The route of entry of cytoplasmic ally synthesized proteins into: Chloroplasts of algae possessing chloroplast ER

Although chloroplasts contain DNA and have a complete protein-synthesizing apparatus, most chloroplast proteins are synthesized on the 80-s ribosomes of the cytoplasm and are subsequently transported to the chloroplast. At present very little is known about how proteins cross the 2 membranes of the chloroplast envelope. Recently Dobberstein, Blobel & Chua (1977) working with Chlamydomonas and Highfield & Ellis (1978) working with pea seedlings have shown that one cytoplasmically synthesized plastid protein, the small subunit of ribulose bisphosphate carboxylase, can be synthesized in vitro on free polysomes as a larger precursor than the final protein. It is believed that the additional segment on the precursor protein is recognized by a specific binding site on the outer membrane of the chloroplast envelope and that the protein is cleaved prior to crossing the envelope membranes.

In most groups of algae, however, the chloroplasts, in addition to having a doublemembraned envelope, are completely enclosed by a cisterna of endoplasmic reticulum which has ribosomes on its outer membrane. A layer of vesicles and tubules is present at one place between the chloroplast ER and the chloroplast envelope. In this paper it is proposed that nuclear-coded plastid proteins are synthesized on the polysomes on the chloroplast ER and pass during synthesis into the lumen of the ER cisterna. From there these proteins are transported via the vesicles to the lumen of the chloroplast envelope. From there the proteins in an unknown manner cross the inner membrane of the chloroplast envelope to enter the chloroplast matrix. A preliminary account of this work has been presented (Gibbs, 1977).

Stocks of Olisthodiscus luteus Carter were obtained from Dr R. R. L. Guillard of the Woods Hole Oceanographic Institute, Woods Hole, Mass., U.S.A. The strain was originally isolated by R. J. Conover from Long Island Sound at Milford, Conn., in 1953. Cultures were grown at 20 °C in half strength ‘f’ medium (Guillard & Ryther, 1962) on a 13-h light, 11-h dark cycle at a light intensity of 450 ft-c. (4·84 × 10⁶ lux). Cells were fixed in a solution of 5 % glutaraldehyde in 0 ·1 M sodium phosphate buffer, pH 7 · 8, to which 0 · 25 M sucrose had been added to give an osmolarity of 1218 mosm. Cells were fixed at o °C for 15 h, rinsed, and then postfixed in 1 % osmium tetroxide in 0 ·1 M sodium phosphate buffer, pH 7 · 8, for 1 h at 0 °C. After dehydration in an ethanol series, cells were embedded in Spurr’s (1969) low viscosity epoxy resin. Sections were stained with 2% potassium permanganate in 0 ·1 M sodium phosphate buffer, pH 7·o, followed by Reynold’s (1963) lead citrate.

Stocks of Ochromonas danica Pringsheim were obtained from the Culture Collection of Algae, which is now located at the University of Texas in Austin (Culture no. 1298). Cells were grown at 29 °C in Aaronson & Baker’s (1959) medium either in the dark or under a bank of fluorescent and incandescent lamps adjusted to give a light intensity of 450 or 600 ft-c. (4·8 4 × 10^s or 6·46 × 10⁶ lux). In the experiments with chloramphenicol and cycloheximide, log phase dark-grown control and inhibitor-treated cultures were allowed to green in the light for 3, 6, 9, 12, 24, 48, and 96 h before being prepared for electron microscopy. O-threo chloramphenicol (Sigma Chemical Co., St Louis, Mo.) was added to dark-grown cultures at a concentration of 300 μg/ml; cycloheximide (Sigma·Chemical Co., St Louis, Mo.) was added to similar cultures at a concentration of 8 μg/ml. In other experiments, spectinomycin (The Upjohn Co., Kalamazoo, Michigan) was added to both fully green cultures and dark-grown cultures at a concentration of 100 μg/ml and the cultures were allowed to grow in the light for a further 24 h prior to fixation. Two methods of fixation were employed in both experiments. The first was a standard fixation in which cells were fixed in 2 · 5 % glutaraldehyde in 0 ·1 M sodium phosphate buffer, pH 7 · 2, for 30 min at room temperature, rinsed, and postfixed in 1 % osmium tetroxide in the same buffer for 1 h at room temperature. In the second fixation, designated cofixation, the cells were fixed simultaneously in glutaraldehyde and osmium tetroxide by the method of Falk (1969). Cells were fixed for 30 min at room temperature in a freshly mixed solution of 2 % glutaraldehyde and 2% osmium tetroxide in 0 · 05 M sodium phosphate buffer, pH 7 · 2, and subsequently postfixed in 1 % osmium tetroxide in 0 ·1 M sodium phosphate buffer, pH 7 · 2, for 1 h. Cells were embedded by the same method as the Olisthodiscus cells and stained by lead citrate alone. The figure illustrating polyribosomes (Fig. 4) was from an Ochromonas culture which had been starved in the dark for 3 days in substrate-free medium (Aaronson & Baker, 1959) and then placed in the light to green for 12 h. These cells were fixed by the standard method and embedded in Araldite.

In the red and green algae, the chloroplasts are enclosed solely in a doublemembraned chloroplast envelope. In the Dinophyceae and Euglenophyceae, chloroplasts are surrounded by 3 membranes (Gibbs, 1970; Dodge, 1973). In all other classes of algae, namely the Cryptophyceae, Raphidophyceae, Haptophyceae, Chrysophyceae, Bacillariophyceae, Xanthophyceae, Eustigmatophyceae, and Phaeophyceae, the chloroplasts are always surrounded by 4 membranes, 2 of the chloroplast envelope and 2 of a layer of endoplasmic reticulum (Gibbs, 1970; Hibberd & Leedale, 1970) which has been named the chloroplast ER (Bouck, 1965).

Fig. 1 is a section through the pyrenoid region of a chloroplast from the Chrysophyte alga, Olisthodiscus luteus. This alga has numerous peripheral chloroplasts which lie so that the pyrenoid of each chloroplast projects inward toward the cell’s central nucleus (Cattolico, Boothroyd & Gibbs, 1976). It can be seen that the chloroplast is limited by a double-membraned chloroplast envelope whose 2 membranes characteristically lie very close to each other. Outside a layer of vesicles and tubules, there is a more widely spaced double membrane, the chloroplast ER, which has ribosomes on its outer membrane (arrow). In favourable sections it can be seen that this layer of chloroplast ER is continuous around the entire chloroplast and as is usual with organisms with many chloroplasts, there are no connexions with the nuclear envelope.

In species of algae with chloroplast ER which have only one or two chloroplasts which lie closely appressed to the nucleus, the chloroplast ER is continuous with the outer membrane of the nuclear envelope (Gibbs, 1962). Fig. 2, which is a section through part of the nucleus and the adjacent chloroplast of the Chrysophyte alga, Ochromonas danica, shows this clearly. The point of continuity between the outer membrane of the nuclear envelope and the chloroplast ER is marked by a star. It can be seen that in these algae the nuclear envelope forms one wall of the sac of ER which encloses the chloroplast.

The cell in Fig. 2 is fixed by the simultaneous glutaraldehyde-osmium tetroxide method and most of the ribosomes on the chloroplast ER and in the cytoplasm have been lost. Figs. 3 and 4 are cells of Ochromonas danica which have been fixed by traditional methods and ribosomes are fairly well preserved. In Fig. 4 the chloroplast ER in the region where it joins the nuclear envelope is cut tangentially and it can be seen that ribosomes on the chloroplast ER are arranged in polysomes. Fig. 3 shows that chloroplast ER may at places be continuous with the rough ER of the cytoplasm (arrow). Ribosomes have been shown to be present on the outer membrane of the chloroplast ER in all the groups of algae which possess it (Gibbs, 1978) and connexions between the chloroplast ER and rough ER of the cytoplasm have also been demonstrated for most of the groups (Gibbs, 1978).

An almost universal feature of algae which have chloroplast ER is the presence of a layer of tubules and vesicles between the chloroplast envelope and the chloroplast ER. These vesicles and tubules have been named the periplastidal reticulum (Falk & Kleinig, 1968). The periplastidal reticulum does not extend around the whole chloroplast but is characteristically restricted to a particular location. In Olisthodiscus luteus, it is located outside the pyrenoid region of the chloroplast (Fig. 1). In Ochromonas danica, the periplastidal reticulum lies in the space separating the nuclear envelope and the chloroplast envelope (Fig. 2). Although most cells of Ochromonas studied have periplastidal reticulum only in this location, some cells show smaller regions of periplastidal reticulum at the anterior tips of the 2 lateral lobes of the chloroplast or on the outside edge of a lobe of a chloroplast opposite the periplastidal reticulum adjoining the nucleus.

Fig. 5 is a tangential section through the periplastidal reticulum of Ochromonas danica. It can be seen that it consists of a tubular reticulum with vesicles lying in the interstices of the reticulum. The periplastidal reticulum of Olisthodiscus luteus appears to have a similar structure. The 3-dimensi0nal structure of the periplastidal reticulum has seldom been demonstrated in the literature. Falk & Kleinig (1968) diagrammed it as a reticulum in Tribonema (Xanthophyceae), but their tangential view of it (fig. 6a) shows clearly that vesicles are present in some of the interstices of the reticulum. A tangential view of the periplastidal reticulum in the Haptophycean alga Chrysochromulina chiton (Manton, 1967, fig. 5) shows that it too has vesicles lying between tubular elements which probably form a reticulum.

Periplastidal reticulum cut in cross-section, i.e., a close-spaced row of vesicles and tubules lying between the chloroplast envelope and the chloroplast ER, has been seen in the Raphidophyceae (Heywood, 1972, fig. 7 and Mignot, 1976, fig. 3), Haptophyceae, Chrysophyceae, Bacillariophyceae, Xanthophyceae, Phaeophyceae (references in Gibbs, 1970) and in the Eustigmatophyceae (Hibberd & Leedale, 1972, fig. 6b and Hibberd, 1974, fig. 18). The only group of algae which has chloroplast ER, but does not have a defined periplastidal reticulum is the Cryptophyceae. In the Cryptophyceae, the space between the chloroplast envelope and the chlorcplast ER is relatively wide and contains starch grains, ribosome-like structures, and double membrane-limited bodies called nucleomorphs (Oakley & Dodge, 1976; Greenwood, Griffiths & Santore, 1977). Scattered vesicles of various sizes are also present (Gibbs, 1962, fig. 10; Lucas, 1970, fig. 8; Wehrmeyer, 1970, fig. 1; Heath, Greenwood & Griffiths, 1970, fig. 18; Oakley & Dodge, 1976, fig. 8). Since periplastidal reticulum appears in cross-section as a row of vesicles and tubules, these scattered vesicles, although they could serve the same function as the vesicles of the periplastidal reticulum, are distinctly different in structure so should not be called by the same name.

The presence of polysomes on the chloroplast ER indicates that proteins are synthesized on the ribosomes located on the cytoplasmic surface of the chloroplast ER. These proteins presumably pass during synthesis into the lumen of the chloroplast ER. At the region of the periplastidal reticulum, vesicles can be seen in the process of pinching off the chloroplast ER or its extension, the nuclear envelope. This is shown for Ochromonas in Figs. 2 and 6 (arrows). It is presumed that these vesicles contain protein and that the periplastidal reticulum originates from them. Vesicles can also be seen in the process of fusing with the chloroplast envelope. The vesicle indicated by the arrow in Fig. 7 appears to be fusing with the outer membrane of the chloroplast envelope, but the connexion is only suggestive because the membranes are not cut perpendicularly. In Fig. 8 one can clearly see that the membrane of the vesicle is continuous with the outer membrane of the chloroplast envelope, although slightly above the point of vesicle fusion the membranes of the chloroplast envelope become oblique for a short distance. In Fig. 9 the continuity of the vesicle membrane with the outer membrane of the chloroplast envelope is also indisputable. By this process of vesicles pinching off the chloroplast ER and subsequently fusing with the outer membrane of the chloroplast envelope, chloroplast proteins synthesized on the ribosomes bound to CER could reach the intra-envelope space of the chloroplast. From there proteins presumably cross the inner membrane of the chloroplast envelope. In several thousands of micrographs of Ochromonas and Olisthodiscus, I have never seen vesicles pinching off the inner membrane of the chloroplast envelope into the chloroplast matrix. Even if they did, chloroplast proteins would still be separated from their destination by a membrane.

Although no one has previously suggested that the vesicles of the periplastidal reticulum carry protein from the chloroplast ER to the lumen of the chloroplast envelope, a number of investigators have observed that the periplastidal reticulum appears to form from vesicles that pinch off the chloroplast ER or its extension, the nuclear envelope. In the Phaeophyceae, vesicles which appear to be pinching off the chloroplast ER can be seen in Giffordia (Bouck, 1965, fig. 7) and in 2 species of Sphacelaria (Evans, 1966, fig. 5 and Galatis, Katsaros & Mitrakos, 1977, fig. 8). In the Xanthophyceae, there are inward facing projections on the chloroplast ER in the region of the periplastidal reticulum in Tribonema (Falk & Kleinig, 1968, fig. 6b, 6c). In the Bacillariophyceae, Crawford (1973) notes that in Melosira varians connexions are infrequently present between the chloroplast ER and the periplastidal reticulum. And in another Chrysophyte, Chrysamoeba radians, Hibberd (1971, figs. 21, 22) points out that vesicles of the periplastidal reticulum can be seen pinching off the chloroplast ER and the nuclear envelope.

To date no one has observed fusions between the periplastidal reticulum and the chloroplast envelope. In a search of the literature I have found one micrograph which shows a vesicle which appears to be in the process of fusing with the outer membrane of the chloroplast envelope. This micrograph is of the brown alga Petalonia debilis (Cole, 1970, fig. 5).

An interesting anomaly has been observed in the brown alga Ectocarpus by Oliveira & Bisalputra (1973). Although this alga has typical periplastidal reticulum, a direct connexion between the chloroplast ER and the chloroplast envelope was also seen. Thus it would be possible in this species for protein to pass directly from the lumen of the chloroplast ER to the lumen of the chloroplast envelope.

Studies with inhibitors of protein synthesis on chloroplast and cytoplasmic ribosomes have provided experimental evidence that the periplastidal reticulum is involved in the transport of cytoplasmically synthesized proteins to the chloroplast.

When dark-grown cells of Ochromonas are placed in the light in the presence of 300 μg/ml chloramphenicol, an inhibitor of protein synthesis on chloroplast ribosomes, chloroplast growth is markedly reduced as is the development of chloroplast thylakoids (Smith-Johannsen & Gibbs, 1972). In many of the cells treated with chloramphenicol for 6 h or more, the periplastidal reticulum becomes hypertrophied. The vesicles become 2 layers deep (Fig. 12) or at places 3 or 4 layers deep (Fig. 10). Also in other chloramphenicol-treated cells, the periplastidal reticulum extends a considerable distance laterally. The junction of the chloroplast ER and the nuclear envelope in these cells is not at the border of the chloroplast but a considerable distance away and the periplastidal reticulum proliferates to fill this space (Smith-Johannsen & Gibbs, 1972). In cells treated with spectinomycin, another inhibitor of protein synthesis on chloroplast ribosomes, hypertrophy of the periplastidal reticulum also occurs, but not to as great an extent as it does in cells treated with chloramphenicol. In some green cells which are exposed to 10μg/ml spectinomycin for 24 h, a proliferation of the vesicles of the periplastidal reticulum occurs. In greening cultures which have been exposed to spectinomycin for 24 h, more cells show an increase in periplastidal vesicles (Fig. 13). It is concluded that treatment of both green and greening cultures of Ochromonas with inhibitors of plastid protein synthesis causes a build up of the periplastidal reticulum presumably because transport of cytoplasmically synthesized proteins into the chloroplast is slowed down.

Conversely, in cells of Ochromonas danica treated with 8 μg/ml cycloheximide, an inhibitor of protein synthesis on cytoplasmic ribosomes, there is a reduction in the amount of periplastidal reticulum. This often appears as a complete absence of vesicles and tubules in the space between the nuclear envelope and the chloroplast envelope (Fig. 11). For example, after 24 h of greening in control cells, 91 % of the sections of chloroplast-nucleus junctions contained periplastidal reticulum, whereas in cycloheximide-treated cells only 58% of the chloroplast-nucleus junctions contained periplastidal reticulum, the remaining 42 % having no reticulum at all. This suggests that in cycloheximide-inhibited cells, periplastidal reticulum frequently fails to form as there are no cytoplasmically synthesized proteins to be transported into the chloroplast.

Several investigators have speculated on the function of the chloroplast ER, but no one has considered that it is involved in the synthesis of nuclear-coded plastid proteins. I originally suggested it might be involved in the transport of material from the nucleus to the chloroplast (Gibbs, 1962). Bouck (1965) suggested that photosynthates diffused through the chloroplast envelope and entered the chloroplast ER. From there he postulated that some of the products of photosynthesis could pass via the nuclear envelope to the Golgi apparatus to be fashioned into secretory granules. Cole & Wynne (1973) have concurred with this suggestion. On morphological grounds alone it is impossible to determine whether the chloroplast ER —periplastidal reticulum —chloroplast envelope pathway demonstrated in this paper is concerned with transporting proteins into the chloroplast or carrying products of photosynthesis away from the chloroplast. However, the responses of the periplastidal reticulum to inhibitors of protein synthesis on both cytoplasmic and chloroplast ribosomes suggests strongly that the periplastidal reticulum is involved in the transport of proteins into the chloroplast.

Proteins destined for the chloroplast may come to be synthesized exclusively on the chloroplast ER by the same mechanism that assures that secretory proteins in animal cells are synthesized on rough endoplasmic reticulum (Blobel & Dobberstein, 1975). Messenger RNAs for nuclear-coded chloroplast proteins may have a signal sequence that when translated binds the polysomes synthesizing plastid proteins to the chloroplast ER.

Since the chloroplasts of red and green algae are simply enclosed in a doublemembraned envelope which never has endoplasmic reticulum associated with it in any way, it would not be possible for the mechanism of transporting proteins into chloroplasts proposed in this paper to function in these algae. Instead Dobberstein et al. (1977) have found that in the green alga Chlamydomonas reinhardi, the small subunit of ribulose bisphosphate carboxylase is translated on free cytoplasmic polysomes. They also found that when translated in a cell-free wheat germ system the small subunit contained an additional sequence which they postulated was involved in the specific binding of the protein to the chloroplast envelope. Nothing is known yet about how the protein crosses the 2 membranes of the chloroplast envelope although Dobberstein et al. (1977) suggest proteins cross where the 2 membranes of the chloroplast envelope lie appressed to each other (Ohad, Siekewitz & Palade, 1967).

Neither is it known yet how cytoplasmically synthesized proteins cross the barrier imposed by the 3 membranes which surround the chloroplasts of dinoflagellates and euglenoids. None of these membranes appears to be related to the chloroplast ER system. There are no ribosomes on the outermost membrane, nor is it ever connected to rough ER of the cytoplasm (Gibbs, 1978). Nor are vesicles similar to periplastidal reticulum present. It may be that the third membrane represents the plasmalemma of an original eukaryotic symbiont which has become progressively reduced so that today only the chloroplast and the cell membrane of the symbiont persist (Gibbs, 1978).

Higher green plants are the direct descendants of green algae and like them have chloroplasts solely enclosed by a double-membraned chloroplast envelope. In a few cell types of a few species, a single ER element lies close to the chloroplast (Larson, 1965; Camefort, 1965; Wooding & Northcote, 1965; Cecchi Fiordi & Maugini, 1972; Rodríguez-García & Sievers, 1977). However, this association between ER and chloroplast is distinctly different from that which occurs in algae. The ER elements do not completely enclose the chloroplast, and also in well fixed cells it can be seen that ribosomes are present on both sides of the ER cisterna (Cecchi Fiordi & Maugini, 1972, fig. 3, and Rodríguez-García & Sievers, 1977, fig. 5). Also no vesicles are ever found between the ER element and the chloroplast. Thus there is nothing in the morphology of these ER-chloroplast associations to suggest a transfer of protein from the ER cisterna to the chloroplast. However, in the microspores of Lycopersicum esculentum and Solanum tuberosum (Abreu & Santos, 1977), there is a much more intimate association between the endoplasmic reticulum and the proplastids. At places on the proplastid’s periphery, cisternae of ER with ribosomes on their outer membrane lie appressed to the chloroplast envelope. The inner membrane of each ER cisterna has no ribosomes on it and lies closely appressed to the outer membrane of the plastid envelope. In such an arrangement proteins might pass directly from the ER lumen to the proplastid envelope lumen through the fused ER-plastid envelope membranes. Another type of close association between ER and chloroplast envelope has been observed in ferns (Crotty & Ledbetter, 1973; Cran & Dyer, 1973) and liverworts (Diers, 1966). In these cases, an element of ER is seen in direct continuity with the outer membrane of the chloroplast envelope and proteins synthesized on the ER element could pass without hindrance into the lumen of the chloroplast envelope.

However, these few records of an intimate association between ER and chloroplast envelope are so exceptional in green plants that they cannot be the main method by which cytoplasmically synthesized proteins enter chloroplasts in these organisms. Recently, Cashmore, Broadhurst & Gray (1978) and Highfield & Ellis (1978) have shown that in peas as in Chlamydomonas the small subunit of ribulose bisphosphate carboxylase is synthesized as a precursor of higher molecular weight (20000) than the final molecule. Highfield & Ellis (1978) further showed that processing of the 20 000 mol. wt. precursor to the small subunit occurs maximally in the presence of chloroplasts which retain their envelopes. They propose that removal of the extra sequence by the chloroplast envelope triggers a conformational change that leads to the transport of the small subunit across the chloroplast envelope.

I wish to thank my former students, Dr. Tiiu Slankis, Dr. Heidi Smith-Johannsen and Mrs Lily Chu for the use of many of the micrographs on which this study is based. This work was supported by the National Research Council of Canada (Grant no. A-2921) and the Quebec Department of Education.