Biogenic membranes of the chloroplast in Chlamydomonas reinhardtii

Subcellular Fractionation Reveals Chloroplast Translation Membranes.

To study where proteins are synthesized in the chloroplast, we used analytical subcellular fractionation to determine whether the membranes bound by the chloroplast translation machinery have the density of thylakoid membranes, the chloroplast envelope membrane, or an unknown membrane type (12). C. reinhardtii cells were broken by French Pressure Cell because other breakage methods leave unbroken cells that contaminate the membranes of interest in subsequent steps. Using isopycnic ultracentrifugation, membranes were floated from a 2.5 M sucrose cushion into a 0.5–2.2 M sucrose concentration gradient, where they banded according to density. Nonmembrane material either remained in the 2.5 M sucrose cushion or pelleted. Most previous studies of ribosome-bound chloroplast membranes used discontinuous sucrose density gradients to isolate bulk membranes in broad density ranges. Chloroplast envelope membranes are isolated in the density range of 0.4–1 M sucrose (13). Thylakoid membranes are considered to be the densest membrane of the chloroplast and, therefore, are collected as bulk dense membranes in the range of 1–2 M sucrose (14, 15). As rough ER membranes are denser than thylakoid membranes (11), it seemed plausible that previous studies inadvertently analyzed an analogous ribosome-bound CTM with thylakoids (10). Therefore, to separate membranes in this density range with high resolution, we used continuous sucrose gradients with a high maximal concentration (2.2 M). Gradient fractions were analyzed by immunoblotting to determine the density of membranes associated with markers of the chloroplast translation machinery and known chloroplast compartments (Fig. 1). Gel lanes had the same proportions of fractions so that the amount of a marker would reflect the proportion of its total cellular pool (12). In other words, samples were not normalized on the basis of mass amounts of protein because this would drastically overrepresent markers, on a per cell basis, in fractions with the least amount of protein, and vise versa.

Our major finding was that thylakoid membranes can be resolved from denser membranes that are associated with markers of the chloroplast translation machinery and the T-zone. As seen for the three experimental trials of our subcellular fractionation scheme (Fig. 1), the fractions with thylakoid membranes could be identified by their enrichment in chlorophyll and the subunits of PSI and PSII, PsaAp, and D2, respectively (Fig. 1A, lanes 7–10; Fig. 1B, lanes 7–10; and Fig. 1C, lanes 6–9). Envelope membranes are less dense than thylakoids and should be in lanes 4–6 of Fig. 1 (see below) (13). Interestingly, denser membrane fractions had substantial proportions of the total pools of chloroplast ribosomal (r)-proteins and RBP40, and yet they had minor amounts of thylakoid membranes (Fig. 1A, lanes 11–13; Fig. 1 B and C, lanes 10–12). As predicted, these dense membranes had similar density to the canonical ribosome-bound membrane of the rough ER revealed by an r-protein of the 60S subunit of the cytoplasmic ribosome (Fig. 1C, lanes 6–12). In contrast, CTM were distinct from stroma-exposed thylakoid membranes, the accepted site of PSII subunit synthesis (Fig. 1 A and B, PsaAp). These results provided evidence of a CTM privileged for the synthesis and membrane insertion of PSII subunits and localized in the T-zone.

In PSII assembly, newly synthesized subunits associate to form subcomplexes, which then associate to form the monomeric PSII complex RCC1 (reaction centre core 1) (16). In an attempt to identify membranes associated with the assembly of chloroplast-encoded subunits, fractions were immuno-probed with an antiserum against a PSII assembly factor in Synechocystis sp. PCC 6803, YCF48, the homolog of HCF136 of Arabidopsis thaliana (Fig. 1A) (17). This antiserum detected a protein of the expected molecular mass (Fig. S1). This putative YCF48/HCF136 was detected in a broad membrane density range (Fig. 1A, lanes 7–13) but not in the nonmembrane material (Fig. 1A, lanes S and 16). Notably, Fraction 7 contained YCF48/HCF136 but had little thylakoid membrane and no detectable CTM (Fig. 1A, lane 7). This result suggests that this fraction contains yet another unique biogenic membrane, one involved in PSII assembly. Although these results are preliminary, they are consistent with the general theme here: that the C. reinhardtii chloroplast may have diverse biogenic membranes.

CTM should be physically bound by chloroplast ribosomes. Alternatively, a ribosome-associated membrane could be generated artifactually during cell breakage if free chloroplast ribosome subunits and RBP40 in the chloroplast stroma become trapped within vesicles that form by fragmenting membranes. Detracting from this possibility, however, is that a marker protein for the chloroplast stroma, HSP70B, was not in CTM fractions, but trace amounts were detected in the thylakoid fractions in lanes 10–12 and 7–9, respectively, of Fig. 1C. To more directly address whether chloroplast translation marker proteins are bound to the CTM, we asked whether they can be extracted by agents that remove peripheral membrane proteins. Membranes of fraction 10 in Fig. 1B were washed with one of the following: 500 mM KCl, 20 mM NaCO₃, 1.0 M NaCl, or 2.0 M urea or, as a negative control, without agent. Supernatant and pellet fractions were analyzed by immunoblot to determine the degree of extraction of marker proteins for chloroplast ribosome subunits and RBP40 (8). To ensure pelleting of the membrane, we followed the low amount of thylakoid membrane in this fraction by immuno-probing for D2. The results revealed that RBP40 and the 50S subunit were extracted by each of the agents, either partially or completely (Fig. 2). Therefore, these translation components are peripherally bound to CTM. The 30S subunit r-protein was only extracted by high ionic strength (2.0 M NaCl) and only partially. Although this result alone is consistent with either of the two possibilities outlined above, the 30S ribosome subunit is probably bound to CTM because it seems improbable that it would be trapped in vesicles when most 50S subunits and RBP40 are not. Therefore, the 30S subunit is probably bound to CTM with particularly high affinity.

Blue-Native PAGE and Immunoblot Analyses Reveal Markers of PSII Biogenesis.

Newly synthesized PSII subunits assemble in specific combinations to form precomplexes, which then associate to form the monomeric PSII complex RCC1 (16). We reasoned that unassembled subunits could serve as a marker for a CTM, and precomplexes for specific steps in PSII assembly. Therefore, membrane fractions were compared for the assembly states of the chloroplast-encoded PSII subunits D1 and D2 using blue-native (BN) PAGE and immunoblot analysis (18). Analyses of equal amounts of membrane ensured comparable solubilisation conditions, which can affect quaternary structure artifactually (19). Because this process necessitated overrepresentation of CTM on a per cell basis, the amounts of sample were normalized to the level of the monomeric PSII complex, RCC1. In other words, we asked whether CTM are qualitatively different from thylakoid membranes in ways that support their having a role in PSII subunit synthesis and assembly. The results revealed RCC1 at constant levels across the lanes, confirming proper normalization (Fig. 3A). The higher mobility complex is RC47, the PSII monomer lacking CP43, which is generated primarily during PSII repair (20). Notably, the dimeric PSII, RCC2, was detected in the thylakoid membrane fractions but not in CTM fractions (Fig. 3A, compare lanes 1–2 with 4–6). This result suggests that thylakoid membranes are the primary location of RCC1 dimerization to form RCC2, a late step in PSII biogenesis (16).

With 1D BN-PAGE we were unable to detect free subunits and subcomplexes for use as CTM markers, possibly because of ill effects of the detergent on detection below 100 kDa (19). Therefore, BN gel lanes with either thylakoid membranes or CTM were subjected to a second dimension of denaturing SDS/PAGE and analyzed by immunoblotting. We normalized based on the relative levels of RCC1 in these samples determined by 1D BN-PAGE (Fig. 3A). (RCC2 was not detected for unknown reasons on the 2D gel immunoblot analyses.)

The results revealed that D1 is present in RCC1 and RC47, as well as in an early assembly intermediate subcomplex, the PSII reaction center, and as a free unassembled subunit. All were detected in both thylakoid and CTM samples (Fig. 3 B and C). Unassembled D1 could not serve as a marker for CTM because it is associated with both the repair and de novo biogenesis of PSII. Nevertheless, this result revealed that the assembly step in which the reaction center forms RCC1 does not occur preferentially in CTM over thylakoid membranes. Furthermore, the finding that unassembled D1 can be detected on both immunoblots serves as a positive control for the subsequent experiments.

These blots were immunoprobed for D2, a PSII subunit the synthesis of which is not induced for PSII repair. Therefore, unassembled D2 can serve as a marker for PSII biogenesis. D2 was detected in RCC1, RC47, and the reaction center (RC) in both thylakoids and CTM (Fig. 3 D and E). Most notably, however, unassembled D2 was detected only in the CTM sample (Fig. 3E). These results suggest that CTM is a privileged location of the synthesis of the plastid genome-encoded subunits for de novo assembly of PSII.

Markers of PSI Assembly Cofractionate with Thylakoid Membranes.

To determine whether CTM has a role in PSI biogenesis, we immuno-probed the 2D blots for PsaAp (Fig. 3 F and G). Although PsaAp was not detected unassembled, it was detected in the PSI monomeric complex, the PSI monomeric complex lacking PsaK and PsaG, and a larger unidentified complex of approximately 550 kDa, possibly the PSI dimer (21). Notably, the PSI monomer lacking PsaK/G was more abundant in the thylakoid membrane fraction than in the CTM fraction (Fig. 3 F and G). This result and previously reported evidence that this complex is a late intermediate in PSI assembly (21), suggest that later steps in PSI assembly occur primarily in thylakoid membranes. Earlier steps in PSI assembly may also occur in thylakoids because we detected an early PSI assembly factor, YCF4, only in thylakoid membrane fractions (Fig. 1C, compare fractions 7–9 and fractions 10–12) (22). Finally, the mRNA encoding PsaAp was not recruited to the T-zone under the same conditions that recruited two PSII subunit mRNAs (6). Taken together, these results suggest that PSI subunit synthesis and assembly occur at thylakoids, and not CTM.

Envelope Membranes with the TOC-TIC Translocons Have Higher than Expected Density.

The gradient fractions were also tested for the envelope markers, Toc75 and Tic110, subunits of the TOC and TIC protein import complexes. Instead of finding these proteins in the density range of the envelope membranes (Fig. 1, fractions 4–6) (13), they were detected with thylakoid fractions and, in certain preparations, also with denser CTM (Fig. 1 B and C). Although the basis of this unexpectedly high and variable density of envelope membranes with the TIC and TOC complexes is unknown, their occasional separation from CTM reveals these are distinct membrane types.

Chloroplast Protein Import Machinery Localizes to Unique Envelope Domains.

To explore the basis of the unexpectedly high density of envelope membranes with Toc75 and Tic110, the in situ distribution of these proteins was characterized by immunofluorescence (IF) staining and epifluorescence microscopy. All cells were costained for the chloroplast psbA mRNA by FISH to reveal the T-zone with strong signal localized around the pyrenoid and to stain the chloroplast with weaker diffuse signal. For a description of chloroplast anatomy, see Fig. 4A (6).

A striking pattern was observed in many cells in which the Toc75 or Tic110 IF signal localized around lobes specifically where they adjoined the basal region (Fig. 4 B and D). We named these sites “lobe junctions” (Fig. 4A). In some cases, the lobe at such a lobe junction could be seen to form a hole in the cloud of IF signal, indicating that the envelope surrounding it was enriched in the TOC-TIC protein import machinery (Fig. 4B). Examples of the cells that did not show this localization pattern are shown in Fig. 4 C and E. Of the cells examined from moderate light growth condition (ML cells), 48% showed the Toc75 IF signal localized around one or two lobe junctions (Fig. 4B) (n = 188). Similarly, of the ML cells IF-stained for Tic110, 47% showed this pattern (Fig. 4D) (n = 199). This pattern is interesting for three reasons. First, it is specific to import machinery because it was not observed for many other chloroplast proteins, the localization of which we have examined with this method (6, 7, 23). Second, this localization pattern may be physiologically relevant because the percentage of cells showing it dropped during incubation in the dark for 2 h, a condition associated with reduced rates of PSII biogenesis and chloroplast protein import in C. reinhardtii (6, 24). When ML cells were dark-adapted immediately before fixation, the percentages showing localization around lobe junctions dropped from 48 to 11% for Toc75 (n = 74) and from 47 to 15% for Tic110 (n = 52). Finally, Toc75 or Tic110 localization at lobe junction was probably present but undetected in many cells. C. reinhardtii cells are polarized and must be oriented longitudinally in the microscopy field to reveal cytological landmarks necessary to locate lobe junctions (e.g., the cytosol and pyrenoid) (Fig. 4A). Moreover, no protein marker exists for which co–IF-staining can reveal lobe junctions in a particular cell.

These results suggest that lobe junctions have special envelope domains enriched in the TOC-TIC protein import machinery. In light of these results, the unexpectedly high and variable density of envelope membranes with Toc75 and Tic110 could be explained if these import envelope domains have higher density than previously described envelope membranes, and had formed to different degrees in the various cultures used for subcellular fractionation.

Culture Conditions.

C. reinhardtii strains CC-4051 or CC-503 were cultured photoautotrophically in high-salt minimal medium with aeration at 24 °C, under a light intensity of approximately 100 µE⋅m⁻²⋅sec⁻¹ until midlog phase (2–4 × 10⁶ cells·ml⁻¹) (33).

Analytical Subcellular Fractionation.

Cells from a 500-mL culture were pelleted by centrifugation at 4,000 × g for 5 min at 4 °C, resuspended in MKT-buffer [25 mM MgCl₂, 20 mM KCl, 10 mM Tricine-Cl pH 7.5, protease inhibitor (Sigma-Aldrich)]. Cells were broken by three passes through an ice-chilled French Pressure Cell at 1,000 psi. Breakage was verified by light microscopy (400× and 1,000× magnification). The lysate was ultracentrifuged at 100,000 × g for 1 h at 4 °C. The supernatant was collected and stored at −80 °C. The pellet was resuspended in 2.5 M sucrose and overlaid with a linear sucrose gradient (0.5–2.2 M). All sucrose solutions were prepared in MKT-buffer. The gradient was ultracentrifuged at 100,000 × g for 16 h at 4 °C. Fractions (0.75 mL) were collected and the pellet was resuspended in KHEG-Buffer [60 mM KCl, 20 mM Hepes, 0.2 mM EDTA, 20% (vol/vol) glycerol]. Gradients contained only membrane and associated material based on the buoyant density of bacterial ribosomes in equilibrium CsCl gradient ultracentrifugation (1.67–1.69 g/mL) would be equivalent to 4.9 M sucrose (34).

Quantification of Protein and Chlorophyll.

Protein concentration was determined using the bicinchoninic acid assay (35). Chlorophyll was quantified as described previously (36).

Immunoblot Analysis.

Equal proportions of the fractions were solubilized in SDS/PAGE loading buffer, denatured at 42 °C for 30 min. SDS/PAGE and immunoblot analyses were performed as described previously (37). The antisera were: αD1 (Agrisera), αS-20 (30S r-protein), αL-30 (50S r-protein) and αcyL4 (60S r-protein) (38, 39), αPsaAp (40), αHSP70B (41), and αRBP40 (42).

FISH, IF-Staining, and Fluorescence Microscopy.

FISH and IF-staining of cells of strain CC-503 were as described previously (6, 43). The psbA FISH probes were labeled with Alexa Fluor 488 and the IF-staining used Alexa Fluor 568-conjugated anti-rabbit secondary antibody (Invitrogen). Images were captured on a Leica DMI6000B microscope (Leica Microsystems) using a 40×/0.75 objective, a Hamamatsu OrcaR2 camera and Volocity acquisition software (Perkin-Elmer).

BN PAGE.

BN-PAGE was performed as described previously with the following minor modifications (18, 44). Aliquots of sucrose gradient fractions containing 6 μg of chlorophyll were concentrated by centrifugation (100,000 × g; 1 h; 4 °C) and resuspended in ACA 750 (750 mM aminocaproic acid, 50 mM Bis•Tris, and 0.5 mM EDTA, pH 7.0). Membranes were then solubilized on ice in 0.8% n-Dodecyl-β-d-Maltoside (β-DM) for 5 min. Samples were centrifuged at 17,000 × g for 30 min at 4 °C. The supernatant was added to 1/10 Vol of 5% (wt/vol) Coomassie Brilliant blue G-250, 750 mM aminocaproic acid whereupon protein complexes were then separated by electrophoresis in a 4.5–12% (wt/vol) acrylamide BN gel. To ensure that D2 signal on different 2D gels was normalized to the level of RCC1, comparable amounts of RCC1 were loaded, as determined by results from 1D BN gels, and all steps were carried out in parallel. Results of maximal ECL exposure times are shown for both.

Membrane Washing.

Aliquots of fraction 10 in Fig. 1B were diluted 25-fold in washing buffer [20 mM KCl, 10 mM Tricine and 2.0 mM EDTA pH 7.2, protease inhibitor mixture (Sigma-Aldrich)] and pelleted by centrifugation in a microfuge for 1 h at 17,000 × g at 4 °C. Pellets were resuspended in 30 µL of one of the following: washing buffer, 500 mM KCl, 20 mM NaCO₃, 1.0 M NaCl, 2.0 M urea, incubated on ice for 30 min, and then subjected to the same centrifugation step. The supernatants were collected and the pellet was washed once and then resuspended in 30 µL SDS/PAGE sample buffer. SDS/PAGE and immunoblot analysis were as described previously (37).