efficient than that of the small subunit precursors, especially in suboptimal reconstitution systems (e . g . 1 1 ) . This could b e due to aggregation o f the LHCP precursors i n the i n vitro translation systems , a deficiency or low avidity of transport factors specific for the LHCP precursors relative to those for the small subunit, or differences in the n umber and/or affinity of receptors for the two classes of precursors . Cline et al calculated that isolated chloroplasts bind 70% more small subunit precursor than LHCP II precursor when nigericin is present to block import (69). Of

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

897

potential relevance to these differences, inclusion of 5 mM EDTA in transport incubation mixtures completely blocks LHCP II precursor uptake but has no effect on the level of small subunit accumulation (69, M. L. Mishkind, G. W. Schmidt, unpublished) . Although the EDTA effect is consistent also with distinct import pathways for soluble and membrane proteins, other explana­ tions are possible. EDTA may directly alter the conformation of the LHCP precursors since their mature forms avidly bind divalent cations. Also, selective effects on import may be misleading since EDTA might induce degradation of imported LHCP apoproteins in the isolated chloroplasts . The mitochondrial outer membrane appears to be similar to the chloroplast envelope with regard to the occurrence of receptors with affinity for specific precursors . Precursors to cytochrome c [equivalent to apocytochrome c because the precursor is not proteolytically processed after transport ( 1 6 1)] , the ADP/ ATP carrier ( 162), cytochrome hz, and citrate synthase ( 1 63) tightly bind to the outer mitochondrial membrane at sites from which transport may occur. Un­ labeled apocytochrome c competes for transport of trace amounts of labeled apocytochrome c but does not affect transport of precursors to the adenine nucleotide translocator or subunit 9 of the ATP synthetase ( 1 62) . As in chloro­ plasts, there is likely to be more than one class of receptors at the organelle surface.

ENERGETICS OF TRANSPORT Unlike precursor binding to envelope membranes, the transport of both thyla­ koid and stromal proteins into chloroplasts requires energy. It is certain that ATP, rather than an electrochemical gradient, supplies the energy for this process. Grossman et al demonstrated that light stimulates four to sevenfold the uptake of soluble and membrane proteins by isolated pea chloroplasts ( 1 64) , in contrast to earlier work ( 1 1 ) . Blocking photosynthetic noncyclic electron flow with DCMU (dichlorophenyldimethylurea) has no effect, however, on protein import in the light because dithiothreitol, present in transport mixtures, presum­ ably serves as a reductant for cyclic photophosphorylation under these con­ ditions. Complete inhibition of proton gradient formation and, consequently, ATP synthesis can be achieved with the protonophores salicylanilide XlII (Sal) and 3 ,5-di-tert-butyl-4-hydroxybenzylidenemalononitrile (SF6847) ( 1 64) or the ionophore nigericin (69, 1 65). In the presence of these inhibitors, import of proteins into illuminated chloroplasts decreases to levels observed in darkness. Inhibition of transport by these reagents substantiates the requirement for energy but does not distinguish its direct source. Fortunately, exogenous ATP supports import of the small subunit in darkness and in the presence of the inhibitors at levels comparable to that of uninhibited, illuminated chloroplasts (69, 1 64) . Although ATP restoration of LHCP II polypeptide import appears to

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

898

SCHMIDT & MISHKIND

be less complete than that of the small subunit, the differential effect on the two classes of precursors may not be significant: the degree oflight stimulation (e. g. 1 1) and ionophore inhibition (69) of import of precursors fluctuates consider­ ably. Moreover, endogenous ATP levels vary in chloroplasts from different plant species and in plants of different physiological or developmental states ( 1 64). The effectiveness of exogenously supplied ATP might vary also with the activity of the adenylate transporter of the envelope. This carrier exchanges stromal ADP for external ATP and presumably functions to supply the chloro­ plast with ATP in darkness ( 1 34) . Further understanding of the energetics of transport of proteins into chloro­ plasts requires resolution of the mode by which ATP-derived energy is coupled to the transfer of proteins across the envelope membranes. It is not known whether the ATP relevant to transport is employed outside the chloroplast, between the envelope membranes, or in an internal compartment. Analyses of transport under conditions that selectively deplete specific ATP pools, such as incubation of isolated chloroplast with hexokinase and glucose, a nonpenetrat­ ing ATP-consuming system (e.g. 1 66) , are j ust beginning (U. I. Flugge, personal communication). An outer envelope membrane protein kinase that phosphorylates in vitro the RuBPCase small subunit has been suggested to be involved ( 167), but there are no other clues about the energetics of the in­ termediate steps of transport in chloroplasts In other systems, electrochemical gradients are the energy source for trans­ port of proteins through lipid bilayers . This is the case for translocation of proteins through the mitochondrial inner membrane ( 15, 1 68-1 70). These structures possess the electron transport chain, ATP synthetase complexes, and a substantial proton motive force. The topologically analogous inner envelope membranes of chloroplasts , especially those of the bundle sheath cells of maize ( 1 7 1 ) , do not support a large electrochemical gradient. In E. coli, electrochem­ ical gradients also have been considered to be the energy source for protein export, based on the accumulation of precursors in cells treated with ionophores ( 172-174) . It has been proposed that in bacteria this energy is utilized in a manner analogous to electrophoresis as the membrane polarity is positive outside ( 1 75, 1 76). Since mitochondrial protein import would be toward the negative side of the membrane, Schatz & Butow suggested that the electro­ chemical gradient in these organelles may serve to "labilize" lipid bilayers to induce formation of inverted micelles that conduct the transmembrane move­ ment of proteins ( 1 77). For certain mitochondrial proteins, import as well as binding occurs in the absence of energy input ( 15, 168, 1 78) . These polypep­ tides, however, are limited to the outer envelope membrane, and in the case of apocytochrome c, to the intermembrane space. Energy-independent transport has not been demonstrated for any chloroplast polypeptide. In contrast to the above, secretory proteins in eukaryotes are vectorially .

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

899

discharged into the lumen of the endoplasmic reticulum by the ATP-derived force of protein chain elongation; consistent with the absence of sources for formation of a large membrane potential in these membranes, uncouplers reportedly do not impair cotranslational transport in reconstituted microsome preparations ( 1 79) . Although the necessity for ATP in posttranslational trans­ port has been thought to be unique to chloroplasts , there are recent indications that it is also the energy source for the posttranslational transport of proteins into plasma membrane vesicles from E. coli ( 1 66). While it is possible that bacteria and chloroplasts have similar energetic means for transporting proteins through membranes, electrochemical gradients in bacteria are also of apparent im­ portance, enabling precursor maturation to occur with optimal efficiency ( 1 66).

TRANSIT PEPTIDASES All cytoplasmically synthesized chloroplast proteins so far examined experi­ ence proteolytic maturation during or shortly after transport into the organelle. Toward elucidating where and how precursor processing occurs, current objec­ tives are to define the number of maturation steps, to identify the amino acid determinants in transit sequences that constitute processing sites, and to characterize the number, specificity, and location of the maturation enzymes. Of the proteases that remove transit sequences from chloroplast precursors, only enzymes (transit peptidases) from pea ( 1 80) and Chla mydo monas ( 10, 63, 1 8 1 ) have been partially purified and characterized. The transit peptidase from Chlamydomonas that correctly and completely matures the small subunit pre­ c ursor is a soluble protease that is inhibited by sulfhydryl reagents such as N-ethyl maleimide and mercuric chloride. This enzyme is a nuclear gene product ( 1 6, 1 82) and is enriched in high-molecular-weight fractions of cell extracts ( 1 8 1) . It processes the algal precursor in vitro ( 1 8 1) and in vivo ( 1 82) in the absence of large subunits of RuBPCase, but does not recognize the small s ubunit precursor of pea ( 1 8 1 ). An analogous protease from pea chloroplasts, although originally thought to be membrane associated ( 1 2), is a soluble enzyme ( 1 83) of about 1 80 kilodaltons as estimated by gel filtration ( 1 80). This protease has been purified 348-fold but, as yet, is not sufficiently homogeneous to enable identification of its polypeptide(s) ( 1 80). However, we note that enrichment of its activity by anion exchange chromatography closely coincides with the occurrence of a high-molecular-mass polypeptide (> 100 kilodaltons) in silver-stained polyacrylamide gels of the column fractions (Figure 3 of 1 80) . In contrast to the Chlamydomonas transit peptidase, the pea enzyme is not inhibited by sulfhydryl-modifying reagents ( 1 80). Similar to the matrix pepti­ dase that processes mitochondrial precursors ( 1 84-1 86), the pea enzyme is a metallo-protease with sensitivity to 1 , 1O-o-phenanthroline and EDTA ( 1 80) . The pea transit peptidase preparation also has been reported to mature the

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

900

SCHMIDT & MISHKIND

precursors of wheat and barley plastocyanin and ferredoxin-NADP-oxidore­ ductase (180). As indicated by the underlined segments near the carboxy termini of the transit sequences displayed in Figure I, the precursors of the soluble and some of the membrane proteins of vascular plants possess a highly similar stretch of three amino acids beginning five residues before their maturation sites. If this amino acid series (small-basic-hydropathic residues) constitutes the transit peptidase recognition site, the enzyme hydrolyzes peptide bonds two residues toward the carboxy terminus from where it binds. Smeekens et al also noted that the sequence three residues from the maturation site of pre-ferredoxin (glycine­ arginine-valine) is homologous to the processing region of the small subunit precursors of vascular plants (75) . Maturation of pre-ferredoxin could actually occur two residues toward the carboxy terminus from the hydrophobic amino acid, as postulated for small subunit precursors, but the resulting NHrterminal methionine is subsequently removed, a common occurrence ( 1 87). It is signifi­ cant that the small subunit precursor of Chlamydom onas lacks this sequence and is processed only to an intermediate form upon transport into vascular plant chloroplasts (65). In their analysis of the precursor of plastocyanin, Smeekens et al noted that there is no homology at the processing site with that of the small subunit precursors, indicating that these may have different maturation path­ ways ( 1 33). However, the transit peptidase preparation obtained by Robinson & Ellis is active with pre-plastocyanin as well as precursors to the vascular plant small subunits and ferredoxin-NADP-oxidoreductase (180). More highly puri­ fied transit peptidase preparations are required before strong conclusions can be drawn about their substrates and sequence specificity. The sequence similarities adjacent to the processing sites in the LHCP II polypeptide and small subunit precursors of vascular plants indicate that the transit peptidases for these two classes of proteins could be identical. Un­ fortunately, the activity of a partially purified transit peptidase from pea chloroplasts with precursors of the LHCP II polypeptides has not been de­ scribed (180). Crude preparations of a soluble Chlamydom onas transit pepti­ dase, although highly active with small subunit precursors of the alga, will convert only 60% of the LHCP II precursors to their mature size even when high concentrations of cell extracts are employed (188). Although the two classes of precursors are distinguishable by their susceptibility to processing, it is not necessarily due to distinct transit peptidases . As Marks et al ( 1 88) suggest, the soluble forms of the LHCP II precursors may be poor processing substrates in vitro because of their conformation. Upon becoming membrane-associated in vivo, efficient maturation of the LHCP II precursors may be achieved by the same enzyme responsible for small subunit maturation. Nascent chains of the precursors to the LHCP polypeptides appear to be better substrates for transit peptidases than full-length polypeptides, supporting

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

901

the notion that the tertiary structure of the LHCP II precursors affects their ability to be processed. This is seen in the work of Pfisterer et aI, who employed a wheat g�rm readout system to translate polysomes containing mRNA for the R uBPCase small subunit and LHCP polypeptides ( 1 59) . Apparently, transit peptidase present in polysome preparations generates discrete , im­ munoprecipitable polypeptides corresponding in size to the mature forms of both protein classes ( 1 59). Perhaps the conformation of incompletely syn­ thesized precursors of the LHCP polypeptides resembles that which they assume when undergoing envelope membrane translocation in vivo . Associa­ tion of transit peptidase activity with polysomes is precedented in the studies of Dobberstein et al ( 10), Roy et al ( 1 89), and Gray & Kekwick ( 1 90). Verifica­ tion that polysome readout products are correctly matured, however, requires analysis of the polypeptides by amino acid sequencing or, at least, two­ dimensional electrophoresis. There also have been incidental reports of the recovery of mature chloroplast polypeptides from mRNA translation products generated in the wheat· germ (58 , 1 59) and reticulocyte lysate (58) systems. There are no satisfactory explanations for these observations, especially in the case of the reticulocyte lysate translations. Recent studies indicate that at least some imported polypeptides experience multiple proteolytic cleavages during their maturation in chloroplasts. In a kinetic analysis of: processing by the partially purified pea transit peptidase, transient appearance of an intermediate form ( 1 8 kilodaltons) of the pea small s ubunit was detected ( 1 9 1 ) . Toward assessing whether the intermediate arises by proteolysis within the transit sequence, the precursor was treated with iodoacetate. This carboxymethylates the cysteine residue located at the car­ boxyl end of the transit sequence, introducing charge and changing side chain length at the j unction with the mature polypeptide. Upon reaction of the modified precursor with transit peptidase, the 1 8-kilodalton intermediate was the only proteolytic product (191). The carboxymethylated precursor retains competence for import into isolated chloroplasts but is recovered as a series of polypeptides ranging in molecular mass from about 1 2 to 1 8 kilodaltons. These results indicate that the pea small subunit precursor contains two processing sites, both of which are recognized by the partially purified processing enzyme. If carboxymethylated, the site for generation of the matlJre polypeptide is not recognized. To explain the apparent degradation of the 1 8-kilodalton in­ termediate upon import of the modified precursor in chloroplasts, it was proposed that the intermediate fails to assemble into RuBPCase holoenzyme and consequently becomes a substrate for other proteases ( 1 9 1 ) . The amino­ terminal sequence of the l 8-kilodalton derivative has not been determined. Consequently, it is not known if the intermediate arises by cleavage within the transit sequence or, alternatively, at the carboxy terminus of the mature protein. As discussed for maturation of the plastocyanin precursor, whether a single

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

902

SCHMIDT & MISHKIND

protease mediates cleavage at both sites will be resolved only upon purification of the enzyme to homogeneity. Two-step processing of the RuBPCase small subunit is also evident from heterologous in vitro transport studies performed with Chlamydom onas pre­ cursors and chloroplasts isolated from spinach or pea (65). The algal small subunit precursor is transported into these organelles but is recovered as a form intermediate in size between the mature polypeptide and its precursor. Although it is present in the stroma, the intermediate is not assembled with vascular plant large subunits to form RuBPCase holoenzyme. The inability of vascular plant transit peptidases to completely mature the algal precursor probably are due to transit sequence differences described above . The Chlamydomonas-specific maturation site is retained, however, since the algal transit peptidase processes the intermediate to its mature form. These data demonstrate that an enzyme in vascular plant chloroplasts specifically cleaves the algal small subunit within its transit sequence. From amino acid sequence analysis , the intermediate results from processing within a segment (Domain II) that is represented in the transit sequences of vascular plant small subunit precursor sequences and is indicated by the

arrow

in Figure 1 . This region is

characterized by a hydrophobic amino acid (leucine) followed by a positively charged residue (lysine or arginine) and preceded by a small amino acid (glycine or alanine). Invariably, proline and valine are located at positions -4 and -6, respectively, relative to leucine. A hydrophobic amino acid (phenylala­ nine or methionine) occurs at position -3. The structural conservation in this segment in all small subunit precursors so far characterized suggests that it mediates critical steps in transport such as the binding of precursors to envelope receptors or proteolytic maturation (65). Cleavage within this region may allow the release of receptor-bound precursors to the stroma. Robinson & Ellis have concluded that both processing steps of the pea small subunit precursor are catalyzed by the high-molecular-weight soluble metallo­ endoprotease ( 1 9 1 ) . Whether this is the enzyme that generates the algal small subunit intermediate is uncertain but seems unlikely. Since pea chloroplast lysates process small subunit precursors from pea ( 1 80, 1 83), but not from Chlamydomonas, (M. L. Mishkind, G . W. Schmidt, unpublished), it appears that there must be two different enzymes. Unlike the chelator-sensitive en­ zyme, the activity in pea chloroplasts that cleaves the algal precursor appears to be quite labile. A similarly labile activity has been detected in Chlamydom onas: chloroplasts isolated from the alga will transport and partially mature pea small subunit precursors whereas crude cell lysates exhibit no processing activity with the vascular plant precursor (M. L. Mishkind, G. W. Schmidt, un­ published) . Evidence that processing intermediates occur in vivo has been obtained for the 1 5 . 5-kilodalton chloroplast ribosomal protein, L- 1 8 , from Chlam ydom onas

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

903

( 192) . Whereas the primary translation product synthesized in vitro is an 1 8 .5-kilodalton polypeptide, an intennediate of 1 7 kiIodaltons is detected by immunoprecipitating extracts of pulse-labeled cells . The intennediate's half­ life nonnally is less than 5 min but this can be extended when chloroplast protein synthesis is inhibited; these conditions lead to degradation of the intennediate rather than processing to its mature size ( 1 92). The L- 1 8 intennediate can also be detected during in vitro processing of its precursor ( 1 92) . Like the pea small subunit intennediate, the 17-kilodalton polypeptide appears transiently during incubation of the primary translation product with algal extracts (58 , 1 92) . It was suggested that, in vivo, the L- 1 8 precursor is rapidly converted to the 1 7-kilodalton intermediate by a processing event closely coupled to transport. Because it is degraded rather than processed to its mature form when cells are treated with chloroplast protein synthesis inhibitors, further maturation of the intermediate appears to require a product of chloroplast protein synthesis. Assembly of L- 1 8 into a ribosome is not a prerequisite for the second processing step since complete maturation occurs in vitro with postribosomal supernatant fractions (58 , 1 92) . Two-step processing has not been observed for other imported Chlamydomonas chloroplast ribosom­ al proteins that are products of cytoplasmic protein synthesis ( 1 92). Several imported mitochondrial proteins are also processed to their mature fonns in two steps. The primary processing step of these polypeptides, all of which are located in the intermembrane space or within the inner membrane, is performed by the soluble matrix metallo-endoprotease ( 193-1 96) .

DIRECTIONS AND PROSPECTS Translocation to Inner Compartments When stromal proteins reach the inner envelope membrane, their transit se­ quences probably are removed by proteolysis, an event that may facilitate release from receptors. In the case of the RuBPCase small subunit, the L- 1 8 ribosomal protein, and perhaps other stromal proteins, this step may occur via envelope-localized proteases to generate maturation intermediates . Sub­ sequently, processing of intermediates is completed by soluble transit pepti­ dases in the stroma. Translocation of thylakoid membrane proteins from the chloroplast envelope to the photosynthetic membrane undoubtedly is more complex than the steps for localization of the stromal proteins. At present, there are few clues about this process, partly because the site where maturation of precursors of thylakoid proteins occurs and whether it involves more than one processing step have not been detennined. Although it is plausible that preLHCPs undergo transit through the stroma as soluble precursors or processing intermediates , there is much merit to proposals that vesicles derived from the inner envelope membrane function as carriers for

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

904

SCHMIDT

&

MISHKIND

these polypeptides. Morphological evidence for envelope-derived vesicles has been reviewed by Douce et al ( 1 36) . Carde et al ( 1 50) have provided convincing ultrastructural evidence for the evagination of the inner envelope membrane in spinach chloroplasts in vivo and in vitro. They noted that plastids in young etiolated seedlings are relatively devoid of inner envelope evaginations. However, during light-dependent chloroplast development, inner envelope evaginations become abundant and often appear as membrane tubules. Vesicles also have been visualized in young chloroplasts subjected to centrifugation; due to their low density, vesicles float to the end of the intact organelles opposite from pelleted thylakoids ( 1 97). Nonthylakoid membranes , when very abun­ dant, form a "peripheral reticulum" of anastomizing tubules which is usually a characteristic of photosynthetic cells of plants with the C4 dicarboxylic acid mode of carbon dioxide fixation . Laetsch has noted that these membranes are different in composition from the envelope membranes and thylakoids on the basis that the peripheral reticulum is preferentially destroyed when per­ manganate is used as a fixative ( 1 98). Douce et al ( 1 36) envisage that transit peptides interact with the external surface of the inner membrane to induce the formation of vesicles for transloca­ tion to thylakoids of products of envelope biosynthetic pathways. Since the hydrophilic precursor extensions may be removed as soon as they become exposed in the stroma, we believe they cannot be very effective in causing membrane evagination. We propose that vesicles are formed, at least in part, from interactions among the mature portions of imported proteins that are integral components of thylakoids . The LHCP II polypeptides are suitable candidates partly because they are major products of cytoplasmic protein synthesis. When LHCP precursors reach the inner envelope membrane they are likely to be oriented, similar to their mature forms ( 1 0 1 ) , with the NH2-terminal transit sequences plus 2 kilodaltons of the mature proteins exposed in the stroma. Upon maturation, the LHCP apoproteins may then be able to interact with each other to form higher order complexes like those that they form in thylakoids (95-97). They also could associate with some of the components synthesized in the inner membrane that intimately associate with the mature pigment-protein complexes. These include xanthophylls and lipids which have been determined to be specific components of LHCP II (94, 9 8 , 1 99) . Other envelope products , including l3-carotene, the remaining thylakoid-destined lipids, plastoquinone, and geranylgeranyl precursors for the phytol chain of chlorophyll ( 1 36) , could subsequently become sequestered in the forming vesicles through less specific lipophilic interactions. Thus, when the membrane evaginates , several constituents of photosynthetic membranes would undergo bulk transport as proposed by Douce et al ( 1 36) . This scheme presents a solution to the problem of the absence of stromal proteins for lipid transfer from the envelope to thylakoids (200) . In addition, it is possible that proteins

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

905

destined for the thylakoid lumen are sequestered inside these vesicles and thereby would not undergo transport through the lipid bilayers of either the thylakoid or inner envelope membranes. Formation of vesicles from membranes is topologically difficult and must involve localized disruption of lipid bilayers . Although this might be accom­ plished locally by hydrophobic portions of proteins, "nonbilayer-forming" lipids , possessing "intrinsic curvature" properties, have been proposed to he essential for evagination/invagination processes (20 1 , 202) . The major non­ bilayer forming lipid in both the chloroplast envelope membrane and in the thylakoid is monogalactosyldiacylglycerol (MGDG) ( 1 36) . As a lipid of this class, novel procedures must be used for its reconstitution into bilayers (203). MGDG appears to associate rather specifically with LHCP II complexes or PS II reaction centers or both . Sonication of Triton X- I OO-solubilized thylakoids with MGDG, but not other thylakoid lipids , leads to restoration of exciton transfer from LHCP II to PS II, as measured spectrofluorometrically and by formation of particles that could be sedimented at 1 7 ,000 x g (98). Xanthophylls also may be necessary components for vesicle evagination since these are found in the mature LHCP II complexes and are important effectors of the conformation of the LHCP polypeptides in the thylakoid membrane (F. G. Plumley, G. W. Schmidt, in preparation). Consistent with this hypothesis, inhibitors of carotenogenesis and mutants blocked in their synthesis exhibit severe defects in thylakoid formation (204, 205). Moreover, a mutant of maize which is blocked in transcription of genes encoding the LHCP polypeptides is nearly nonpigmented and lacks thylakoids (205) . It was sug­ gested that products of the carotenogenic pathway are directly involved in the regulation of expression of the LHCP genes (205), a hypothesis that may be correct. However, we feel that the studies of the maize mutants also indicate that pigment accumulation in chloroplasts requires the continual synthesis of the LHCP polypeptides. Although the above model emphasizes a role of LHCP II polypeptides in the formation of envelope vesicles , other imported membrane proteins might perform this role as well. NADPH-protochlorophyllide reductase , which also can form polymeric complexes as the major integral membrane component in prothylakoids of etioplasts (82-85), is possibly another vesicle-forming pro­ tein. Transit Sequences Transit sequences are potent effectors for the biogenesis of chloroplast proteins and appear to be all that are necessary for import of soluble proteins into the stromal compartment. Whether transit sequences of proteins of the envelope membranes, the thylakoid membrane, and the thylakoid lumen serve auton­ omously in the suborganellar localization of these classes remains to be

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

906

SCHMIDT & MISHKIND

examined through genetic engineering techniques. If integral membrane pro­ teins of thylakoids, such as the LHCP polypeptides , are translocated as a function of their mature sequences as well as their transit sequences, identifica­ tion of the functional determinants for this process may be difficult. However, determining where fusion proteins containing the transit sequences of pre­ cursors to the LHCP polypeptides or plastocyanin become localized is a worthwhile endeavor. Comparative sequence analysis of different classes of precursors, especially when evolutionarily divergent species are employed that retain the capacity for heterologous transport, can aid also in the further identification of functional determinants of transit sequences. In this manner, more information can be obtained about the specificity of envelope receptors and the enzymes that are involved in precursor maturation. Finally, site­ directed mutagenesis of cloned precursors will be of enormous use in pinpoint­ ing the necessary amino acid sequences for chloroplast import and maturation. Import Factors Isolated intact chloroplasts and in vitro synthesized precursor proteins provide uptake. Since precursors syn­ thesized in nonplant cellfree translation systems such as E. coli (67) and reticulocyte lysates (206) are capable of transport, the translation system is unlikely to provide plant-specific components other than the precursor polypeptides . It is possible, however, that cytoplasmic factors associated with isolated chloroplasts are part of the transport machinery. In yeast, a soluble, trypsin-sensitive factor of about 40 kilodaltons has been shown to be required for the import of the �-subunit of the mitochondrial ATP synthetase (207) . Although large complexes of precursors and cytosolic factors appear to be important in the import of proteins into mitochondria, the occurrence of chloroplast precursors in high-molecular-weight complexes has not been observed. The small subunit precursor, for example, is not incorporated into a rapidly sedimenting particle (65). Evidence that the small subunit precursor is not sequestered completely in a high-molecular-weight complex is apparent also from the ability of the IgG fraction of affinity-purified antibodies to selectively and almost completely block import of the small subunit precursor into chloroplasts (M. L. Mishkind, G . W. Schmidt, unpublished) . Precursors of the LHCP polypeptides of Chlamydomonas do form higher-molecular­ weight complexes in cell-free translation systems (Figure 5C of 65) but these complexes are much smaller than those of precursors of animal mitochondrial proteins. Pre-ornithine transcarbamylase (40 kilodaltons) of rat liver associates with a ribonucleoprotein in reticulocyte lysates to form a 400-kilodalton com­ plex that is essential for organelle import (208). Aggregates (or complexes with cytosolic factors) have been noted also for the 45-kilodalton precursors of rat mitochondrial fumarase which elutes as a 300-400-kilodalton aggregate of 6-8 factors sufficient for supporting pol ypeptide

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT

907

s ubunits on gel filtration columns (5 1 ) and the precursor of subunit 9 of the mitochondrial ATP synthetase in Neurosp ora (209). The Neurosp ora ADP/ ATP carrier protein, which is not made as a larger precursor, has a molecular mass of 3 1 kilodaltons but elutes in gel filtration columns as two bands, one of 1 20 kilodaltons and one of 500 kilodaltons or higher (2 10, but also see 2 1 1 ) . Except for the data showing that LHCP precursors synthesized i n the wheat germ system sediment partially into sucrose gradients (65), possibly due to their self-associating properties, there is no evidence for a role of precursor com­ plexes in chloroplast transport. However, cytosolic complexes and factors warrant more detailed study. In the case of mitochondria, porphyrins have been documented to participate in the import of some cytoplasmically synthesized proteins. Import of the precursor of 8-aminolevulinate synthase, the first enzyme for porphyrin synthe­ sis, is blocked in vivo and in vitro by the presence of hemin (2 1 2 , 2 1 3) . The precursor of cytochrome Cl is imported into the yeast mitochondrial matrix where it is converted to a maturation intermediate, then is translocated back to the outer face of the inner mitochondrial membrane. If heme is present, a second maturation step takes place resulting in the functional association of cytochrome CI with other components of the respiratory chain in the in­ termembrane space (194). The evidence for this pathway was derived from studies of a heme-deficient mutant of yeast which accumulates the cytochrome CI intermediate. So far, whether chlorophyll a and/or its precursors participate in the import and maturation of LHCP apoproteins is not resolved. However, Johanning­ meier & Howell (214) have suggested that chlorophyll precursors may partici­ pate in this process, perhaps through the involvement of magnesium pro­ toporphyrin methyl ester which is formed by chloroplast envelope enzymes (2 1 5 , 2 16). Their proposal derives from studies of accumulation of mRNAs for the LHCP II polypeptides in Chlamydomonas using chlorophyll biosynthesis inhibitors and mutants that accumulate chlorophyll precursors. Under con­ ditions where protoporphyrin IX, magnesium protoporphyrin methyl ester, or protochlorophyllide accumulate, LHCP mRNA levels are low relative to nor­ mal cells, indicating porphyrins somehow play a role in regulation of nuclear gene expression in photosynthetic cells. From in vivo pulse labeling studies on chlorophyll biosynthesis mutants of Chlamydomonas, we find that the mRNAs for the LHCP II polypeptides, although at low levels, are translated and mature LHCP apoproteins transiently accumulate in membranes (G. W. Schmidt, unpublished). Since thylakoids are virtually absent in these strains, we suppose they are blocked in transport at the level of the envelope and work is in progress toward verifying this possibility. If this is the case, it is possible that LHCP precursors or their mature forms must interact with the chlorophyll precursors as an early event in the import process. -

908

SCHMIDT

&

MISHKIND

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

Transport Mutants Further resolution of the pathway and components for the transport of proteins into chloroplasts could be gained from genetic approaches. As examples, yeast m utants blocked in the import of mitochondrial proteins have been isolated by screening temperature-sensitive mutants for the accumulation of precursors of the f3-subunit of the F1-ATPase (2 17). Two nonallelic nuclear mutants were characterized and found to accumulate precursor in the cytosol; other proteins also are imported at reduced rates in these mutants . The primary lesions for these phenotypes have not been defined, but molecular analyses of these mutants may lead to identification of important components for mitochondrial protein transport. Mutants also have provided invaluable insights into the complexities of protein export in bacteria (9) . Although genetic analyses provide an alternative and useful approach to the study of protein uptake by chloroplasts, the development of selection and screening protocols for mutants must be carefully considered: defects in the import machinery are likely to be lethal or lead to slow growth. However, conditional mutants should be obtain­ able and these will use genetic approaches to dissect the steps involved in the transport of proteins to the six chloroplast compartments.

ACKNOWLEDGMENTS We are grateful to F. Gerald Plumley for his many contributions to this review. Preparation of this manuscript and research cited from our laboratory was s upported from grants from the National Science Foundation and the US Department of Energy . Literature Cited I . Milstein, C . , Brownlee, G. G . , Harrison, 2. 3. 4.

5.

6. 7. 8. 9.

T. M . , Mathews, M. B. 1 972. Nature 239: 1 l7- 1 20 Blobel, G . , Dobberstein, B. 1 975 . J. Cell Bioi. 67:835-5 1 Blobel, G . , Dobberstein, 1;3. 1 975 . J . Cell Bioi. 67:852-62 Inouye , S . , Wang, S . , Sekizawa, J . , Halegoua, S . , Inouye, l . J . 1977. P�oc. Natl. Acad. Sci. USA 74: 1 004-8 BUIT, B . , BUIT, F. A . , Rubenstein, I . , Simon, M . N . 1 978. Proc. Natl. Acad. Sci. USA 75:696--700 B lobel, G. 1 980. Proc. Natl. Acad. Sci. USA 77: 1 496--500 Si 1havy , T. J . , Benson, S. A . , Ernr, S. D. 1 983. Microbiol. Rev. 47:3 1 3-44 Walter, P. , Gilmore, R . , Blobel, G . 1984. Cell 38:5-8 Benson, S. A . , Hall, M. N . , Silhavy , T. J. 1 985 . Ann. Rev. Biochem. 54: 1 0 1-34

10. Dobberstein, B . , Blobel, G . , Chua, N.-H. 1 977. Proc. Natl. Acad. Sci. USA 74: 1 082-85 I I . Chua, N .• H . , Schmidt, G. W . 1 978. Pmc. Natl. Acad. Sci. USA 75:6 1 1 ()"" 1 4 1 2 . Highfield, P . E., Ellis, R . J . 1978. Na­ ture 27 1 :42()""24 1 3. Kindl, H. 1 982. Int. Rev. Cytol. 80: 193229 1 4 . Lord, J . M . , Roberts , L. M. 1983. Int. Rev. Cytol. Suppl. 1 5 : 1 1 5-56 1 5. Hay, R . , Bohni, P. , Gasser, S. 1 984. Biochim. Biophys. Acta 779:65-87 1 6. Chua, N . ·H. , Schmidt, G . W. 1 979. J. Cell Bio i. 8 1 :461-83 1 7 . Oaks, A . , Hirel, B. 1985. Ann. Rev. Plant Physio l. 36:345-65 1 8 . Roughan, G . , Slack, R. 1 984. Trends Biochem. Sci. 9:383-86 1 9. Miflin, B . J . , Lea, P. J. 1977. Ann. Rev. Plant Physiol. 28:299--329

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

CHLOROPLAST PROTEIN TRANSPORT 20. Camara, B. 1 984. Plant Physiol. 74: 1 12-16 2 1 . Castelfranco, P. A . , Beale, S . I. 1983. Ann. Rev. Plant Physiol. 34:241-78 22. Ellis, R. J. 1 983 . Suhcell. Biochem. 9:237-61 23. Whitfeld, P. R. , Bottomley, W. 1983. Ann. Rev. Plant Physiol. 34:279-3 1 0 24. Watson, J. c . , Surzycki, S. J. 1982. Proc. Natl. Acad. Sci. USA 79: 2264-67 25. Cashmore, A. R. 1 979. Cell 1 7:383-88 26. Bedbrook, J. R . , Smith, S. M . , Ellis, R. 1. 1 980. Nature 287:692-97 27. Coruzzi, G . , Broglie, R . , Cashmore , A. R . , Chua, N.-H. 1 98 3 . 1. Bioi. Chem . 258: 1399-402 28. Berry-Lowe, S . L . , McKnight, T. D . , Shah, D. M . , Meagher, R. M . 1982. 1. Mol. Appl. Genet. 1 :482-98 29. Stiekema , W. J . , Wimpee , C. F. , Tobin , E . M . 1983 . NucleicAcids Res. 1 1 :8051-

61

30. Dunsmuir, P. , Smith, S . , Bedbrook, J. 1 983. Nucleic Acids Res. 1 1 :41 77-83 3 1 . Pinck, L. , Fleck, J . , Pinck, M . , Hadi­ dane, R . , Hirth, L. 1983. FEBS Lett. 154: 1 45-48 32. Smith, S. M . , Bedbrook , J . , Speirs , J. 1983 . Nucleic Acids Res. 1 1 :27 19-34 33. Mazur, B. J . , Chui, C.-F. 1 985. Nucleic Acids Res. 1 3 : 2373-86 34. Broglie. R . , Coruzzi, G . , Lamppa, G . , Keith, B . , Chua, N . -H . 1983. Bioi Technology 1 : 55-61 35. Shinozaki, K . , Sugiura, M . 1983. Nucle­ ic Acids Res. 1 1 :6957-64 36. Reith, M. E . , Cattolico, R. A. 1985. Biochemistry 24:2556-6 1 37. Grossman, A. R . , Barlett, S. G . , Schmidt, G . W . , Mullet, J. E . , Chua, N . -H . 1982. 1. Bioi. Chem . 257: 1558-63 3 8 . Hague, D. R . , Uhler, M . , Collins, P. D. 1 983. Nucleic Acids Res. II :4853-65 39. Aoyagi, K . , Bassham, J. A. 1984. Plant Physiol. 76:278-80 40. Perlman, D . , Raney, P. , Halvorson, H. O . 1984. Mol. Cell. Bioi. 4: 1 682-88 4 1 . Hopper, A. K. , Furukawa, A. H . , Pham , H. D . , Martin, N. C. 1982. Cell 28:54350 42. Murphy, D . J . , Walker, D . A. 198 1 . FEBS Lett. 134: 1 63-66 43. Kruger, I . , Schnarrenberger, C. 1983. Eur. J. Biochem . 1 36 : 1 0 1-6 44. Lebherz, H. G . , Leadbetter, M . M . , Bradshaw, R . A. 1984. 1 . Bioi. Chem . 259: 1 0 1 1 - 1 7 45. Simcox, P. D., Dennis, D . T. 1978. Plant Physiol. 62:287-90 46. Muta, S. 1982. FEBS Lett. 147: 1 6 1-64 47. Biro, R . L . , Daye, B. S . , Serlin, M. E . , Terry, N . , Datta, N . , e t at. 1984. Plant Physiol. 75:382-86

909

48. Dieter, P . , Marme, D. 1 984. 1. Bioi. Chem. 259:1 84-89 49. Simon, P. , Bonzon, M . , Greppin , H . , Marme, D . 1984. FEBS Lett. 1 67:33238 50. Roberts, D. M . , Zielinski, R. E . , Schleicher, M. , Watterson , D. M. 1983. 1. Cell. Bioi. 97: 1 644-47 5 1 . Ono, H . , Yoshimura, N . , Sato, M . , Tuboi, S . 1985. 1. Bioi. Chem . 260: 3402-7 52. Pichersky, E . , Gottlieb, L. D . , Higgins, R. C. 1984. Mol. Gen. Genet. 1 93 : 1 5861 5 3 . Kurzok, H.-G . , Feierabend, J. 1 984. Biochim. Biophys . ACla 788:222-33 54. Cerff, R . , Kloppstech, K. 1982. Proc. Natl. Acad. Sci. USA 79:7624-28 55. Sabatini, D. D . , Kreibich, G . , Morimo­ to, T. , Adesnik, M. 1982. 1. Cell. Bioi. 92: 1-22 56. Neupert, W . , Schatz, G. 1 98 1 . Trends Biochem. Sci. 6: 1-4 57. Small, I. S . , Gray, J. C. 1984. Eur. 1. Biochem . 145:29 1-97 58. Schmidt, R . J . , Myers, A. M . , Gillham, N. W . , Boynton, J . E. 1984 . 1. Cell Biol. 98:201 1 - 1 8 59. Van der Mark, F . , van den Briel, W . , Huisman, H. G . 1 983. Biochem. J. 2 1 4:943-50 60. Shure, M . , Wessler, S . , Federoff, N. 1983. Cell 35 :225-33 6 1 . Ohlrogge , J. B . , Kuo, T. -M . 1984. What's New in Plant Physiol. 1 5:4 1-44 6 1 a. Vierling, E . , Mishkind, M . L., Schmidt, G. W . , Key, J. L. 1986. Proc. Natl. Acad. Sci. USA . 83:361-65 62. Kloppstech, K . , Meyer, G . , Schuster, G . , Ohad, I. 1985. EMBO 1. 4: 1 90 1-9 63. Schmidt, G. W . , Devillers-Thiery , A . , Desruisseaux, H . , Blobel, G . , Chua, N . ­ H. 1979. 1. Cell. Bioi. 83:61 5-22 64. Coruzzi , G . , Broglie, R . , Lamppa, G . , Chua, N.-H. 1983. I n Structure and Function ofPlant Genomes, ed. O . Cifer­ ri, L. Dure III, pp. 47-59. New York: Plenum 65. Mishkind, M. L . , Wessler, S. R . , Schmidt, G. W . 1985. 1 . Cell BioI. 1 00: 226-34 66. Broglie, R . , Coruzzi , G . , Fraley, R. T . , Rogers, S . G . , Horsch, R. B . , et aI . 1984. Science 224:838-43 67. Van den Broeck, G . , Timko, M. P . , Kausch, A. P. , Cashmore , A . R . , Van Montagu, M . , Herrera-Estrella, L. 1 985. Nature 3 1 3:358-63 68. Schreier, P. H . , Seftor, E. A . , Schell, J . , Bohnert, H . J. 1985. EMBO 1 . 4:25-32 69. Cline, K . , Werner-Washburne , M . , Lub­ ben, T. H. , Keegstra, K. 1985. 1. Bioi. Chem . 260:3691-96

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

910

SCHMIDT & MISHKIND

70. Horwick, A. L . , Fcnton, W. A . , Fir­ gaira, F. A. , Fox, J. E . , Kolansky, D . , et al. 1 985. J. Cell. Bioi. 100: 1 5 1 5-21 7 1. Von Heijne, G . 1 984. J. M ol. Bioi. 1 73:243-5 1 72. BrandIs , J. F . , Halvorson, H. R . , Bren­ nan, M. 1975. Biochemistry 14:4953--63 73. Kim, P. S . , Baldwin, R. L. 1 982. Ann. Rev. Biochem. 5 1 :459-89 74. Davies, D. R. , Metzger, H. 1983 . Ann. Rev. lmmunol. 1 : 87-177 75 . Smeekens, S . , Binsbergen, V . , Weis­ beek, P. 1985. Nucleic Acids Res. 1 3: 3 1 79-94 76. Pongor, S . , Szalay, A. A. 1985 . Proc. Natl. Acad. Sci. USA 82:366--70 77. Garnier, F . , Gaye, P. , Mercier, F.-C . , Robson, B . 1980. Biochimie 6:231-39 7 8 . Emr, S . D . , Silhavy, T. 1 983. Proc. Natl. Acad. Sci. USA 80:4599--603 7 9 . Briggs, M. S . , Gierasch, L. M. 1984. Biochemistry 23:3 1 1 1 - 1 3 80. Tobin, E. M . , Silverthorne , J . 1 985. Ann. Rev. Plant Physiol. 36:569-93 8 1 . Apel, K. 1 98 1 . Eur. J. Biochem. 1 20: 89-93 82. Apel, K . , Santel, H .-J . , Redlinger, T. E., Falk, H. 1 980. Eur. J. Biochem. 1 1 1 :25 1-58 83. Meyer, G . , Bliedung, H . , Kloppstech, K. 1 983. Plant Cell Rep . 2:26-29 84. Hauser, I . , Dehesh, K . , Apel, K . 1 984. Arch. Biochem. Biophys. 228:577-86 85. Kay, S . A . , Griffith, W. T. 1 983. Plant Physiol. 72:229-36 86. Meyer, G . , Kloppstech, K. 1 984. Eur. J. Biochem. 1 38:201-7 87. Chua, N . -H . , Gillham, N . W. 1977. J. Cell Bioi. 74:441-52 88. De1epe1aire , P. , Chua, N.-H. 1 98 1 . J . Bioi. Chem. 256:9300-7 89. Schmidt, G. W . , Bartlett, S. G . , Gross­ man, A . R . , Cashmore, A. R . , Chua, N . -H . 1981. J. Cell Bi oi . 9 1 :468-78 90. Suss, K.-H. 1 983. Photobiochem. P hotobi op hys. 5 : 3 1 7-24 9 1 . Chua, N.-H . , Blomberg , F. 1 979. J. Bioi. C he m. 254:21 5-23 92. Plumley, F. G . , Schmidt, G. W. 1983. Annl. Biochem. 1 34:86-95 93. Dunsmuir, P. 1 985. Nucleic Acids Res. 1 3:2504- 1 8 94. Braumann, T . , Weber, G . , Grimme, L . H . 1 982. Photobiochem. Photobiophys.

4: 1-8

95. Kuhlbrandt, W . 1 984. Nature 307: 478-80 96. McDonnel, A . , Staehelin, L. A. 1 980. J. Cell Bioi. 84:40-56 97. Li, J. 1985. Proc. Natl. Acad. Sci. USA 82:386-90 98. Siefermann-Hanns, D . , Ross, J. W. ,

Kaneshiro , K. H . , Yamamoto, H. Y . 1 982. FEBS Lett. 1 49 : 1 9 1-96 99. Henriques, F . , Park, R. B . 1 976. Biochim. B i ophys . Acta 430:3 1 2-20 1 00. Chua, N . -H . , Matlin, K . , Bennoun , P. 1 975 . J. Cell B ioi . 67:361-77 1 0 1. Mullett, J . E. 1 983 . J. BioI. Chem. 258:9941-48 1 02. Bennett, J. 1 983. Biochem. J. 2 1 2 : 1 - 1 3 1 0 3 . Staehelin, L. A . , Arntzen, C . J. 1 983. 1. Cell BioI. 97: 1 327-37 104. Kung, S. D . , Thornber, 1. P . , Wildman, S. G. 1 972. FEBS Le tt . 24:1 85-88 1 05. Hoober, J. K. 1 970. 1. B i oi . Chem. 245:4327-33 1 06. Cashmore, A. R. 1 976. J. BioI. Chern. 2 5 1 : 2848-53 1 07. Machold, 0 . , Aurich, 0. 1 972. Biochim. Biophys. Acta 28 1 : 1 03-12 108. Ellis, R. 1. 1 983. Subcell. Biochern. 9:237--61 1 09. Apel, K . , Kloppstech, K. 1978. Eur. J. Biochem. 85:581-88 l lO. Apel, K . 1979. Eur. J. Biochern . 97: 1 83--8 8 1 1 1 . Cuming, A . C . , Bennett, J. 198 1 . Eur. J. Biochem. 1 1 8:7 1-80 1 1 2. Tobin, E. M. 1 98 1 . Plant Physiol. 67: 1 078-83 1 1 3. Kaufmann, L. S . , Thompson, W. F. , Briggs, W. R. 1 984. Science 226: 1 447-49 1 1 4. Grossman, A. R . , Bartlett, S . , Schmidt, G. W . , Chua, N.-H. 1979. Ann. NY Acad. Sci. 343:266-74 1 1 5. Mullet, J. E . , Grossman, A. R . , Chua, N.-H. 1 982. Cold Spring Harbor Symp. Quant. Bioi. 46:979-84 1 1 6. Mullet, J. E . , Burke , J. J . , Arntzen, C. J. 1 980. Pl ant P hysi ol. 65:823-27 1 1 7. Schwarz, H. P. , Kloppstech, K. 1 982. Planta 155: 1 1 6-23 1 1 8. Bellemare , G . , Bartlett, S. G . , Chua, N.-H. 1982. 1. Bioi . Chem. 257:776267 1 19. Viro, M . , Kloppstech, K. 1982. Pianta 1 54: 1 8-23 1 20. Bennett, J. 198 1. Eur. J. Biochem. 1 1 8:61-70 1 2 1 . Michel, H . , Tellenbach, M . , Boschetti, A. 1 983. Biochim. Bi ophys. Acta 725: 4 1 7-24 1 22. Akerlund, H. E . , Jansson, C. 1 98 1 . FEBS Lett. 124:229-32 123. Haehnel, W . , Berzborn, R. J . , Anders­ son, B . 1981. Biochim. Biophys. Acta 637:389-99 1 24. Boulter, D . , Haslett, B . G . , Peacock, D . , Ramshaw, J . A . M . , Scawen, M . D . 1 977. In Plant Biochemistry, ed. D . H . Northcote, 2: 1-40. Baltimore: Univer­ sity Park Press

CHLOROPLAST PROTEIN TRANSPORT 1 2 5 . Bohner, fl. , Bohme, H . , Boger, P. 1 98 1 . FEBS Lett. 1 3 1 :386-89 1 26 . Merchant, S . , Bogorad, L. 1986. Mol. Cell. Bioi. 6:462-69 127. Westhoff, P., Jansson, C . , Klein-Hit­ pass, L . , Benbom, R . , Larsson, C . , Bartlett, S. G. 1 985. Plant Mol. Bioi.

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

4: 1 37-46

1 28 . Cashmore, A. R. 1984. Proc. Natl. Acad. Sci. USA 8 1 :2960-64 1 29. Larnppa, G. K . , Morelli, G . , Chua, N.­ H . 1985. Mol. Cell. Bioi. 5 : 1 370-78 1 30. Karlin-Neumann, G. A . , Kohom, B . D . , Thornber, J . P. , Tobin, E . M. 1985. 1. Mol. Appl. Genet. 3:45-6 1 1 3 1 . Gilbert, W. 1978. Nature 271 :501 1 32 . Mount, S. M. 1982. Nucleic Acids Res. 10:459-72 1 3 3 . Smeekens, S . , de Groot, M . , van Bins­ bergen, J . , Weisbeek, P. 1985 . Nature 3 17:456-58 1 34. Heber, U . , Heldt, H. W. 1 98 1 . Ann. Rev. Plant Physiol. 32: 1 39-68 1 35 . Keegstra, K . , Werner-Washburne, M . , Cline, K . , Andrews, J. 1984. 1 . Cell. Biochern. 24:55-68 1 36. Douce, R . , Block, M. A . , Dome, A . -J . , Joyard, J. 1984. Subcell. Biochern. 10:184 1 37 . Nikaido, H. , Vaara, M. 1 985. Micro­ bioI. Rev. 49: 1-32 1 3 8 . Colombini, M. 1979. Nature 279:64345 1 39 . Flugge, U. I . , Benz, R. 1984. FEBS Lett. 169:85-89 140. Joyard, J . , Billecocq, A . , Bartlett, S. G . , Block, M . A . , Chua, N.-H . , Douce, R . 1 983. 1. Bioi. Chern. 258: 10000-6 1 4 1 . Cline, K . , Andrews, J . , Mersey, B . , Newcomb, E . H . , Keegstra, K. 1 98 1 . Proc . Natl. Acad. Sci. USA 78:359599 142. Joyard, J . , Grossman, A . , Bartlett, S . G . , Douce, R . , Chua, N.-H. 1982. 1. Bioi. Chern. 257: 1 095- 1 0 1 1 4 3 . Block, M. A . , Dome, A . J . , Joyard, J . , Douce, R. 1983. 1 . Bioi. Chern . 258: 1 3273-80 144. Block, M. A . , Dome, A . -J . , Joyard, J . , Douce, R. 1983 . 1 . Biol. Chern. 258:

1 328 1-86

145 . Werner-Washburne, M . , Cline, K . , Keegstra, K . 1983. Plant Physiol. 73: 569--7 5 146. Augge, U. I., Heldt, H . W. 1 98 1 . Biochirn . Biophys. Acta 638:296-304 147. Somerville, S . c . , Somerville, C. R. 1 985. Plant Sci. Lett. 37: 2 1 7-22 148. Hackenbrock, C. R . , Miller, K. J. 1975. 1. Cell Bioi. 65:615-30 149. Smit, J . , Nikaido, N. L. 1978. 1. Bac­ teriol. 1 35 :687-702

911

150. Carde, J.-P. , Joyard, J . , Douce, R. 1982. Bioi. Cell 44:3 1 5-24 1 5 1 . Wanner, G . , Formanek, H., Theimer, R . R . 1 98 1 . Planta 1 5 1 : 1 09-25 152. DiRienzo, J. M . , Inouye, M. 1 979. Cell

1 7 : 1 55-61

153. Augge, U. I. , Wessel, D. 1984. FEBS Lett. 168:255-59 154. Sagara, Y . , Ito, A. 1982. Biochern. Bio­ phys. Res. Cornrnun. 1 09: 1 1 02-7 155. Gasser, S . M . , Schatz, G. 1 983 . 1. Bioi.

Chern. 258:3427-30

156. Freitag, H . , Janes, M . , Neupert, W. 1982. Eur. 1. Biochern. 126: 1 97-202 157. Shore, G. C . , Power, F., Bendayan, M . , Carignan, P. 1 98 1 . 1 . Bioi . Chern. 256:8761-66 158. Cline, K . , Werner-Washburne , M . , An­ drews, J . , Keegstra, K. 1984. Plant Physiol. 75:675-78 1 59. Pfisterer, J . , Lachmann, P. , Kloppstech, K. 1982. Eur. 1. Biochern. 126:14348 160. Stocking, C . R . , Franceschi, V . R . 1 982. Plant Physiol. 70: 1 255-59 161 . Hennig, B . , Neupert, W. 1 98 1 . Eur. 1. Biochern. 1 2 1 :203- 1 2 162. Zwizinski, C . , Schleyer, M . , Neupert, W. 1983. 1. Biol. Chern. 258:407 1-74 163. Reizman , H . , Hay, R . , Witte, C . , Nel­ son, N . , Schatz, G. 1983. EMBO 1. 2: 1 1 1 3-1 8 164. Grossman, A . , Bartlett, S . , Chua, N . -H . 1980. Nature 285:625-28 165. Werner-Washburne, M . , Keegstra, K .. 1985. Plant Physiol. 78:22 1-27 166. Chen, L . , Tai, P. C. 1985. Proc. Natl. Acad. Sci. USA 82:4384-88 167. Soli, J. , Buchanan, B . B. 1983. J. Bioi. Chern. 258:6686-89 168. Reid, G. A. 1985. Curro Top. Biornernbr. Transp . 34:295-336 169. Gasser, S. M . , Daum, G . , Schatz, G. 1982. 1. Bioi. Chern. 257: 1 3034-41 1 70. Schleyer, M . , Schmidt, B . , Neupert, W. 1982. Eur. 1. Biochern. 1 25 : 1 09-16 171 . Walker, G . H . , Izawa, S. 1979. Plant Physiol. 63: 1 33-38 172. Von Heijne, G. 1985. Curro Top. Mernbr. Transp . 24: 1 5 1-79 173. Duffaude, G. D . , Lehnhardt, S . K . , March, P. E. , Inouye, M . 1985 . Curro Top. Mernbr. Transp. 24:65-103 174. Chen, L . , Rhoads , D., Tai, P. C . 1 985. J. Bacteriol. 1 6 1 : 973-80 175. Inouye, S . , Soberon, X . , Franceschini, T. , Nakamura, K . , Itakura, K . , Inouye, M. 1982. Proc. Natl. Acad. Sci. USA 79:3438-41 176. Copeland, B . R . , Landick, R . , Nazos, P. M. , Oxender, D . L. 1984. 1. Cell. Biochern. 24:345-56

Annu. Rev. Biochem. 1986.55:879-912. Downloaded from www.annualreviews.org Access provided by CONRICYT EBVC and Econ Trial on 09/23/15. For personal use only.

912

SCHMIDT & MISHKIND

1 7 7 . Schatz, G . , Butow, R. A. 1983 . Cell 32:3 1 6- 1 8 178. Freitag, H . , Janes, M . , Neupert, W . 1 982. Eur. J . Biochem. 1 26 : 1 97-202 1 79. Walter, P . , Gilmore, R . , Blobel, G . 1 984. Cell 38:5-8 1 80. Robinson , c . , Ellis, R. J. 1984. Eur. J. Biochern . 142:337-42 1 8 1 . Mishkind, M. L . , Jensen, K. H . , Brana­ gan , A. J . , Plumley, F. G . , Schmidt, G. W . 1985 . Curro Top . Plant Biochem. Physiol. 4:34-50 1 82 . Schmidt, G. W . , Mishkind, M . L. 1983. Proc. Natl. Acad. Sci. USA 80:2632-36 1 8 3 . Smith, S . M . , Ellis, R. J. 1979. Nature 278:662-64 1 84 . McAda, P. C . , Douglas, M. G. 1982. J . Bioi. Chem. 257:31 77-82 1 85 . Miura, S . , Mori , M . , Amaya, Y . , Tati­ bana, M. 1982. Eur. J. Biochem. 1 22:641-47 1 86. Bohni, P. C . , Daum, G . , Schatz, G. 1 983. J. Bioi. Chern. 258:4937-43 1 87. Spencer, N. 1 984. J. Thear. Bioi. 107: 405-1 5 1 88 . Marks, D . B . , Keller, B . J . , Hoober, J . K . 1985. Plant Physiol. 79:108-13 1 89 . . Roy, H . , Patterson, R . , Jagendorf, A. T. 1976. Arch. Biochern. Biophys . 172:6473 1 90. Gray, J . c . , Kekwick, R . G . O. 1973. FEBS Lett. 38:67-69 1 9 1 . Robinson, C . , Ellis, R . J. 1 984. Eur. '­ Biochern. 142:343-46 1 92 . Schmidt, R. J . , Gillham, N. W . , Boyn­ ton, J. E. 1985 . Mol. Cell:Biol. 5 : 1 09399 1 93 . Teintze, M . , Slaughter, M . , Weiss, H . , Neupert, W. 1982. J . Bioi. Chern. 257: 1 0364-7 1 1 94. Gasser, S. M . , Ohashi, A . , Daum, G . , Bohni, P. C . , G ibson , J . , e t a l . 1982. Proc. Natl. Acad. Sci. USA 79:267-7 1 1 95 . Schmidt, B . , Wachter, E . , Sebald, W . , Neupert, W . 1984. Eur. J. Biochem . 144:58 1-88 1 96. Ohashi , A . , Gibson, J . , Gregor, I . , Schatz, G . 1982. J. Bioi. Chern. 257: 1 3042-47 1 97 . Beams, H. W . , Kessel, R . G . , Shih, C. Y . 1979. Bioi. Cell. 35:87-96

198. Laetsch , W. M. 1974. Ann. Rev. Plant Physiol. 25:27-52 1 99 . Tremolieres, A., Dubacq, J. P., Ambard-Bretteville, F . , Remy , R . 1 98 1 . FEBS Lett. 1 30:27-3 1 - 200. Schwitzguebe1 , J: P . ; Nguyen , T. D . , Siegenthaler, P . 1984. In Function and Metabolism of Plant Lipids. ed. P. A . Siegenthaler, W . Eichenberger, p p . 299302. Amsterdam: Elsevier 20 1 . Verkleij , A. J. 1982. Biochim. Biophys. Acta 779:43-63 202, Gruner, S. M. 1985. Proc. Nat!. Acad . Sci. USA 82:3665-69 203. Sprague , S. G . , Staehelin, L. A. 1984. Plant Physiol. 502-4 204. Harpster, M . H . , Mayfield, S. P . , Taylor, W . C . 1984. Plant Mol. BioI. 3:59-7 1 205. Mayfield, S. P., Taylor, W. C. 1984. Eur. J . Biochern. 1 44:79-84 206. Mishkind, M. L . , Greer, K. S . , Schmidt, G. W. 1986. Methods Enzyrnol. In press 207. Ohta, S . , Schatz , G. 1 984. EMBO J. 3:65 1-57 208. Firgaira, F. A . , Hendrick, J. P . , Dalousek, F . , Draus, J. P . , Rosenberg, L. E. 1984. Science 226: 1 3 1 9-22 209. Schmidt, B . , Henning, B . , Zimmer­ ,mann, R . , Neupert, W. 1983. J. Cell . BioI. 96:248-55 2 10 . Zimmermann, R . , Neupert, W. 1980. Eur. '- Biochem. 109:217-29 2 1 1 . Miura, S . , Mori, M . , Amaya, Y . , Tati­ bana , M . , Cohen, P. P. 1 98 1 . Biochern. Int. 2:302- 1 2 2 1 2. Ades, 1 . Z. 1983. Biochem. Biophys. Res. Cornrnun . 1 10:42-47 2 1 3 , Yamaoto, M . , Watanabe, N . , Kikuchi, G. 1983. Biochern . Biophys. Res. Corn­ rnun. 1 1 5:700-6 2 1 4 . Johanningmeier, U . , Howell, S. H. 1 984. J. BioI. Chem. 259 : 1 3541-49 2 1 5 . Fues1er, T. P. , Castelfranco, P. A . , . Wong, Y.-S. 1 984. Plam Physiol. 74: 928-33 2 1 6 . Crawford, M. S. , Wang, W . -Y . , Jensen, K. G. 1982. Mol. Gen. Genet. 1 88 : 16 2 1 7 . Yaffe , M. P . , Schatz, G. 1984. Proc. Natl. Acad. Sci. USA 8 1 :4819-23