The YlmG protein has a conserved function related to the distribution of nucleoids in chloroplasts and cyanobacteria

Isolation of A. thalianachloroplast division mutants from FOX lines

Several proteins required for chloroplast division have been identified and characterized by both forward and reverse genetics. These studies yielded the identification of FtsZ [2], MinD [4], MinE [5], DRP5B/ARC5 [6, 7], ARC3 [24], ARC6 [3], PDV1/PDV2 [8], and MCD1 [9].

Although analyses by using conventional loss-of-function mutants have contributed to the identification of these chloroplast division proteins, certain key chloroplast division genes may as yet still be uncovered, e.g. because of the lethality of a given mutation or the functional redundancy provided by duplicate genes. In order to identify any such unidentified chloroplast division genes by an alternative approach, we searched for genes that affect on chloroplast division when the genes are overexpressed.

To this end, we screened ~15,000 T2 plants of the full-length cDNA overexpresser (FOX) gene-hunting lines of A. thaliana [22] by microscopic observation of mesophyll cell chloroplasts. As a result, we isolated 18 mutant lines that contained a smaller number and larger size of chloroplasts than the wild type. Of these mutants, two lines (FN026 and FN028) contained cDNA of the same gene (At3g07430) downstream of the cauliflower mosaic virus (CaMV) 35S promoter (Figure 1A). All the F1 progeny, after crossing FN026 and FN028 with the wild type, displayed the mutant phenotype, indicating that the phenotype occurs in a dominant manner. When At3g07430 was overexpressed in the wild-type plants by a newly constructed 35S-At3g07430 transgene, the resulting plant contained chloroplasts significantly larger than the wild type (two times on average, Figure 1B). Reverse transcriptase-polymerase chain reaction (RT-PCR) analyses confirmed that the level of At3g07430 transcript was increased in the FN026, FN028, and 35S-At3g07430 lines (Figure 1C). In order to examine the protein level of At3g07430, we prepared antibodies against At3g07430. On immunoblots, the antibodies detected a band of ~20 kDa, which is consistent with the predicted size (23 kDa) of the At3g07430 protein [the transit peptide was predicted by the TargetP http://www.cbs.dtu.dk/services/TargetP/ and the predicted transit peptide was omitted for calculation of the molecular mass](Figure 1D). Immunoblot analyses confirmed that At3g07430 protein was over-produced in the FN028 and 35S-At3g07430 lines and showed that the At3g07430 protein level of the 35S-At3g07430 line was higher than that of the FN028 line (Figure 1D). Therefore there is a correlation between the level of At3g07430 protein (Figure 1D) and the chloroplast size (Figure 1A and 1B). In contrast to the protein level, the At3g07430 transcript level of the FN028 line was higher than that of the 35S-At3g07430 line. This is likely due to the difference between inserted transgenes. The transgene in the FN028 line contains full-length At3g07430 cDNA whereas the 35S-At3g07430 line contains no 5'-UTR. The presence or absence of the 5'-UTR probably affects on the efficiency of the translation of At3g07430 protein.

Although the function of the At3g07430 protein has not been determined, the database [The Arabidopsis Information Resource (TAIR); http://www.arabidopsis.org/] indicates that a homozygous T-DNA insertion mutant of this gene (CS24080) resulted in an embryonically defective phenotype. Because a BLAST search showed that At3g07430 is homologous to the bacterial YlmG protein which is of unknown function (for details, see below), we named the gene AtYLMG1-1.

In order to explore how chloroplast division is impaired in the AtYLMG1-1 overexpresser, we compared the localization of FtsZ in the overexpresser and the wild type. Immunofluorescence microscopy using an anti-FtsZ2-1 antibody [9] showed FtsZ localization at the chloroplast division site in the wild type (Figure 1E). In contrast, the localization of FtsZ was perturbed in the overexpresser, with fragmented filaments, small rings, and dots observed in almost all of the chloroplasts (Figure 1E). Therefore overexpression of AtYLMG1-1 perturbs the FtsZ ring formation and consequently impairs chloroplast division.

Phylogenetic relationships in the YlmG family

BLAST and PSI BLAST searches along with sequence alignment indicated that AtYLMG1-1 is homologous to the bacterial YlmG proteins and the chloroplast-encoded Ycf19 of unknown function (Figure 2A; see also Additional file 1). The YlmG protein contains a putative membrane spanning YGGT domain (according to the name of E. coli gene yggT of unknown function). In addition, the searches showed that YlmG-related sequences are conserved in and specific to bacteria and plastid-carrying eukaryotes. In the genome of A. thaliana, four homologs of YlmG were identified (At3g07430, At4g27990, At5g21920, and At5g36120).

In the genomes of gram-positive bacteria, the ylmG gene locates downstream of the cell division gene ftsZ in the dcw (division and cell wall) cluster in the order ylmD, ylmE, ylmF, ylmG, ylmH, and divIVA [23]. Of these genes, ylmF (sepF) [25–27] and divIVA [28] have been shown to be involved in cell division in several bacterial lineages. These results raise the possibility that the YlmG protein is also involved in cell division, although inactivation of ylmG in Streptococcus pneumoniae did not in fact result in any apparent effect on cell division [23]. On the other hand, the Chlamydomonas reinhardtii YlmG homolog, CCB3, has been implicated in cytochrome b₆ maturation, based on the result that CCB3 complemented the defect in cytochrome b₆ maturation in ccb mutants [29]. In addition, disruption of the ortholog in Synechocystis sp. PCC6803 impaired photosynthetic activity [30].

In order to clarify the relationship of YlmG homologs in bacteria and eukaryotes, we performed phylogenetic analyses (Figure 2B). The analyses indicate that oxygenic-photosynthetic organisms (i.e. cyanobacteria and chloroplast-carrying eukaryotes) have two distinct families (Group I and II) with high support values (local bootstrap value by the maximum likelihood method/posterior probability by Bayesian inference, 92/1.00 for group I, 94/1.00 for Group II). Group II contains only proteins of oxygenic photosynthetic organisms, consistent with the reports suggesting a relationship between CCB3 and photosynthesis. In contrast, group I, to which AtYLMG1-1 belongs, contains proteins of apicomplexan parasites, which have non-photosynthetic plastids (apicoplasts) acquired by a red algal secondary endosymbiosis [31]. This pattern of gene distribution suggest that both group I and II of cyanobacterial ylmG were transferred to the nuclear genomes of plants by a primary endosymbiosis of the chloroplast, and that group II, but not group I, was lost in parallel with the loss of photosynthetic activity in the ancestor of apicomplexans. Given this scenario and the existence of ylmG in non-photosynthetic bacteria, it is suggested that group I (including AtYLMG1-1) and bacterial YlmG (other than cyanobacterial group II) are not related to photosynthesis. Supporting this suggestion, recent studies have shown that the inactivation of the members of group I (A. thaliana, At4g27990 and At5g21920; Synechocystis sp. PCC 6803, ssr2142) had no effect on the photosynthesis [30, 32].

The phylogenetic analyses categorized four Arabidopsis YLMGs into the Group I and II and showed a very close relationship between AtYLMG1-1 and At4g27990. We therefore named At4g27990 AtYLMG1-2, At5g21920 AtYLMG2, and At5g36120 AtYLMG3, respectively.

The relationship between AtYLMG1-1 and nucleoid structure

The chloroplast division defect observed above was caused by AtYLMG1-1 overexpression, and we were not able to obtain a homozygote of the AtYLMG1-1 T-DNA insertion mutant (CS24080) as reported in the TAIR database. To further examine the function of AtYLMG1-1, we expressed the antisense RNA in the wild-type plant to knockdown AtYLMG1-1 (Figure 3A). Immunoblot analysis confirmed that the AtYLMG1-1 protein was hardly detectable in two antisense lines (Figure 3B). The result further confirmed that the band detected by the antibodies is AtYLMG1-1 protein and showed that the antibodies do not cross-react with AtYLMG1-2 protein. RT-PCR analyses confirmed that there was a decrease of the AtYLMG1-1 transcript level in the antisense line and no effect on the accumulation of other YLMG transcripts (Figure 3C). In the antisense line, young emerging leaves and the basal part of expanding leaves exhibited a pale-green phenotype. As leaves matured, the leaf color shifted to green, with no obvious difference compared to the wild-type leaves (Figure 3A). These phenotypes were not observed in the AtYLMG1-1 overexpresser (Figure 1A).

Then the morphology of chloroplasts in the AtYLMG1-1 antisense line was observed under microscopy (Figure 3D). In contrast to the overexpresser, the shape and size of chloroplasts in expanding leaves were similar to those in the wild type. In the basal part of expanding leaf of the antisense line, chloroplasts were pale and smaller than those of the wild type (Figure 3D). We compared the localization of FtsZ in the antisense line and the wild type (Figure 3E). In the antisense line, FtsZ localization was observed at the chloroplast division site, as in the wild type. These results suggest that the knockdown of AtYLMG1-1 had no effect on chloroplast division.

Although the knockdown did not impair envelope division, the existence of ylmG in the dcw cluster of gram-positive bacteria suggests the gene product may be involved in bacterial division. To examine whether AtYLMG1-1 is required for a process other than envelope fission that is related to chloroplast division, we observed chloroplast nucleoids in the antisense line and the wild type. By 4', 6-diamidino-2-phenylindole (DAPI) staining, nucleoids were observed as small particles dispersed in mature chloroplasts of the wild type (Figure 4A). In contrast, nucleoids were concentrated in a few large structures in both the tip and basal part of the expanding leaves of the antisense line. When the AtYLMG1-1 overexpresser was examined, the nucleoids were observed as irregular networks. These networks of nucleoids are similar to those in dividing chloroplasts, although the fluorescent intensity by DAPI-staining was higher in the overexpresser than in the wild type (Figure 4B, [16]). In both the antisense line and the overexpresser, DNA gel blot analyses showed that the amount of chloroplast DNA compared to the nuclear DNA is similar to that of the wild type (Figure 4C). These results suggest that knockdown or overexpression of AtYLMG1-1 does not affect the replication of chloroplast DNA, but does affect the morphology of nucleoids.

To examine whether abnormal structure of nucleoids causes the chloroplast division defect in the AtYLMG1-1 overexpresser or chloroplast division defects result in the abnormal nucleoids, we observed nucleoids in bona fide chloroplast division (envelope division) mutants. In contrast to the AtYLMG1-1 overexpresser, the morphology of chloroplast nucleoids in ftsZ2-1, arc5, and arc6 mutants was similar to the wild type (Figure 4D). Taken together, the above results suggest that AtYLMG1-1 is required for the proper distribution of nucleoids in chloroplasts. It is also suggested that abnormality of the nucleoid structure is not due to a chloroplast division defect, but rather, the abnormal nucleoids induced by AtYLMG1-1 overexpression might be a cause of the chloroplast division defect.

Localization of AtYLMG1-1

In order to obtain insight into whether AtYLMG1-1 directly affects the distribution of nucleoids, we examined the localization of AtYLMG1-1. Immunoblot analyses showed that AtYLMG1-1 was enriched in the isolated chloroplasts as compared with the whole plant protein (Figure 5A). When the chloroplasts were lysed in hypotonic solution, AtYLMG1-1 was detected in the membrane fraction (pellet), as was the membrane protein TOC34, suggesting that AtYLMG1-1 is a chloroplast membrane protein (Figure 5B), as predicted in the database ARAMEMNON http://aramemnon.uni-koeln.de/. Further fractionation showed that AtYLMG1-1 is exclusively associated with the thylakoid membranes, as is Lhcb1 (Figure 5C).

We further examined the intrachloroplast localization of AtYLMG1-1 by immunofluorescence microscopy using AtYLMG1-1 antibodies. The fluorescent signals were detected on the punctate structures dispersed in chloroplasts of the wild-type leaves (Figure 5D). These results, together with the results of the immunoblotting, indicate that AtYLMG1-1 localizes in the puncta on thylakoid membranes. Comparison of the immunofluorescence and the DAPI fluorescence showed that some of the AtYLMG1-1 puncta co-localize with a subset of nucleoids (Figure 5E).

Effect of overexpression and gene disruption of ylmG in the cyanobacterium S. elongatus

Knockdown of AtYLMG1-1 had no effect on chloroplast division, but the lack of an evident chloroplast division phenotype might be due to the existence of two other genes related to AtYLMG1-1 (AtYLMG1-2 and AtYLMG2, shown in Figure 2B). Our phylogenetic analyses indicated that cyanobacterial species have only single genes encoding group I and group II YlmG proteins, respectively. Therefore, in order to examine whether the group I ylmG is involved in bacterial cell division, and whether the function of ylmG is conserved between chloroplasts and cyanobacteria, we examined the effects of group I ylmG (ORF ID; Synpcc7942_0477, SylmG1) disruption and overexpression in Synechococcus elongatus. We disrupted the SylmG1 gene by homologous recombination and insertion of a kanamycin-resistant gene into the SylmG1 locus (Figure 6A). Because cyanobacteria have multiple genomes [33, 34], PCR was used to determine whether the mutations were completely or incompletely segregated. In the wild type, a 2.5 kbp DNA fragment that contains SylmG1 gene was amplified. In contrast, the 2.5 kbp DNA fragment was not detected and a 3.4 kbp DNA fragment was detected in the five independent kanamycin-resistant transformants (Figure 6A). These results indicate that the 0.9 kbp nptII gene cassette was integrated into the SylmG1 genomic locus and that the mutation was completely segregated. To overexpress SylmG1, the bacterial consensus II promoter and SylmG1 orf fusion was integrated into a neutral site of the S. elongatus genome [35]. RNA gel blotting indicated that SylmG1 was overexpressed in the transformants (Figure 6B).

To examine the effect of disruption and overexpression of the SylmG1 gene on cell division as well as nucleoid structure, cells in the exponential phase were stained with DAPI and observed under microscopy. Although the shape and length of the ΔSylmG1 cells were similar to the wild type, the intensity of DAPI fluorescence was higher in ΔSylmG1 (Figure 6C). However, the amount of total DNA extracted from the same number of cells did not differ between the wild type and ΔSylmG1 (ΔSylmG1 /wild type, 1.03 ± 0.01). These results suggest that nucleoid compaction occurred in ΔSylmG1. On the other hand, SylmG1 overexpressers frequently contained cells significantly longer than the wild-type cells (two times longer on average, Figure 6D), suggesting that cell division is partially impaired in the overexpresser. In addition, abnormal distribution of nucleoids was observed in the overexpresser (Figure 6D and 6E middle panel) and ~2% cells exhibited extremely biased segregation of nucleoids during cell division (Figure 6E right panel). These results suggest that the overexpression of SylmG1 impairs nucleoid segregation during cell division. To further examine how cell division is impaired in the overexpresser, we examined FtsZ localization by immunofluorescence microscopy using anti-FtsZ antibodies. The antibodies detected the FtsZ ring at the mid-cell position in the wild type (Figure 6E). In the SylmG1 overexpressers, the FtsZ rings had a tendency to be biased towards the side of the cell to which nucleoid density was biased (Figure 6E middle panel). In addition, a diffuse but higher concentration of FtsZ localization was observed around the region where nucleoid density was biased (Figure 6E right panel). These results suggest that SYlmG1 is required to maintain normal nucleoid structure, and that the FtsZ localization might be related to the nucleoid partitioning by YlmG.

Growth of organisms

A. thaliana (Col-0) was used as the wild type. A. thaliana seeds were surface-sterilized, sown on Murashige and Skoog (MS) plates, and stratified at 4°C for 2 days in the dark before germination. Plants were grown in controlled-environment chambers with 16 h of light (100 μmol/m²s) and 8 h of dark at 20°C. Seedlings were transferred onto soil and were grown in the controlled-environment chambers.

Synechococcus elongatus was grown in BG-11 medium [45] in 50 ml flasks on a rotary shaker or 1.2% agar plates at 30°C in continuous light (100 μmol/m²s). Growth of cells in the liquid cultures was measured by determining OD₇₃₀.

Isolation of chloroplast division mutants in the FOX library

T2 seeds of the A. thaliana FOX lines were germinated and grown on MS plates for 3 weeks. Tips from expanding leaves were put on a glass slide without fixation, covered with a cover slip, and smashed gently. Samples were observed with Nomarski differential interference contrast optics.

To identify the inserted 35S-cDNA in the FOX lines, the insertion was amplified by PCR using primers 5'-GTACGTATTTTTACAACAATTACCAACAAC-3' and 5'-GGATTCAATCTTAAGAAACTTTATTGCCAA-3', and then sequenced by a primer 5'-CCCCCCCCCCCCD (A or G or T)-3'.

Construction and generating transgenic A. thaliana

For overexpression of AtYLMG1-1, a genomic region containing AtYLMG1-1 orf was amplified by primers 5'-ATGTCTAGAATGGCCGCCATTACAGCTCTC-3' (the XbaI site is underlined) and 5'-ATGGAGCTCCGTTTCAACAAAACCATTAGC-3' (the SacI site is underlined). For expression of antisense AtYLMG1-1 gene, a genomic fragment was amplified by primers 5'-ATGTCTAGATCACAGAGATCTCTAATGGCA-3' (the XbaI site is underlined) and 5'-AGTGAGCTCTCTTCAACAGGCGGAATAAC-3' (the SacI site is underlined). These amplified products were digested with XbaI and SacI and were inserted between XbaI and SacI sites of pBI121 vector. Above constructs were transferred to Agrobacterium tumefaciens GV3101 and introduced into A. thaliana plants as described [46]. Transformants were selected on the MS medium containing 30 mg/L kanamycin and T2 plants were used for further analyses.

Construction and generating transgenic S. elongatus

For targeted disruption of SylmG1 gene (ORF ID; Synpcc7942_0477), a unique restriction site (XbaI site) was added to the genomic fragment containing SylmG1 by overlap-extension PCR. We amplified a 100 bp of SylmG1 orf franked by an 850 bp upstream sequence by primers 5'-CCGCGATCGGCTCTCGCGTGATTGCCAGCG-3' and 5'-CATTCTAGAGGAACCAGCTCAGTAAGACGC-3' (the XbaI site is underlined). A 200 bp of SylmG1 orf franked by an 850 bp downstream sequence was amplified by primers 5'-ACTGAGCTGGTTCCTCTAGAGAGCAGTCAGTTCATGCTGAT-3' and 5'-ACGGTGGCGATGAGCACGGCTACACCGACT-3' (the XbaI site is underlined). These two amplified fragments were mixed and fused by PCR using primers 5'-CCGCGATCGGCTCTCGCGTGATTGCCAGCG-3' and 5'-ACGGTGGCGATGAGCACGGCTACACCGACT-3'. The fused fragment was cloned into pGEM-T easy (Promega). An orf of nptII, which confers resistance to kanamycin, was amplified by primers 5'-ATGTCTAGAAGCTATGACCATGATTACGAA-3' and 5'-ATGTCTAGAAAGTCAGCGTAATGCTCTGCC-3' (the XbaI site is underlined), digested with XbaI, and inserted into SylmG1 orf in which an XbaI site was introduced as above. A construct in which the nptII cassette was inserted in the same orientation was used for gene disruption. Genotypic character of the wild type and kanamycin-resistant mutants (lines #1-5) which were subjected to PCR analysis with primers 5'-TGACGGACTTCTTCGACCAGATG-3' and 5'-ATTGAACCGCGTTGGGACAAGG-3'.

For overexpression of SylmG1, a DNA fragment containing SylmG1 orf was amplified by primers 5'-ATGCCCGGGGACAGATTTATTGGACGGTGA-3' and 5'-ATGCCCGGGCAAGCGGAGCTCTATCACGAA-3' (the SmaI site is underlined). The amplified product was digested by SmaI and inserted into SmaI site of the pTY1002 vector, which contains a bacterial consensus II promoter, aadA gene which confers resistence to spectinomycin, and S. elongatus neutral site (position is 2577767-2578661 and 2578658-2580657) [35].

Above constructs were transformed into the wild type cells. Transformants were selected on BG-11 plates containing 15 mg/L kanamycin or 10 mg/L spectinomycin. For the SylmG1 disruption, homologous recombination and segregation were confirmed by PCR using primers 5'-TGACGGACTTCTTCGACCAGATG-3' and 5'-ATTGAACCGCGTTGGGACAAGG-3'. Overexpression of SylmG1 transcript was confirmed by a RNA gel blot analysis using digoxigenin-labeled SylmG1 specific probe.

Semi-quantitative RT-PCR

DNA-free total RNA (1 μg) was reverse-transcribed using Primescript (Takara) with oligo(dT)₁₅ primer. PCR was performed by using primers 5'-CACCGAGAAGTCAACAGCTCGGTCATCGAC-3' and 5'-TCAAGTCTTCCAATTTCTACCCAGTGCTGC-3' for AtYLMG1-1, 5'-CCTCAACATATATAACACCATC-3' and 5'-GACAGGTTCAGGTCATAGAAG-3' for At5g21920, 5'-TATCTGAACACTCCGTTGACGGTA-3' and 5'-CAAAGATAAACGGAATACGATC-3' for At4g27990, 5'-GCAATGGGAAGCAGTGGTGG-3' and 5'-GGGAGAAGAGACGGGTTTCG-3' for GAPDH, and 5'-GGCCAAGATCCAAGACAAAG-3' and 5'-GTTGACAGCTCTTGGGTGAA-3' for UBQ1.

Phylogenetic analyses

Deduced amino acid sequences of YlmG homologs encoded by the 41 genes were collected by BLAST searches against public databases. The sequences were aligned by Clustal X 2.0 [47], manually refined and 82 amino acid residues corresponding to the YGGT domain were used for the phylogenetic analyses. Maximum likelihood trees were constructed PHYML [48] based on the WAG model with the discrete gamma model for site-heterogeneity (8 categories with 100 replications). Bayesian inference was performed with the program MrBayes version 3.1.2 [49] using the WAG matrix assuming a proportion of invariant positions and four gamma-distributed rates. For the MrBayes consensus trees, 1,000,000 generations were completed with trees collected every 100 generations.

Microscopy

For observation of chloroplast size and number, tips from expanding leaves were cut and fixed by 3.5% glutaraldehyde for 1 h at room temperature and then incubated in 0.1 M Na₂-EDTA pH 9.0 for 15 min at 50°C. Samples were analyzed with Nomarski differential interference contrast optics.

For observation of chloroplast nucleoids, leaves were cut into small fragments and then digested in 2% cellulase RS (Yakult), 1% macerozyme (Yakult), 550 mM sorbitol, and 5 mM MES-KOH pH5.8 at 25°C for 30 min. Resulting protoplasts were stained by 1 μg/ml DAPI in TAN buffer (0.5 M sucrose, 0.5 mM EDTA, 1.2 mM spermidine, 7 mM 2-mercaptoethanol, 0.4 mM phenylmethyl-sulfonyl fluoride, 20 mM Tris-HCl, pH 7.5) or SYBR Green I (Invitrogen) after fixation with 1% glutaraldehyde.

Localization of FtsZ in A. thaliana expanding leaves was examined by immunofluorescence microscopy using an anti-AtFtsZ2-1 antibodies as described [9]. Localization of FtsZ in S. elongatus was examined by immunofluorescence microscopy using an anti-Anabaena FtsZ antibodies (AS07217, Agrisera) as described [25] except for Can Get Signal (Toyobo) was used for the antibody reactions.

Localization of AtYLMG1-1 in chloroplasts of expanding leaves was examined using rabbit anti-AtYLMG1-1 antibodies which were raised against two synthetic peptides (FASLRDRPPGYLNT and TEKSTARSSTLTGS). Isolated chloroplasts were fixed by 2% paraformaldehyde dissolved in TAN buffer. For antigen retrieval, fixed chloroplasts were treated by 20 μg/ml proteinaseK in TE buffer containing 0.5% Triton X-100 at 37°C for 10 min. Chloroplasts were washed twice with PBS. After blocking with 3% skim milk in PBS for 30 min, the samples were incubated for 16 h at 4°C with the anti-AtYLMG1-1 antibodies diluted 1:200 in Can Get Signal. Chloroplasts were then washed twice with 3% skim milk in PBS and then incubated with anti-rabbit IgG conjugated with AlexaFluor 488 (Invitrogen) for 2 h at room temperature. To observe chloroplast DNA, samples were stained with 1 μg/ml DAPI.

DNA-blot analysis

Total DNA were extracted from ~3 weeks plants grown on the MS medium. Three-micrograms DNA were digested with HindIII and the DNA fragments were separated on 0.8% agarose gels. The restriction fragments were transferred to Hybond-N+ membranes. Hybridization and detection were carried out as described using digoxigenin-labeled DNA probes [50]. A psbA probe was amplified with the primers 5'-TGCATAAGAATGTTGTGCTCAGCC-3' and 5'-CTACTTCTGCAGCTATTGGATTGC-3', and a PsbO probe was amplified with the primers 5'-CAATCGTGCGATTTCACAGCCACTC-3' and 5'-TTCTCTTCCAAGTTGTGTCGTCTCC-3'.

Immunoblot analyses

Intact chloroplasts were isolated from ~3 weeks plants grown on the MS medium as described [20]. To obtain insoluble (pellet) and soluble (sup) fraction of chloroplasts, intact chloroplasts were resuspended in hypotonic buffer (10 mM Tris-HCl, pH 8.0 and 1 mM EDTA), incubated for 15 min on ice, and then centrifuged at 15,000 g for 30 min at 4°C. To separate the thylakoid and the envelope membranes, intact chloroplasts were resuspended in Tricin buffer (50 mM Tricin pH 7.6 and 5 mM MgCl₂), frozen at -80°C and thawed. The chloroplast lysate was separated in a sucrose density gradient (lysate/0.6 M/1.0 M/1.2 M/1.5 M) by a centrifugation at 113,000 g for 1 h at 4°C. The envelope membrane (at the 0.6 M/1.0 M interface) and the thylakoid membrane (at the 1.2 M/1.5 M interface) were collected.

Immunoblot analyses were performed as described [50] except that 17.5% polyacrylamide gel was used and 10 mM CAPS buffer containing 10% methanol were used for the transfer onto PVDF membrane. AtYLMG1-1 was detected with the anti-AtYLMG1-1 antibodies (1: 1,000). Anti-TOC34 (AS07 238, Agrisera) and Anti-Lhcb1 (AS01 004, Agrisera) were used at dilutions of 1: 5,000 and 1: 2,000, respectively. After detection of these proteins, the Rubisco small subunit was detected by staining the same membranes with CBB.