Chloroplast biogenesis: The use of mutants to study the etioplast–chloroplast transition

Abstract

In angiosperm plants, the etioplast–chloroplast transition is light-dependent. A key factor in this process is the protochlorophyllide oxidoreductase A (PORA), which catalyzes the light-induced reduction of protochlorophyllide to chlorophyllide. The import pathway of the precursor protein prePORA into chloroplasts was analyzed in vivo and in vitro by using homozygous loss-of-function mutants in genes coding for chlorophyllide a oxygenase (CAO) or for members of the outer-envelope solute-channel protein family of 16 kDa (OEP16), both of which have been implied to be key factors for the import of prePORA. Our in vivo analyses show that cao or oep16 mutants contain a normally structured prolamellar body that contains the protochlorophyllide holochrome. Furthermore, etioplasts from cao and oep16 mutants contain PORA protein as found by mass spectrometry. Our data demonstrate that both CAO and OEP16 are dispensable for chloroplast biogenesis and play no central role in the import of prePORA in vivo and in vitro as further indicated by protein import studies.

Chloroplasts and mitochondria originated from two independent unique endosymbiotic events. In the course of organelle evolution, most of the endosymbiont's genes were transferred to the host nucleus (1, 2). Therefore, today chloroplasts and mitochondria must import the vast majority of their protein constituents in a posttranslational event from the cytosol (3–5). In general, chloroplast-destined proteins are made as precursor forms in the cytosol and contain an N-terminal transit peptide that is both necessary and sufficient for targeting and translocation into chloroplasts. Upon translocation, the transit peptide is cleaved off by a stromal processing peptidase and the mature protein is formed (6). For the majority of preproteins, recognition and translocation into chloroplasts is achieved by two distinct translocons, the Toc and Tic complexes, situated in the outer envelope or the inner envelope of chloroplasts, respectively. The Toc complex is composed of the three core subunits, i.e., the GTP-dependent Toc33/Toc34 receptor isoforms, the GTP-dependent Toc159 precursor binding and motor protein, and the Toc75-composed translocation channel (for review, see refs. 3–5). Toc64 is transiently attached to the Toc core complex and represents a receptor for Hsp90-associated preproteins (7). The Toc complex seems to act in close cooperation with the Tic complex, which is mediated by an intermembrane space complex consisting of Toc12, an Hsp70 homolog, Toc64, and Tic22 (7). The inner-envelope-integral Tic translocon consists of the channel-forming subunits Tic110 and Tic20 and less well characterized subunits like Tic62, Tic55, Tic40, and Tic32 (for review, see refs. 3–5).

However, recent findings suggest that further import pathways exist in chloroplasts. Proteins like a quinone oxidoreductase homolog (8) and Tic32 (9) are imported without a cleavable presequence, but targeting is provided by internal sequence information. Furthermore, NTP requirements as well as the involvement of the classical Toc complex differ from the standard import route. The proteome analysis of chloroplasts from Arabidopsis thaliana suggests that presequence-independent targeting might be much more widespread than previously anticipated (10, 11).

An even more complex import pathway has been proposed for protochlorophyllide oxidoreductase (POR) (12–17). This enzyme catalyzes the light-dependent conversion of protochlorophyllide (Pchlide) to chlorophyllide (Chlide), which represents an essential step in chloroplast formation in angiosperms (18, 19). POR has three isoforms in Arabidopsis: PORA, PORB, and PORC (20). PORA is present only in etiolated tissue in the dark but is rapidly degraded in the light. In seedlings, PORB occurs concomitantly with PORA but remains stable in the light and is regulated in a circadian rhythm (20, 21). PORC is present in the leaves of light-grown plants (22). Whereas prePORB seems to follow a general import pathway, the import of prePORA into isolated chloroplasts or etioplasts was suggested to depend on its substrate Pchlide (14, 15, 17). Although these results could not be verified in vitro (23, 24), in vivo data demonstrated that prePORA requires its substrate Pchlide to become imported into chloroplasts of cotyledons (12, 13). The substrate dependence is lost in vivo in true leaves and therefore seems developmentally regulated (12). A closer analysis of the components involved in the import pathway of prePORA in isolated chloroplasts of barley and Arabidopsis identified Toc33 and OEP16 by chemical cross-linking as proteinaceous subunits of the prePORA import pathway (14–16). In vivo data, however, indicate that neither Toc33 nor its isoform Toc34 are necessary for prePORA recognition and translocation (13).

OEP16 was identified originally as a solute channel in the outer envelope of chloroplasts of pea with selectivity for amino acids and amines (25). In Arabidopsis, four genes named OEP16.1–OEP16.4 were designated to code for OEP16 isoforms (16). In phylogenetic analysis, however, OEP16.3 outgroups and does not belong to the true OEP16 subfamily (26). Protein sequencing data suggest that OEP16.1 is the most prominent isoform in the outer envelope of Arabidopsis (10, 11) and barley (16) and shows the strongest sequence identity with OEP16 from pea. Barley OEP16.1 was suggested to form the import channel for prePORA (16). In addition, import of prePORA via this nonstandard pathway was further indicated to depend on Pchlide b (15), the existence of which is highly disputed. Although Pchlide b seems completely absent from etiolated tissue, minor amounts could be detected in green leaves (27). To our knowledge, no biochemical evidence is present that indicates that Pchlide b is a true intermediate in the formation of chlorophyll (Chl); rather, it is a fortuitous degradation product in vitro (27).

To address these conflicting results, we have made use of Arabidopsis mutants. Arabidopsis mutant lines have been used successfully to study photomorphogenesis (28). Here, a similar approach was applied to specifically address the etioplast–chloroplast transition. T-DNA insertion lines were studied that were either devoid of OEP16.1, OEP16.2, or OEP16.1 and OEP16.4 in a double mutant, respectively. In no case could we detect impaired prePORA import in vitro or a retardation of greening in vivo. In addition, we analyzed a Chlide a oxidase (CAO) loss-of-function mutant that is devoid of Chl b and Pchlide b. Again, the absence of Pchlide b did not cause a failure or decrease of prePORA to import into chloroplasts in vitro. Plants without Chl b developed normally and showed a normal etioplast-to-chloroplast transition. Together, our data demonstrate that neither OEP16 nor Pchlide b is involved in prePORA import.

Results

Previous reports indicated that the import of prePORA into isolated chloroplasts involves two factors: first, OEP16 was tentatively identified as a translocon subunit for prePORA import, and second, prePORA import was reported to depend on the presence of its substrate Pchlide, particularly on Pchlide b (14–17). To test this latter assumption, we isolated a loss-of-function mutant in Arabidopsis that is devoid of the enzyme CAO, which converts Chlide a to Chlide b during Chl b synthesis (29, 30). Homozygous cao-1 plants were able to grow on soil under standard growth conditions. When compared with wild type, they were slightly less green in color but flowered normally and seeds had similar germination rates [supporting information (SI) Fig. 6]. Pigment analysis demonstrated the complete absence of Chl b from cao-1 lines (Fig. 1). In addition, when cao-1 seedlings were incubated with δ-aminolevulinic acid (ALA), Pchlide a accumulated like in wild-type plants (Fig. 1 b and d). In no case were we able to detect Pchlide b, Chlide b, or Chl b in homozygous cao-1 lines (Fig. 1 and data not shown). These findings corroborate earlier reports showing that CAO is the only enzyme present in chloroplasts to catalyze the formation of Chl b in vivo (31).

Fig. 1.

Homozygous cao-1 mutants do not contain Chl b or Pchlide b. Eight-day-old seedlings of Col-0 wild-type (a and b) and cao-1 (c and d) were used for the extraction of pigments either without ALA (a and c) or after the addition of ALA (b and d) for 16 h in the dark. Arrowheads indicate the expected occurrence of Pchlide b. HPLC runs were monitored with a diode array detector. The relative absorbance at wavelengths of 430 nm (Chl a), 448 nm (Pchlide a/b), and 460 nm (Chl b) is depicted. Note that the extraction protocol used for the enrichment of Pchlide results in separation of most of Chl a and Chl b before HPLC analysis, resulting in different amounts of residual Chl (see Materials and Methods).

To further analyze the putative role of Pchlide b in the import of prePORA, we isolated intact chloroplasts from either light-grown seedlings, which contain almost no detectable amounts of Pchlide, or plantlets that were kept in the dark in the presence of ALA, which results in the accumulation of Pchlide (Fig. 2a) (12, 17). Incubation with ALA resulted in the accumulation of Pchlide a only; the occurrence of Pchlide b as a general intermediate of Chl is generally not accepted (27). Both prePORA and prePORB imported into isolated chloroplasts with similar yield and were processed to the mature form either in the absence or presence of large amounts of Pchlide (Fig. 2b). The processed mature forms were protected from externally added protease thermolysin, indicating that they had reached the inside of the organelle in contrast to the precursor forms, which were protease-susceptible, demonstrating that they were still on the surface of the chloroplast (Fig. 2b, compare lanes 3, 5, 7, and 9 with lanes 4, 6, 8, and 10). Furthermore, chloroplasts isolated from cao-1 plantlets were able to import prePORA and prePORB with similar yield when compared with wild-type, although, if these plants contained any Pchilde, it was Pchlide a and not Pchlide b. Etioplasts from dark-grown cao-1 plantlets contained both PORA and PORB (SI Table 1); in addition, etioplasts from cao-1 cotyledons exhibited a normal ultrastructure, especially of the prolamellar body, which contains the Pchlide holochrome (see below) (32). We conclude that Pchlide b plays a role in neither etioplast/chloroplast development in vivo nor import of prePORA and prePORB in vitro.

Fig. 2.

Import of prePORA and prePORB into chloroplasts is independent of Pchlide and can be competed for by preOE33. Chloroplasts were isolated from 8-day-old cotyledons of light-grown seedlings (light) or from plantlets harvested after a 16-h dark period, either without (dark) or with addition of ALA (dark + ALA). (a) Determination of Pchlide a from Col-0 wild-type (wt) and homozygous cao-1 plants grown under different conditions as outlined above. The Pchlide a content (n = 2) is given in nanomoles per gram of fresh weight (FW). (b) Import of prePORA and prePORB into chloroplasts from wild-type and cao-1 plantlets, either light-grown or treated with ALA (see above). Chloroplasts equivalent to 10 μg of Chl were incubated with the in vitro translated precursor proteins (Tl). After import, chloroplasts were either not treated (−) or treated (+) with the protease thermolysin. Precursor and mature proteins are indicated by asterisks and arrowheads, respectively. The letter A or B above each lane denotes the prePOR isoform used. (c) The import of prePORA and prePORB was tested in wild-type chloroplasts using urea-denatured precursor proteins (urea) or sucrose as an osmoticum during chloroplast isolation (Suc). All other conditions were as described for b. (d) The import of preSSU, prePORA, and prePORB can be competed for by preOE33 but not by mOE33. OE33 proteins were heterologously expressed as soluble proteins and added at a final concentration of 5 μM to the import reaction before precursor proteins. Import rates were determined by measuring the signal density of the respective mature radiolabeled protein. Signals of imports with added competitor were normalized to the signals in control imports (without OE33), which were set to 1.0 (arbitrary units). A mean import rate ± SD of three (prePORA and prePORB) and five (preSSU) independent reactions is shown.

Previous studies have used urea-denatured preproteins, because the in vitro translation products were not import-competent (14–17). Therefore, we also used urea-denatured prePORA and prePORB and tested its import in either the presence or absence of Pchlide (Fig. 2c). Although the overall import yield was much lower than under our standard import conditions, again no influence of the absence or presence of Pchlide was detectable. Finally, we isolated chloroplasts with sucrose but not sorbitol as an osmoticum. Sucrose has been used before as an osmoticum to isolate plastids (14–17), although it causes higher osmotic stress because of its membrane permeability compared with impermeable sorbitol. In this case, the import of prePORA and prePORB was slightly stimulated when seedlings were fed with ALA, i.e., in the presence of Pchlide.

So far our data demonstrate that the import of prePORA does not depend on Pchlide b in vitro and in vivo and that it does not depend on the presence of Pchlide a in vitro. To analyze the import characteristics in more detail, we carried out competition experiments in the presence of the precursor or the mature form of the oxygen-evolving complex subunit 33 (OE33), each at 5 μM. The import of both prePORA and prePORB as well as of the stromal marker protein preSSU was effectively competed in the presence of preOE33 but not of mOE33 (Fig. 2d). Import yields were reduced between 75% to 90%, dependent on the preprotein, indicating that these proteins share at least in part a common import route.

In Arabidopsis, four proteins were grouped into the OEP16 family by Reinbothe et al. (16): AtOEP16.1–AtOEP16.4. Detailed phylogenetic analysis including all homologous proteins from vascular plants, however, revealed that the Arabidopsis OEP16 subfamily comprises only three members, namely AtOEP16.1 (At2g28900), AtOEP16.2 (At4g16160), and AtOEP16.4 (At3g62880) (26). The Arabidopsis protein designated as AtOEP16.3 (At2g42210) did not belong to this group (see SI Fig. 7). In addition, AtOEP16.3 (17 kDa) was localized in mitochondria (Fig. 3a) and, thus, was not considered in the present study. AtOEP16.1 shows the highest protein sequence identity (62%) to OEP16 from pea (25) and to HvOEP16-1;1 from barley (52%), which was identified as the import channel of prePORA (16). The protein identity of HvOEP16-1;1 to AtOEP16.2 and AtOEP16.4 is 28% and 21%, respectively.

Fig. 3.

Differential and developmental occurrence of OEP16 isoforms in Arabidopsis. (a) Western blot analysis of the distribution of OEP16 isoforms between chloroplasts (C) and mitochondria (M). Except for the detection of OEP16.2 in chloroplasts of 8-day-old cotyledons, organelles from leaves of 3-week-old plants were used. Antisera against the marker proteins Toc159, OEP37, and VDAC (outer membrane of mitochondria) were used as controls. Numbers indicate the molecular mass of proteins in kDa. (b) The presence of OEP16.1 and OEP16.2 was tested by immunoblot analysis of dry seeds and during germination in 0-, 2-, 4-, 6-, and 8-day-old seedlings. (c) Presence of POR and OEP16.1 proteins in 8-day-old green (lane 1) and etiolated (et; lane 2) cotyledons and during the exposure of etiolated cotyledons to light for the times indicated (lanes 3–6). POR polypeptides are highly identical; therefore, the antibody used recognizes all POR isoforms in Arabidopsis. Please note that OEP16.2 is absent in light-grown and etiolated 8-day-old seedlings (see b, rightmost lane) and therefore was not detectable during the deetiolation of cotyledons.

The proteins AtOEP16.1 (16 kDa) and AtOEP16.2 (20 kDa) were localized to chloroplasts of leaves and cotyledons with the help of specific antibodies raised against each protein (Fig. 3a). Screening the available public microarray databases revealed that OEP16.1 is expressed predominantly in green tissues (AtGenExpress project at http://affymetrix.arabidopsis.info/narrays/experimentbrowse.pl). In contrast, OEP16.2 transcripts are exclusively present in early pollen development, in late stages of embryo and seed development, and during seedling germination (see also Fig. 3b). OEP16.4 is expressed in all tissues but much less so than OEP16.1. So far we were unable to detect the OEP16.4 protein in different tissues or even in enriched envelope membranes from Arabidopsis by immunoblot analysis (data not shown). The expression profile of AtOEP16.1 and AtOEP16.2 was recently supported by ref. 26 and by our own analysis, which shows that the OEP16.1 protein is accumulating during seedling growth and light-dependent greening of cotyledons and thus seems to parallel the development of chloroplasts (Fig. 3 b and c; see SI Fig. 8). In contrast, OEP16.2 is predominantly present in seeds and disappears in germinating seedlings after ≈4 days. Furthermore, OEP16.1 was the only form found in the proteome of Arabidopsis chloroplast envelopes (10, 11).

To estimate the role of OEP16 in the import of prePORA, we isolated knockout mutants in Arabidopsis for OEP16.1 (oep16.1-1), OEP16.2 (oep16.2-1), and OEP16.4 (oep16.4-2) and generated a double knockout mutant for OEP16.1 and OEP16.4 (oep16.1-1/oep16.4-2). As shown by RT-PCR and immunoblots, we detected neither mRNA nor protein in the homozygous T-DNA insertion lines (SI Fig. 9). Seeds from all different homozygous mutant lines germinated with rates between 80% and 90%. They showed neither a delay in greening nor any other morphological phenotype at any stage of development under standard growth conditions (data not shown). Chloroplast development in angiosperm plants is light-dependent because of the light-dependent reduction of Pchlide to Chlide by POR (18, 19). In the dark, etioplasts develop whose internal membrane structure consists of a few prothylakoids and a large prolamellar body instead of thylakoids (32). The prolamellar body consists, to a large extent, of the Pchlide holochrom, comprising POR, Pchlide, and NADPH (33). Etioplasts in homozygous oep16.1-1 and oep16.2-1 (SI Fig. 10) as well as in oep16.1-1/oep16.4-2 mutants (Fig. 4) look similar to wild type. The size and structure of the prolamellar body also is indistinguishable between mutant and wild type. Furthermore, a protein analysis of etioplasts from oep16.1-1, oep16.2-1, and oep16.1-1/oep16.4-2 clearly demonstrated the presence of specific peptides belonging to either PORA or PORB (SI Table 1). In the oep16.1/oep16.4 double mutant, transcript and protein levels of OEP16.2 do not change, as has been determined by quantitative RT-PCR, microarray analysis, and immunoblotting (data not shown). For this reason and because of its expression profile in wild type (see above), OEP16.2 probably cannot replace OEP16.1 and OEP16.4 for the import of prePORA or prePORB, although very low levels of OEP16.2, which are beyond the detectability of the antibody used, could still be present and support some import. We feel that this possibility is highly unlikely, because we never detected any ultrastructural differences or developmental retardation of oep16.1-1/oep16.4-2 double mutants. We conclude that OEP16 function is neither essential in etioplast formation nor in the differentiation of etioplasts to chloroplasts.

Fig. 4.

Etioplast-to-chloroplast transition in Arabidopsis cotyledons. Plastids are shown from etiolated seedlings (Left), after exposure to light for 16 h (Center), or grown under a normal day-night cycle (Right). (a) Col-0 wild-type control. (b) Homozygous oep16.1/16.4 double knockout mutant. (Scale bars: 2.5 μm.)

Thus far, we have been unable to find any influence of OEP16 in plastid biogenesis, as we should have expected if it was crucial for prePORA import. To obtain more direct evidence for the role of OEP16 in the import of either prePORA or prePORB, we isolated intact chloroplasts from the different Arabidopsis oep16 knockout mutants and compared their import properties to chloroplasts from wild-type plants. Neither chloroplasts from the homozygous, single knockout lines oep16.1-1 and oep16.2-1 nor the double knockout line oep16.1-1/oep16.4-2 showed a reduced capacity to import either prePORA or prePORB [Fig. 5a, compare lanes 3–6 (wild type) with lanes 7–10 (mutant)]. Finally, we wanted to know whether the import of prePORA was influenced by the absence or presence of Pchlide in oep16 mutants. In no case was the import of prePORA dependent or stimulated by the presence of endogenous Pchlide (Fig. 5b). As an internal control, the import of prePORB was analyzed, which also was not influenced by the presence or absence of Pchlide. We conclude that OEP16 plays no central role in the import of prePORA in vivo and in vitro.

Fig. 5.

Import of prePORA and prePORB is not impaired in chloroplasts from oep16 knockout mutants. (a) Typical import reactions are shown to compare the import yield of prePORA and prePORB into chloroplasts from Col-0 wild-type (lanes 3–6) or oep16 mutants (lanes 7–10) as indicated. Chloroplasts equivalent to 10 μg of Chl were incubated with the in-vitro-translated precursor proteins (Tl). After import, chloroplasts were either not treated (−) or treated (+) with the protease thermolysin. Precursor and mature proteins are indicated by asterisks and arrowheads, respectively. (b) Import yield of prePORA and prePORB into chloroplasts isolated from light-grown or ALA-treated (dark + ALA) oep16 knockout mutants. All other conditions were as outlined in a and in Fig. 2 b and c.

Discussion

In angiosperm plants, chloroplast development is strictly light-regulated (18, 19). In the dark, etioplasts develop that contain few prothylakoids and a large prolamellar body, in which the Pchlide holochrom accumulates awaiting catalysis of the light-triggered reduction of Pchlide to Chlide (33). After this conversion, the prolamellar body disintegrates rapidly and thylakoids are formed to accommodate the photosynthetic protein–pigment complexes (34). CAO converts Chlide a to Chlide b during Chl b synthesis (30). Although different oxygenases, which could theoretically also catalyze this reaction (35), exist in plastids, our present analysis shows that cao loss-of-function mutants are devoid of detectable amounts of either Pchlide b or Chl b. At the same time, etioplasts from cao-1 mutants have normal-looking prolamellar bodies, contain both PORA and PORB, and differentiate and green normally upon exposure to light. We conclude that the cao-1 line is able to import prePORA and other preproteins normally. Currently, we have no experimental explanation for the previously described Pchlide dependence of prePORA import into isolated plastids.

The outer envelope of plastids contains multiple solute channels whose substrate specificity has not been fully defined (36). Many of these channel proteins are abundant and prominent constituents of the outer membrane. OEP16 represents one of these channels (25). It belongs to a distantly related family of protein and amino acid transporters called PRAT (37). PRAT proteins contain four α-helical transmembrane regions and are further characterized by a loosely conserved amino acid motif in the hydrophobic region. Other members of the family include the mitochondrial protein translocon subunits Tim17 and Tim22. It is therefore tempting to speculate that OEP16 is also involved in protein translocation (14–16). So far, our results from using in vitro reconstituted OEP16 indicate the permeation of amino acids but not of other solutes like sugars or triosephosphates (25). In vivo, oep16 knockout mutants show no ultrastructural changes at the etioplast or chloroplast stage or a retardation in the light-dependent organellar transition phase. Furthermore, etioplasts from cotyledons of oep16 knockout lines contain PORA and Pchlide a, clearly demonstrating that OEP16 has no vital role in the import of prePORA in vivo. But how can we explain the results that show that prePORA can be cross-linked to OEP16 in vitro in isolated chloroplasts and therefore seemingly indicate the role of OEP16 in translocation? In vitro binding of prePORA to OEP16 might be nonspecific and nonproductive, but it occurs because of a nonselective interaction between both polypeptides, which in vivo could be inhibited by cytosolic proteins proposed to play a role in the early events of maintaining import compatibility (3). This notion is, in our opinion, supported by the experimental design followed by Reinbothe and coworkers (16), who used urea-denatured preproteins for the import and cross-linking studies. Upon rapid dilution of the denatured and unfolded polypeptides into the import reaction, the protein is prone for aggregation and nonnative folding routes that, in turn, favor nonspecific interactions (38). When a native precursor is used, as in our study, prePORA imports with similar yields into chloroplasts from oep16 knockout mutants as from wild-type plants, independent of the presence or absence of Pchlide. How Pchlide influences the accumulation of PORA in plastids in vivo was not studied here, but see ref. 12.

In summary, we could not detect any role of OEP16 in import reaction of prePORA in vivo and in vitro and conclude that other results are untenable. Furthermore, the accompanying implications and consequences for the light-dependent greening process are not relevant.

Materials and Methods

Transcription and Translation.

The coding regions for prePORA (At5g54190) and prePORB (At4g27440) were PCR-amplified and EcoRI/XbaI-subcloned into the vector pSP65 (Promega, Madison, WI). The following oligonucleotide primers were used for amplification: prePORA, 5′-ccggaattcatggcccttcaagctg-3′ (forward) and 5′-ctagtctagattaggccaagcctacg-3′ (reverse); PORB, 5′-ccggaattcatggcccttcaagctg-3′ (forward) and 5′-ctagtctagattaggccaagcccacg-3′ (reverse) on cDNA synthesized from etiolated and light-grown Col-0 seedlings, respectively. All constructs were controlled by DNA sequencing. The pSSU control was described previously (39). Transcription was performed in the presence of SP6 RNA polymerase, and the resulting mRNA was translated in a wheat germ system (wheat germ extract; Promega) in the presence of [³⁵S]methionine (39) at 25°C for 45 min. The translation mixture was centrifuged at 50,000 × g at 4°C for 20 min, and the supernatant was used for all import studies.

Plant Material and Growth Conditions.

All experiments were performed on A. thaliana (L.) Heynh. Columbia plants (cv. Col-0; Lehle Seeds, Round Rock, TX) or the respective mutant plants. Before sowing, seeds were surface-sterilized with 5% hypochloride. To synchronize germination, all seeds were kept at 4°C for 3 days. Plants were grown on soil or on 0.3% Gelrite medium (Serva, Heidelberg, Germany), containing 1% d-sucrose and 0.5× Murashige and Skoog salts at pH 5.7. Plant growth occurred in growth chambers with a 16-h light (21°C, photon-flux density of 100 μmol·m⁻²·sec⁻¹) and 8-h dark (16°C) cycle.

Chloroplast Isolation and Protein Import.

Chloroplasts for import assays were isolated from 8-day-old cotyledons. Therefore, 0.2 g of seeds was grown on Murashige and Skoog agar plates. Etiolated and ALA-treated seedlings were grown in darkness under the same temperature regime as light-grown plants. To induce Pchlide synthesis, seedlings were sprayed with 1 mM ALA in 20 mM Tris, pH 8.0, or in Tris buffer as a control. Seedlings were then vacuum-infiltrated for 3 min and incubated for 16 h in darkness. Harvesting, plastid isolation, and import into plastids of etiolated and ALA-treated cotyledons was performed under dim green safe light. All chloroplasts were isolated 3 h after onset of the day period as described in ref. 40.

Import reactions were carried out in HMS buffer [20 mM gluconic acid/50 mM ascorbic acid/10 mM NaHCO₃/0.2% (wt/vol) BSA/20 mM Hepes·KOH, pH 7.6/4 mM MgCl₂/300 mM sorbitol]. Import reactions with a volume of 200 μl contained chloroplasts equivalent to 20 μg of Chl, 3 mM ATP, 10 mM methionine, and the translation products, which did not exceed 10% of the total reaction volume. To import denatured precursors, translation products were precipitated in NH₄SO₄ and denatured in 8 M urea. The final urea concentration did not exceed 150 mM in the import reaction. Import reactions were carried out at 25°C for 10 min. Afterward, one half of the reaction was treated with 100 μg/ml thermolysin on ice for 30 min. All import reactions were stopped by adding HMS buffer containing 50 mM EDTA, followed by a short centrifugation burst to pellet the chloroplasts, and separated by SDS/PAGE. Import was documented by autoradiography and quantified by a FLA-3000 phosphoimager (FujiFilm, Tokyo, Japan) and AIDA software (Raytest Isotopenmessgeräte, Straubenhardt, Germany).

Protein Extraction.

Plant material was homogenized in liquid nitrogen, and proteins were extracted in 50 mM Tris·HCl, pH 8/50 mM EDTA/2% LDS/10 mM DTT/100 μM PMSF on ice for 30 min. Cell debris was pelleted at 4°C for 15 min by centrifugation at 14,000 × g. Intact mitochondria were isolated as described in ref. 41. Appropriate amounts of organellar or total cellular proteins were resolved by SDS/PAGE followed by immunoblot analysis using antibodies in 1:200 to 1:1,000 dilutions. All antisera were raised in rabbit against oligopeptides for OEP16.3 (⁴⁴PRVERNVALPGLIRT⁵⁸ and ¹³⁸TRVDNGREYYPYTVEKRAE¹⁵⁶) or heterologously expressed proteins from Arabidopsis (OEP16.1, OEP16.2, OEP37, and PORB) and pea (Toc159 and VDAC) (Pineda Antibody Service, Berlin, Germany).

Pigment Analysis.

Plant tissue (30–120 mg of fresh weight) was pulverized in liquid nitrogen and extracted twice with 500 μl of acetone:water (4:1, vol/vol). After centrifugation, the liquid phases were collected in a 2-ml test tube. The extraction was repeated until the pellet was colorless. For detection of Pchlide, the combined acetone extracts were mixed with 500 μl of n-hexane. The n-hexane phase, which contained most of the Chl, was separated. The acetone phase was extracted several times with ethylacetate, and the combined ethylacetate phases were washed with distilled water. The dried ethylacetate phase was diluted in 50 μl of acetone and applied for HPLC. The HPLC system used for analytical separations consisted of a Waters 600E pump and controller (Waters, Eschborn, Germany) and a diode array detector (Tidas; J&M, Aalen, Germany). Pigments were separated on a column (4 × 250 mm) filled with RP18 (Gromsil 120; Grom Analytic, Herrenberg, Germany). The flow rate was 1.2 ml/min with the following gradient consisting of 60% acetone:40% water, pH 3.5 (HAc) (solvent A), and of 100% acetone (solvent B):100% solvent A for 3 min, followed by 100% solvent B for 20 min and 100% solvent B for 10 min.

Abbreviations

Pchlide

protochlorophyllide

prePOR

precursor Pchlide oxidoreductase

CAO

chlorophyllide a oxygenase

Chlide

chlorophyllide

Chl

chlorophyll

ALA

δ-aminolevulinic acid.