First published online April 13, 2007; 10.1105/tpc.106.042895
The Plant Cell 19:1329-1346 (2007)
© 2007 American Society of Plant Biologists
The Nuclear-Encoded Factor HCF173 Is Involved in the Initiation of Translation of the psbA mRNA in Arabidopsis thaliana[W]
Kerstin Schulta,
Karin Meierhoffa,1,
Susanne Paradiesa,
Thomas Töllerb,2,
Petra Wolffb and
Peter Westhoffa
a Institut für Entwicklungs und Molekularbiologie der Pflanzen, Heinrich-Heine-Universität, 40225 Düsseldorf, Germany
b Institut für Physikalische Biologie, Heinrich-Heine-Universität, 40225 Düsseldorf, Germany
1 To whom correspondence should be addressed. E-mail karin.meierhoff{at}uni-duesseldorf.de; fax 49-211-81-14871.
 |
ABSTRACT
|
|---|
To gain insight into the biogenesis of photosystem II (PSII) and to identify auxiliary factors required for this process, we characterized the mutant hcf173 of Arabidopsis thaliana. The mutant shows a high chlorophyll fluorescence phenotype (hcf) and is severely affected in the accumulation of PSII subunits. In vivo labeling experiments revealed a drastically decreased synthesis of the reaction center protein D1. Polysome association experiments suggest that this is primarily caused by reduced translation initiation of the corresponding psbA mRNA. Comparison of mRNA steady state levels indicated that the psbA mRNA is significantly reduced in hcf173. Furthermore, the determination of the psbA mRNA half-life revealed an impaired RNA stability. The HCF173 gene was identified by map-based cloning, and its identity was confirmed by complementation of the hcf phenotype. HCF173 encodes a protein with weak similarities to the superfamily of the short-chain dehydrogenases/reductases. The protein HCF173 is localized in the chloroplast, where it is mainly associated with the membrane system and is part of a higher molecular weight complex. Affinity chromatography of an HCF173 fusion protein uncovered the psbA mRNA as a component of this complex.
|
INTRODUCTION
|
|---|
Photosystem II (PSII) is one of the three multisubunit protein complexes of the photosynthetic electron transport chain. It is localized in the thylakoid membranes of photosynthetically active organisms and functions as a water-plastoquinone oxidoreductase. Like other complexes of the thylakoid membrane, PSII is a genetic mosaic consisting of plastid- and nuclear-encoded subunits (Wollman et al., 1999 ); therefore, a coordination of gene expression in the plastid and the cytosol is required (Goldschmidt-Clermont, 1998).
The core complex of the PSII reaction center consists of the proteins D1 and D2, which bind all important cofactors needed for the primary charge separation (i.e., the P680 chlorophyll a, pheophytin, and plastoquinone). The 32-kD D1 protein is encoded by the psbA gene of the chloroplast genome. Regulation of D1 protein synthesis is important for the correct biogenesis of PSII during chloroplast development and for the maintenance of a functional photosystem, since D1 exhibits a high turnover due to damage conferred by light. The accumulation of the D1 protein depends on environmental signals, such as light and the developmental stage of the plant (Klein et al., 1988; Gamble and Mullet, 1989; Klein and Mullet, 1990; Klaff and Gruissem, 1991; Kim et al., 1993; Staub and Maliga, 1994).
In general, the expression of plastid-encoded genes is mainly regulated by posttranscriptional mechanisms (Deng and Gruissem, 1987). Several nuclear-encoded factors have been identified to be involved in RNA-stabilizing processes, RNA degradation, and translation by in vitro UV cross-linking of RNA with proteins, in vitro translation experiments, and characterization of nuclear mutants of green algae and higher plants (reviewed in Goldschmidt-Clermont, 1998; Barkan and Goldschmidt-Clermont, 2000; Monde et al., 2000; Zerges, 2000; Leister and Schneider, 2003). One of the best-studied examples was described for the regulation of the psbA mRNA translation in Chlamydomonas reinhardtii. A complex of at least four distinct proteins (RB38, RB47, RB55, and RB60) seems to activate translation in a light- and redox state–dependent manner (Danon and Mayfield, 1991; Trebitsh et al., 2000; Kim and Mayfield, 2002; Somanchi et al., 2005). The RNA binding activity of the complex is thought to be regulated by the proteins RB60 and Tba1, which are able to modulate the redox state of RB47 (Kim and Mayfield, 2002; Somanchi et al., 2005). Binding of the complex to the 5'-untranslated region (5'-UTR) of the psbA mRNA is mediated by RB47 and RB38 (Yohn et al., 1998a, 1998b; Barnes et al., 2004). It has been proposed that RB47 takes part in D1 synthesis for de novo PSII biogenesis because it localizes to the low-density membrane system (Zerges and Rochaix, 1998), whereas a 63-kD protein, which also binds to the 5'-leader of the psbA RNA but is localized to the stroma thylakoids, might participate in D1 synthesis for PSII repair (Ossenbühl et al., 2002).
Less is known about protein binding to the 5'-UTR of the psbA transcript in higher plants. Cross-linking experiments identified several distinct proteins, which bind to the 5'-leader of the Spinacia oleracea psbA mRNA. Among these, the protein CS1 is a homolog of the Escherichia coli ribosomal S1 protein (Alexander et al., 1998). In Arabidopsis thaliana, two proteins were identified, which bind to the 5'-UTR of the psbA mRNA in a redox-dependent manner. The binding occurs in a region comprising the ribosome binding sites, indicating a role in translation initiation of the D1 protein (Shen et al., 2001). However, the identity of these two proteins is still unclear.
We are interested in understanding the biogenesis of the multisubunit protein complex PSII. To identify genes that are involved in this process, we investigate nuclear mutants of Arabidopsis that exhibit the high chlorophyll fluorescence phenotype (hcf), which indicates defects in the photosynthetic light reactions and the electron transport (Miles, 1982). Here, we report on hcf173, a mutant that is especially affected in D1 protein synthesis. The altered polysome association of the psbA mRNA in hcf173 suggests that translation initiation of this transcript is impaired. The mutant phenotype was rescued by complementation experiments with the At1g16720 cDNA. At1g16720/HCF173 encodes a protein that shows weak similarities to the short-chain dehydrogenase/reductase (SDR) superfamily. The protein is mainly localized to the chloroplast membranes and part of a high molecular weight complex, which is associated with the psbA mRNA, emphasizing the possible role of HCF173 in psbA gene expression.
|
RESULTS
|
|---|
Hcf173 Is Affected in PSII Function
The mutant hcf173 was isolated by its recessive high chlorophyll fluorescence phenotype from a collection of M2 families that were obtained by ethyl methanesulfonate (EMS) mutagenesis of Arabidopsis seeds as described (Meurer et al., 1996b). Mutant seedlings of hcf173 were not able to grow photoautotrophically on soil but could be maintained on sucrose-supplemented medium. However, they did not develop any fertile flowers.
Chlorophyll fluorescence induction measurements (Figure 1A
) revealed a drastically reduced ratio of variable fluorescence to maximal fluorescence (Fv/Fm) in hcf173 (0.15 ± 0.02) when compared with the wild type (0.80 ± 0.01). This is indicative of a severe defect of PSII in hcf173, which was further supported by an extremely low quantum yield of PSII ( II), with 0.065 ± 0.02 in hcf173 versus 0.74 ± 0.03 in the wild type. Photochemical quenching was less affected in the mutant plants [qp(hcf173) = 0.73 ± 0.07 and qp(WT) = 0.97 ± 0.02]. This shows that the few electrons released by PSII were transported through the electron transport chain, thus indicating an active cytochrome b6f complex and photosystem I (PSI). This finding was supported by measurements of P700 redox kinetics in wild-type and hcf173 mutant plants (Figure 1B). In hcf173, we detected a clear absorbance change induced by far-red light, suggesting that PSI is functional. Application of a saturating light pulse under far-red light induces PSII electron transport leading to a complete rereduction of P700 in the wild type. By contrast, only a weak rereduction of the PSI reaction center was detectable in the mutant, which is in line with its strong PSII defect.
View larger version (7K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 1. Spectroscopic Analyses of Wild-Type, hcf173 Mutant, and Complemented hcf173 Plants.
Chlorophyll fluorescence induction (A) and P700 absorbance kinetics (B) were measured with 3- or 4-week-old plants, respectively.
|
|
Taken together, these spectroscopic analyses indicate that the mutant hcf173 is affected mainly in PSII and not in the intersystem electron transport chain or PSI.
hcf173 Fails to Accumulate PSII Subunits
To corroborate the results obtained by spectroscopic measurements, we monitored the steady state levels of distinct subunits of the photosynthetic membrane complexes with specific antibodies (Meurer et al., 1996b). The plastid-encoded subunits D1, D2, CP43, and CP47 of PSII were hardly detectable in hcf173 (Figure 2
). The cytochrome b559 polypeptides were reduced to  25% of wild-type amounts and the nuclear-encoded subunits of the oxygen-evolving complex (23- and 34-kD proteins) to 25 or 12.5%, respectively. We further analyzed representative polypeptides of PSI (PsaD and PsaA/B), the cytochrome b6f complex (SUIV and Cytf), and the ATP synthase ( -SU). Reduced protein levels were found for PsaD (20%) and PsaA/B (50%). Subunits of the cytochrome b6f complex and the ATP synthase accumulated to normal wild-type levels. Hence, the immunoblot analysis indicates a severe defect in PSII of hcf173. Obviously, subunits of the core complex and the inner antenna cannot accumulate, whereas subunits of the oxygen-evolving complex accumulate to low but detectable amounts. The reduction of PSI proteins was less severe than that found for PSII components. Furthermore, hcf173 mutant plants grown under low-light conditions (2 to 5 µmol·m–2·s–1) accumulated much higher amounts of PSI proteins compared with plants cultivated under standard light conditions (50 to 70 µmol·m–2·s–1) (Plücken et al., 2002). This suggests that the levels of PSI subunits in hcf173 mutant plants are sensitive to light and therefore not directly influenced by the mutation.
View larger version (45K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 2. Immunoblot Analysis of Thylakoid Membrane Proteins from the Wild Type and the Mutant hcf173.
Twenty micrograms (WT 1/1 and hcf173), 10 µg (WT 1/2), 5 µg (WT 1/4), and 2.5 µg (WT 1/8) of protein were separated by SDS-PAGE and transferred to a nitrocellulose membrane. Specific proteins were immunodecorated with the indicated antisera.
|
|
Together, we conclude from these data that the nuclear mutation in hcf173 leads primarily to a drastically impaired accumulation of PSII polypeptides.
The Synthesis of the PSII Reaction Center Protein D1 Is Drastically Reduced in hcf173
To define the mutational defect in the mutant hcf173 more precisely, the synthesis of plastid-encoded subunits of the thylakoid membrane was examined by in vivo labeling of leaf proteins with 35S-Met. Cytoplasmic translation of nuclear-encoded subunits was blocked by application of cycloheximide. Figure 3A
shows that after a labeling period of 20 min, the rates of synthesis of thylakoid membrane polypeptides were similar in wild-type and mutant leaves with the exception of the PSII reaction center protein D1. The incorporation of radioactivity into this protein was eight times lower in hcf173 compared with the wild type. We repeated the in vivo labeling experiments with a shorter pulse of only 5 min. Even under these conditions a reduced D1 labeling was observed, indicating that indeed D1 synthesis is affected and not D1 protein stability (see Supplemental Figure 1 online).
View larger version (47K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 3. In Vivo Protein Synthesis in Wild-Type Plants and hcf173 Mutants.
Newly synthesized proteins of 15-d-old plants were radiolabeled with 35S-Met for 20 min.
(A) Wild-type and hcf173 membrane proteins with equivalent amounts of radioactivity (100,000 cpm) or the indicated dilutions were separated by SDS-PAGE and analyzed by fluorography.
(B) SDS-solubilized membrane proteins with the equivalent of 200,000 incorporated cpm were used for immunoprecipitation with anti-D1. The hcf173 sample was filled up with unlabeled wild-type membrane protein to guarantee equal amounts of D1 in both the wild-type and the mutant sample. Anti-D1 was used in excess for immunoprecipitation.
|
|
To confirm the identity of the supposed D1 protein band, immunoprecipitation experiments were performed. After a labeling period of 20 min, membrane proteins were solubilized and incubated with a specific antiserum raised against the D1 protein. Gel electrophoresis of immunoprecipitated proteins and subsequent autoradiography (Figure 3B) confirmed the reduced labeling of the D1 protein in the mutant hcf173. Together, these results provide strong evidence that a reduced synthesis rate of the PSII reaction center protein D1 is the primary mutational defect in hcf173.
psbA mRNA Accumulation and Ribosomal Loading Are Impaired in hcf173 Mutants
To confirm that the reduced D1 synthesis rate in the hcf173 mutant is caused by a defect in translation, we first determined the amount of the psbA mRNA. Figure 4
shows that the level of the monocistronic psbA transcript was reduced to 12% of the wild-type amount in the mutant hcf173. We further examined the transcript levels of other PSII, PSI, cytochrome b6f complex, and ATP synthase polypeptides (i.e., psbC, psbD, psbE/F, psbO, psbP, psaA, psaF, petA, petD, and atpA), but no significant differences between wild-type and mutant transcript levels were detected (Figure 4; data not shown).
View larger version (53K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 4. Transcript Analysis of psaA, psbA, psbE/F, and psbC in the hcf173 Mutant.
Six micrograms or the indicated dilutions of total leaf RNA from 3-week-old wild-type and hcf173 mutant plants were separated by denaturing agarose gel electrophoresis, blotted onto a nylon membrane, and hybridized with radiolabeled probes as described. The psbC mRNA is transcribed in a polycistronic transcription unit resulting in a complex transcript pattern (Meurer et al., 1996a).
|
|
To address the question whether the mutation in hcf173 affects the ability of the psbA mRNA to associate with ribosomes, we examined ribosomal loading of this transcript. To this end, leaf extracts were fractionated in sucrose gradients under conditions that maintain intact polysomes (Barkan, 1993). Efficiently translated ribonucleic acids are highly associated with ribosomes and therefore migrate deep into the gradient. Specific mRNAs were localized in the gradients by performing RNA gel blot hybridizations with RNA purified from gradient fractions. The distribution of plastidic and cytosolic rRNAs in wild-type and hcf173 polysome gradients was determined, and both gradients showed an equal pattern of rRNA distribution, revealing that there is no general difference in the distribution of ribosomes between the mutant and the wild type (Figure 5A
). By contrast, different sedimentation behavior was observed for the psbA mRNA in the wild type and hcf173 (Figure 5B). Quantification of the signals by phosphor imaging revealed that 27% of wild-type psbA mRNA sedimented in the polysomal region of the gradient (fractions 7 to 11). In hcf173 mutants, the bulk of the psbA mRNA was localized in the fractions containing monosomes or free RNA, and only 4% were detected in the polysomal fraction, indicating a low translational activity of this transcript. To confirm the distribution of small RNA particles (monosomes and free RNA) and polyribosomes in the sucrose gradients, crude polysomal RNA from the wild type and hcf173 was treated with EDTA, which disrupts ribosome association. This treatment resulted in a shift of the wild-type psbA mRNA into the first six fractions (Figure 5C), revealing that the RNA migrating deeper into the gradient represents polysomes. We further examined polysomal association of the psbB mRNA, which is encoded in the psbB-psbT-psbH-petB-petD polycistronic transcription unit (Westhoff and Herrmann, 1988). In the wild type, 65% of the psbB/T transcript was found in the polysomal fractions of the gradient (fraction 7-11; Figure 5D). In the mutant, the corresponding fractions contained lower amounts (42%) of the psbB/T mRNA, and a major amount of the transcript was associated with small polysomes (fraction 7). Because there is no indication for RNA degradation in the polysomal fractions of the mutant, this difference could be an indication of impaired translation of psbB-containing transcripts in hcf173 mutant plants.
View larger version (48K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 5. Association of the psbA Transcript with Polysomes in the hcf173 Mutant.
Whole-cell extracts were fractionated in linear 15 to 55% sucrose gradients by ultracentrifugation. Gradients were divided into 11 fractions, and equal proportions of each fraction were analyzed.
(A) RNA gel blots of polysome gradient fractions were stained with methylene blue to visualize distribution of rRNAs.
(B) The blots shown in (A) hybridized with a probe that detects psbA mRNA. The RNA amounts of the different fractions were quantified by phosphor imaging.
(C) Cell extracts treated with EDTA to disrupt polysome association.
(D) RNA gel blots hybridized with a probe specific for psbB. The signals of the large primary transcripts are faint, but their positions are indicated.
|
|
Together, these results imply that the reduced labeling of the D1 protein in hcf173 is mainly the result of drastically impaired translation. The lack of polysomal association suggests that the hcf173 mutation affects the initiation of psbA mRNA translation or very early steps of translation elongation. Moreover, a direct or indirect effect of the mutation on the translational efficiency of the psbB/T transcript is possible, although the in vivo labeling experiments did not corroborate this assumption. The 5-min pulse indeed showed weaker labeling of CP47 but also for all other synthesized proteins in the mutant (see Supplemental Figure 1 online). Moreover, the 20-min pulse resulted in similar signals of CP47 in the wild type and mutants (Figure 3A).
The altered ribosomal loading possibly causes the diminished accumulation of the psbA mRNA in the mutant. To test whether the reduced psbA mRNA level is caused by decreased mRNA stability, psbA mRNA decay in wild-type plants and hcf173 mutants was investigated in the presence of the transcriptional inhibitor actinomycin D. Total RNA was isolated from single plants after incubation in the inhibitor solution for 0, 2, 6, 12, 18, or 24 h. Isolated RNAs were analyzed by gel blot hybridization with probes for rbcL and psbA mRNAs and subsequent autoradiography. Figure 6A
shows that the stability of the psbA mRNA in the mutant was drastically reduced (t1/2psbA hcf173  6 h) compared with wild-type plants (t1/2psbA WT 47 h). As a control, the stability of the rbcL mRNA was measured. The half-life of this RNA was similar in the wild type (t1/2 rbcL WT 22 h) and mutants (t1/2rbcL hcf173 15 h) (Figure 6B). Thus, the observed RNA degradation is not the result of a general RNA instability in the mutant plants. The reduction of psbA mRNA stability in hcf173 by a factor of eight is sufficient to explain the low steady state level of the psbA transcript. Therefore, our result suggests that the reduced amount of psbA mRNA in the mutant results from increased RNA turnover and not from a reduced transcription rate.
View larger version (7K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 6. Stability of psbA and rbcL Transcripts in the hcf173 Mutant.
Wild-type and mutant plants were harvested and incubated with a solution of the transcriptional inhibitor actinomycin D and water as control. RNAs were isolated 0, 2, 6, 12, 18, and 24 h after the start of actinomycin D or water treatment. Amounts of psbA mRNA (A) and rbcL mRNA (B) were quantified by RNA gel blot analysis and subsequent phosphor imaging.
|
|
Molecular Cloning of HCF173
The mutant hcf173 was generated by EMS mutagenesis of Arabidopsis seeds. A set of microsatellite and cleaved amplified polymorphic sequence (CAPS) markers (Konieczny and Ausubel, 1993; Bell and Ecker, 1994) was used to map the gene HCF173 on the upper arm of chromosome 1 (Figure 7A
). To determine the HCF173 locus more precisely, several simple sequence length polymorphism and CAPS markers were developed and used for genotyping 1279 plants of the mapping population. The markers SNP711/918 on BAC clone F19K19 (one recombination) and SNP872 on BAC clone F17F16 (one recombination) were found to flank the HCF173 locus. The DNA region between these markers (91 kb) comprised three genes that encode proteins with a predicted chloroplast transit peptide (At1g16610, At1g16630, and At1g16720). The transcript levels of these genes were investigated in wild-type and mutant plants. Indeed, one of these genes, At1g16720, showed a drastically reduced RNA level in the hcf173 mutant (Figure 7B). To test whether At1g16720 was affected by a mutation, we amplified the whole sequence by PCR using genomic DNA from hcf173 mutants as a template. Sequence analysis revealed a point mutation in exon 10 at base pair position 2524 (Figure 8
). The exchange from cytosine to thymine at this site caused a premature termination codon and would very likely lead to synthesis of an incomplete HCF173 protein. The incorrect RNA is probably submitted to the pathway of nonsense-mediated mRNA decay (Baker and Parker, 2004), which would explain the reduced transcript level. This mutant allele was referred to as hcf173-1. The discovery of a second allele (hcf173-2) in the GABI-Kat collection (Rosso et al., 2003; line 246C02) supports the assumption that At1g16720 encodes the nuclear factor HCF173. Line 246C02 contains a T-DNA insertion in exon 3 of At1g16720 (Figure 8). The mutant exhibited the hcf phenotype under UV light and at the molecular level showed the same defects as hcf173-1 (data not shown).
View larger version (28K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 7. Mapping of the HCF173 Locus.
(A) The HCF173 locus was localized on the upper arm of chromosome 1 between the two CAPS markers SNP711/918 (BAC clone F19K19) and SNP872 (BAC clone F17F16). Within the mapped region, three genes (At1g16610, At1g16630, and At1g16720) were identified, which encode proteins with a putative chloroplast transit peptide.
(B) The transcripts of the genes At1g16610, At1g16630, and At1g16720 were analyzed by RNA gel blot hybridization with gene-specific probes. To this end, 5 µg of poly(A)+ RNA of 3-week-old wild-type plants and hcf173 mutants were separated on a denaturing agarose gel.
|
|
View larger version (15K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 8. Gene Structure of HCF173 and Mutant Alleles.
HCF173 represents an open reading frame of 3088 bp consisting of 13 exons separated by 12 introns. Sequence analysis of the mutant allele hcf173-1 discovered a point mutation at base pair 2524 in exon 10. This base pair exchange causes a premature stop codon. In hcf173-2 (GABI-Kat line 246C02), the T-DNA insertion was detected in exon 3.
|
|
We further confirmed the identity of the gene At1g16720 by complementation analysis. The cDNA of At1g16720 under the control of the 35S promoter was transferred into heterozygous hcf173-1 plants by Agrobacterium tumefaciens–mediated transformation. In the progeny of these transformants, we identified two transgenic plants that were homozygous for the mutation hcf173 and were able to grow photoautotrophically on soil. Chlorophyll fluorescence induction measurements showed that the transgenic plants exhibit the same kinetics as wild-type plants (Figure 1A), indicating that PSII is fully active. Thus, the introduced cDNA of At1g16720 is able to complement the hcf173 mutant phenotype. Together, the data of the identified T-DNA mutant and the complementation analysis revealed that the gene At1g16720 encodes the factor HCF173.
HCF173 Encodes a Protein with Similarity to the SDR Superfamily
The HCF173 gene encodes a protein of 598 amino acids with a predicted molecular mass of 65.7 kD. The WolfPSORT program (Horton et al., 2006; http://wolfpsort.seq.cbrc.jp) predicted a chloroplast transit peptide of 83 amino acids (Figure 9
).
View larger version (97K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 9. Sequence Alignment of Arabidopsis HCF173 (At-HCF173) and Putative HCF173 Homologs from Rice (Os-HCF173-1 and Os-HCF173-2) and C. reinhardtii (Cr-HCF173).
Predicted ends of chloroplast transit peptides are boxed: white on black, identical to At-HCF173; black on gray, similar to At-HCF173. Gray bars indicate SDR-like regions 1 and 2. The predicted secondary structure of the putative dinucleotide binding site of HCF173 is schematically indicated above the sequences: black bar, -helix; zigzag, ß-strand. Positions of the Gly-rich motif and the conserved Asp are indicated by asterisks; other conserved amino acid residues in the ßß-turn are marked by triangles.
|
|
BLAST analyses (Altschul et al., 1994) indicated two possible HCF173 homologs in rice (Oryza sativa; Figure 9): Os-HCF173-1 consists of 648 amino acids and is encoded on chromosome 6; Os-HCF173-2 has a length of 598 amino acids, and the corresponding gene is localized on chromosome 2. According to prediction programs, both proteins possess an N-terminal chloroplast transit peptide of 89 and 83 amino acid residues, respectively. In pairwise comparisons between the protein sequences of Arabidopsis HCF173 and the rice proteins, 74% identical amino acid residues were observed. Additional BLAST searches in the database of Chlamydomonas revealed an HCF173 homolog in the green alga. The sequence alignment showed an amino acid identity of 45% and total sequence similarity of 65% between the Arabidopsis and the predicted Chlamydomonas protein. However, it has to be mentioned that the deduced protein sequence of Chlamydomonas HCF173 starts with a Leu instead of a Met residue, suggesting that the 5'-end of the proposed gene model is incomplete. Moreover, BLAST analysis revealed similarities of HCF173 with the UOS1 protein and its homologs (see Supplemental Figure 2 online). UOS1 stands for UV-B and ozone similarly regulated protein 1, as the UOS1 transcript decreases due to UV-B irradiation or ozone application (Sävenstrand et al., 2002). HCF173 and several UOS1 from cyanobacteria (Nostoc sp PPC7120 and Anabaena variabilis ATCC 29413) and higher plants (Pisum sativum and Arabidopsis) shared total sequence similarities between 35 and 38% (20 to 23% sequence identity). The distinct role of the UOS1 protein and its site of function are still unknown and therefore allowed no functional prediction for HCF173.
Furthermore, database searches revealed two conserved domains in the protein sequence of HCF173 (Figures 9 and 10
). These domains are located between amino acids 161 to 260 and 461 to 560. Many proteins were identified showing sequence similarities with these two regions. The conserved regions of HCF173 shared highest sequence similarity with a 219–amino acid YCF39-like protein from Synechocystis sp PCC6803, showing 34% identical and 54% similar amino acids (Figure 10). Analysis of the Synechocystis protein sequence by the Conserved Domain Database (Marchler-Bauer et al., 2005) revealed highest similarity to the clusters of the orthologous groups COG0451 and COG0702. These clusters comprise nucleoside-diphosphate-sugar epimerases or predicted nucleoside-diphosphate-sugar epimerases, respectively. The epimerases of COG0451 are members of the SDR superfamily. Predictions of the secondary structure of HCF173 with the web tool Jpred (Cuff and Barton, 2000) suggested a pattern of alternating -helices and ß-strands that is typically observed for SDR proteins (Jörnvall et al., 1995). Therefore, we decided to designate HCF173 as SDR-like protein and the two conserved sequence elements as SDR-like domains.
View larger version (23K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 10. Partial Alignment of Arabidopsis HCF173 and a YCF39-Like Protein from Synechocystis sp.
Amino acid residues 161 to 260 and 460 to 560 of HCF173 were aligned to the first 200 of a total of 219 amino acid residues of the YCF39-like protein. White letters on black indicate identical residues; black letters on gray mean similar residues.
|
|
HCF173 Is Associated with Chloroplast Membranes
The N-terminal part of the HCF173 polypeptide was predicted to function as a chloroplast transit peptide. To test this assumption, we transformed heterozygous hcf173 mutants with two fusion genes encoding C-terminally tagged proteins, which allow the localization of the HCF173 protein: The fusion gene At-HCF173-tHA consisted of HCF173 and the DNA sequence of a triple hemagglutinin tag. In the second construct, At-HCF173-TAP HCF173 was fused to the DNA sequence of the so-called twin affinity purification (TAP) tag (Puig et al., 2001). We identified five transgenic HCF173-ctHA plants and seven transgenic HCF173-cTAP plants that were homozygous for the mutation hcf173 and were able to grow photoautotrophically on soil. Spectroscopic measurements and analyses of the accumulation of thylakoid membrane proteins confirmed a fully functional PSII and therefore the complementation of the mutant phenotype (data not shown).
For localization of HCF173, we prepared total leaf, total chloroplast, chloroplast membrane, and stroma proteins from HCF173-ctHA and wild-type plants. The proteins were separated by gel electrophoresis, electrotransferred to nitrocellulose membranes, and incubated with a high-affinity antihemagglutinin peroxidase conjugate. In extracts obtained from complemented plants, a protein of 70 kD could be detected (Figure 11A
), which was expected from the HCF173-tHA protein sequence. As predicted, the HCF173-tHA protein was indeed detectable in isolated chloroplasts. The purity of this fraction was confirmed by the cytosolic marker protein -tubulin. Compared with the total protein fraction, no enrichment of HCF173-tHA was observed in the total chloroplast protein fraction. This stands in contrast with the chloroplast proteins PsaD and RbcL and indicates that the overexpressed protein is not completely imported into the organelle (Figure 11A). Separation of extracts from isolated chloroplasts into membrane and soluble fractions revealed a heterogeneous distribution of HCF173-tHA within the plastid. Approximately 20% of the protein was localized in the stroma, but the bulk of HCF173-tHA was associated with chloroplast membranes. PsaD and RbcL, two proteins that are membrane or predominantly stroma localized, respectively, served as controls for this experiment.
View larger version (39K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 11. Immunolocalization of HCF173-tHA.
(A) Chloroplasts were isolated from 3- to 4-week-old wild-type and HCF173-ctHA plants by Percoll gradient centrifugation, lysed, and separated into chloroplast membranes and stroma proteins. A protein equivalent to 15 µg of chlorophyll was used for SDS-PAGE and gel blot analysis. In addition, a total leaf protein extract containing the same protein amount as the total chloroplast extract was analyzed.
(B) Intact chloroplasts were isolated from 3-week-old HCF173-ctHA plants by Percoll gradient centrifugation and ruptured in a glass homogenizer. Chloroplast membranes were floated by discontinuous sucrose gradient centrifugation. Fractions (340 µL) were collected from the bottom to the top of the gradient. Aliquots of the fractions were used for SDS-PAGE and gel blot analysis.
(C) Chloroplast membranes were isolated from HCF173-ctHA plants and sonicated in 200 mM NaCl or 250 mM Na2CO3, respectively. Proteins of repelleted membranes and the supernatants were separated by SDS-PAGE. The protein equivalent of 20 µg chlorophyll was analyzed. As controls, a membrane-associated protein that is not releasable (PsaD) and the 34-kD (PsbO) and 23-kD (PsbP) proteins of the oxygen-evolving complex, which are releasable by sodium carbonate treatment or sonification, respectively, were used.
|
|
To confirm if the observed membrane association of HCF173-tHA is a genuine feature or rather due to overexpression of the fusion protein, we performed a membrane flotation experiment, which was originally used to separate different membrane systems on a sucrose gradient by their buoyant densities (Cline et al., 1981). To this end, lysed chloroplasts were placed at the bottom of the gradient. During ultracentrifugation, the chloroplast membranes floated through the gradient to a zone of equal density. Proteins that are associated with the membrane system should migrate in the same manner. Figure 11B clearly shows that HCF173-tHA comigrated with PsaD and therefore with chloroplast membranes into fraction 5, whereas RbcL was evenly distributed in the first five fractions. This result revealed that HCF173-tHA is a bona fide membrane-bound protein and not artificially attached to the thylakoids. Washing of chloroplast membranes with sodium chloride did not release the protein from the membrane; however, with sodium carbonate, 50% of HCF173-tHA was removed (Figure 11C). This finding indicates a firm association of the fusion protein with the chloroplast membranes. Although it is possible that membrane association is mediated by the tHA tag itself, it is unlikely because immunolocalization experiments with fusion constructs of HCF173 containing the TAP tag demonstrated membrane binding as well (data not shown). In addition, the protein Mbb1 was identified as a stroma component of the chloroplast via a triple hemagglutinin tag (Vaistij et al., 2000). Thus, it appears that HCF173 is indeed mainly attached to chloroplast membranes in vivo.
HCF173 Is Part of a Higher Molecular Weight Complex
To investigate whether HCF173 is associated with other proteins, solubilized chloroplast membrane proteins from HCF173-cTAP, HCF173-ctHA, and wild-type plants were analyzed by nondenaturing blue native PAGE (Schägger et al., 1994) in the first dimension followed by separation on denaturing SDS gels in the second dimension. For detection of the fusion proteins, the gels were subjected to immunoblot analyses. As shown in Figure 12A
, in the first dimension, the majority of HCF173-TAP was detected in the range between 130 and 150 kD. In addition, minor amounts could be detected in several complexes ranging from 210 to >670 kD. This finding suggested that HCF173-TAP is part of at least one higher molecular weight complex in vivo. In the case of HCF173-tHA, most of the protein was detected in the range of 100 to 110 kD (Figure 12B). In addition, the protein was located to complexes of 140, 200, 230, and 300 kD, emphasizing the finding that HCF173 interacts with other factors. It is not yet clear if the strong signals at 130 to 150 kD and 100 to 110 kD, respectively, are caused by HCF173 fusion protein monomers or if the protein is already bound to an interaction partner. Although the monomers are smaller (HCF173-TAP 90 kD; HCF173-tHA 70 kD), the strong intensity of the detected signals indicate that they represent the overexpressed fusion protein monomers.
View larger version (31K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 12. Identification of HCF173-TAP/HCF173-tHA–Containing Complexes by Two-Dimensional Blue Native/SDS-PAGE.
Solubilized chloroplast membrane protein complexes of wild-type, HCF173-cTAP, and HCF173-ctHA plants were separated according to their molecular masses by blue native (BN) PAGE (first dimension). Subsequent denaturing SDS-PAGE separated the complexes into their subunits (second dimension). Complexes containing HCF173-TAP (A) or HCF173-tHA (B) were identified by gel blot analysis. The cytochrome f subunit of the cytochrome b6f complex dimer was detected due to its heme-associated peroxidase activity and is indicated by an asterisk.
|
|
The HCF173-tHA Complex Is Associated with psbA RNA
The involvement of HCF173 in RNA translation and stabilization implies that the HCF173-tHA complex could be associated with RNA. To test this assumption, affinity chromatography was performed with protein extracts from wild-type and HCF173-ctHA plants. Chloroplast membranes were solubilized as described in Methods, and the protein extracts were incubated with an anti-HA affinity matrix. Purified proteins were competitively eluted by HA peptide, and nucleic acids were purified by phenol/chloroform extraction followed by alcohol precipitation. RT-PCR and quantification of RNA amounts were performed as described in Methods. Within the exponential phase of the polymerase chain reactions, the ratios of the transcript amounts obtained by HCF173-tHA affinity chromatography to the transcript amounts purified by wild-type affinity chromatography were determined (Figure 13A
). The examined transcripts could be detected in both the wild-type and HCF173-ctHA samples. Therefore, we conclude that the anti-HA affinity chromatography resin unspecifically binds nucleic acids. Nevertheless, the data indicate that in all three independent experiments, the psbA mRNA was enriched in the HCF173-ctHA sample as the ratio of psbA mRNAHCF173-ctHA to psbA mRNAWT was 2.4- to 6.2-fold higher compared with all other transcript ratios. Moreover, RNA samples of two additional experiments were submitted to slot blot hybridization and analyzed with a psbA and a psbB probe, respectively (Figure 13B). Both experiments showed a clear enrichment of the psbA mRNA in the HCF173-ctHA sample. The factor of psbA mRNA enrichment correlates with the different amounts of purified HCF173-tHA (Figure 13C). Therefore, we assume that the HCF173-tHA complex is indeed specifically associated with psbA mRNA.
View larger version (31K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 13. RT-PCR and Slot Blot Hybridization Analysis of Affinity Chromatography Eluates.
Anti-HA affinity chromatography was performed with extracts from wild-type and HCF173-ctHA plants. RNA was purified from the elution samples.
(A) Relative amounts of distinct plastidic transcripts were determined by RT-PCR, DNA gel blot analysis, and phosphor imaging. The ratios of transcript amounts purified from HCF173-ctHA plant extracts and wild-type plant extracts were calculated. Results from three independent experiments are shown.
(B) Purified RNA was submitted to slot blot hybridization with probes for psbA or psbB, respectively. After phosphor imaging, the ratios of transcript amounts purified from HCF173-ctHA and wild-type plant extracts were calculated and indicated adjacent to the signal. Results from two independent experiments are shown.
(C) HCF173-tHA protein that was purified in parallel to the RNA, which was isolated for slot blot hybridization.
|
|
|
DISCUSSION
|
|---|
PSII is a genetic mosaic consisting of nuclear- and plastid-encoded subunits. Therefore, many predominantly nuclear-encoded auxiliary factors are needed for correct and coordinated biogenesis of PSII. Here, we describe the nuclear mutant hcf173 of Arabidopsis, which is specifically affected in the accumulation of the PSII reaction center protein D1 due to defects in initiation of psbA mRNA translation. In addition, impaired stability of the psbA mRNA was observed. We report on the identification of the defective nuclear gene by map-based cloning and further discuss possible functions of the HCF173 protein based on biochemical studies with HCF173 fusion proteins.
The Mutant hcf173 Is Primarily Affected in D1 Expression
Spectroscopic measurements revealed a severe defect in PSII of the hcf173 mutant, while the cytochrome b6f complex and PSI are functional. The levels of thylakoid membrane proteins in the mutant corroborate these results. Whereas polypeptides of the cytochrome b6f complex, PSI (under low light), and the ATP synthase accumulated to nearly wild-type amounts, subunits of PSII and the oxygen-evolving complex were drastically reduced, though to a different degree. Together, these data indicate that the mutation in hcf173 primarily affects the accumulation of PSII, whereas the reduction of PSI protein amounts is caused by secondary effects of the mutational lesion. Radioactive labeling of de novo synthesized plastidic proteins indicates that the synthesis rate of the D1 protein was drastically impaired. This result was confirmed by the observation that only a minority of the psbA mRNA in hcf173 mutants was associated with polyribosomes. The reduced D1 synthesis rate was accompanied by a decreased stability of the psbA transcript, which resulted in a low steady state level of this RNA. Together, these data provide strong evidence that the nuclear-encoded factor HCF173 is primarily involved in translation of the psbA mRNA. The pattern of polysomal association in the mutant indicates that the factor possibly acts during initiation of translation. Moreover, HCF173 may directly or indirectly confer stability to the psbA mRNA. Alternatively, high RNA turnover could also be a secondary effect induced by altered ribosomal loading of this transcript in the mutant.
Recently, it was observed that decreased synthesis of D1 in Chlamydomonas results in impaired translation of the psbB transcript (Minai et al., 2006). This relationship, found also for other photosynthesis proteins, was termed control by epistasy of synthesis (Wollman et al., 1999; Choquet and Wollman, 2002). The polysome association analysis point to a comparable situation in the Arabidopsis mutant hcf173 because the polysome profile of the psbB transcript is significantly shifted to smaller polysomes and monosomes, indicating impaired translational efficiency of this transcript. However, pulse labeling experiments do not show a reduced synthesis rate of CP47 and further studies will be necessary to obtain a conclusive result.
The lack of accumulation of other PSII proteins tested indicated an increased degradation rate of these polypeptides in the mutant. This clearly shows that D1 is a prerequisite for stable PSII assembly, which is in accordance with the fact that D1 and D2 together with cytochrome b559 and PsbI represent the primary intermediate of PSII assembly (Rokka et al., 2005).
The Identified Nuclear Gene Encodes an SDR-Like Protein That Is Associated with Chloroplast Membranes
HCF173 was predicted to possess a functional chloroplast transit peptide in its N terminus. Immunoblot analyses of isolated chloroplasts confirmed the plastidic localization of HCF173. The separation of chloroplasts into membrane and soluble fractions revealed that minor amounts of the protein are localized in the stroma, whereas most of the protein is associated with the chloroplast membranes, indicating the existence of two different HCF173 pools. This finding was further supported by membrane flotation experiments in which the bulk of the HCF173 fusion protein comigrated with the chloroplast membranes. These experiments indicate that membrane association of HCF173 is a genuine feature of this protein. Salt washes with sodium chloride or sodium carbonate indicated peripheral membrane binding of HCF173.
The nuclear-encoded factor HCF173 is related to the SDR protein family. SDR proteins are enzymes, which use NAD(P)(H) as coenzymes, function as dehydrogenases, dehydratases, epimerases, and isomerases of a variety of substrates, and consist of 250 to 350 amino acids. To date, >3000 primary protein sequences are known, which typically show 15 to 30% identical residues in pairwise comparisons. Although the sequence similarity among SDR enzymes is low, their three-dimensional structures show highly similar /ß-folding patterns with a central ß-sheet (Kallberg et al., 2002; Oppermann et al., 2003). A secondary structure prediction revealed the same pattern for HCF173 in the SDR-like regions. Hence, the typical three-dimensional structure might be maintained in this protein despite the separation of the SDR-like domains by 200–amino acid residues. Classical SDR proteins show a Gly-rich motif (TGXXXGXG) and a conserved Asp residue in the N terminus, which are positioned in a ßß-turn and are needed for coenzyme binding (Filling et al., 2002). Additional residues are conserved in the ßß-turn (Filling et al., 2002) possibly due to structural reasons. Those residues are conserved in HCF173 as well; however, the Thr and the second Gly residue of the Gly-rich motif are changed (Figure 9). These findings suggest that HCF173 possesses no NAD(P)(H) binding activity. However, an alcohol dehydrogenase from Drosophila melanogaster was shown to have NAD+ binding activity, although the Gly-rich motif is not strictly conserved (Sofer and Ursprung, 1968). The enzymatic activity of SDR enzymes is associated with a YXXXK motif (Jörnvall et al., 1995), which forms a catalytic tetrad with a conserved Asn and a conserved Ser residue (Filling et al., 2002). This sequence motif is not conserved in Arabidopsis HCF173 and its homologs. From these observations, we predict that HCF173 does not exhibit its original dehydrogenase activity but has possibly gained other functions during evolution.
It has been proposed that several SDR-related proteins have lost their dehydrogenase activity but are adapted to other cellular functions. An interesting example with distant relationship to the SDR family is CSP41, which exhibits endonucleolytic activity (Baker et al., 1998). CSP41 is able to bind to specific RNAs. The protein has been proposed to initiate degradation of plastidic mRNAs and is able to cleave efficiently within the 3' stem-loop structures of psbA, rbcL, and petD mRNA in vitro (Yang and Stern, 1997). RNA binding is realized by the dinucleotide binding domain of the protein (Bollenbach and Stern, 2003), indicating that the domain has gained a function in chloroplast RNA metabolism. RNA binding via this domain has been shown for several NAD+-dependent dehydrogenases and (di)nucleotide binding metabolic enzymes that play an additional role in RNA metabolism (Hentze, 1994; Nagy et al., 2000). Therefore, HCF173 might also be an RNA binding protein, which possibly interacts with the psbA mRNA, affecting its translational activity and possibly also its stability.
HCF173 Is Part of a High Molecular Weight Complex That Contains psbA RNA
Blue native gel electrophoresis indicated that HCF173 is part of a high molecular weight complex, as are several other proteins, which are involved in chloroplast RNA metabolism and/or translation (Fisk et al., 1999; Vaistij et al., 2000; Ostheimer et al., 2003; Sane et al., 2005; Schmitz-Linneweber et al., 2005). Affinity purification of this HCF173-containing complex revealed that the psbA mRNA is specifically bound to this complex. What could be the functional significance of this binding to the psbA mRNA?
It is known that several proteins are enhancers of RNA translation and/or are necessary for their stability (Barkan and Goldschmidt-Clermont, 2000; Monde et al., 2000; Choquet and Wollman, 2002). Binding occurs via the UTRs of the RNAs mainly in adenine- and/or uracil-enriched sequences and in dependence of stem-loop structures (Klaff et al., 1997; Yang and Stern, 1997; Yohn et al., 1998a; Nickelsen et al., 1999; Ossenbühl and Nickelsen, 2000; Shen et al., 2001; Nakamura et al., 2003; Barnes et al., 2004). Possibly, the binding of the HCF173 complex to the psbA mRNA is a prerequisite for successful initiation of translation. In an in vitro translation system of tobacco (Nicotiana tabacum), it has been shown that psbA mRNA translation is dependent on several cis-elements, including RBS1 and RBS2, which are thought to function like a bipartite Shine-Dalgarno sequence for ribosome binding (Hirose and Sugiura, 1996). A third cis-element, the so-called AU-box (UAAAUAAA), is located between RBS1 and RBS2. This sequence is essential for translation of the tobacco psbA mRNA, and it is thought to bind one or more trans-acting factors (Hirose and Sugiura, 1996). The AU-box is conserved in the 5'-UTR of the psbA mRNA of Arabidopsis, and it is conceivable that HCF173 is able to bind this sequence motif via its dinucleotide binding site (see above) acting as an enhancer of translation initiation. Moreover, this protein binding might protect the psbA mRNA against exonucleolytic degradation and/or masks potential endonucleolytic cleavage sites of CSP41 or RNase E/G-like proteins that are thought to initiate chloroplast mRNA decay (Bollenbach et al., 2004).
Several investigations have shown that D1 protein synthesis is coupled to chloroplast membrane systems (Aro et al., 1993; Zhang et al., 2000). Therefore, the observed membrane binding of HCF173 gives rise to the hypothesis that the HCF173 complex could be necessary for membrane targeting of psbA mRNA translation. Possibly, the HCF173 complex is required for association of the psbA mRNA with chloroplast membranes, and this might be a prerequisite for efficient translation of the transcript. A similar tethering function was proposed for the RNA binding protein RBP63 from C. reinhardtii that binds to psbA mRNA and stroma thylakoid membranes (Ossenbühl et al., 2002). The existence of HCF173 in the stroma fraction does not contradict this hypothesis. Rather, this protein pool might be involved in binding and trafficking of the stroma-localized psbA transcripts to the chloroplast membranes. A detailed study of HCF173 localization inside the chloroplast would contribute to test these assumptions. Moreover, for a functional analysis, it will be of great importance to identify the components of the HCF173 complex. To this end, specific protein tags have been developed to isolate interacting partners of HCF173 by affinity purification.
Taken together, we were able to show that HCF173 is weakly related to the SDR superfamily and is mainly attached to chloroplast membranes where it acts as a part of a high molecular mass complex. Our investigations characterized HCF173 as a nuclear-encoded factor of higher plants that is essential for translation of the D1 encoding psbA mRNA. Whether the protein is involved in basic aspects of D1 synthesis or instead is part of the regulatory apparatus adjusting psbA mRNA translation remains to be determined in future studies.
|
METHODS
|
|---|
Growth Conditions
Wild-type and mutant seedlings were grown on sucrose-supplemented Gelrite medium under a 16-h photoperiod at a photon flux density (PFD) of 50 to 70 µmol·m–2·s–1 (Meurer et al., 1998). The mutant hcf173 was selected from a collection of EMS-induced mutants. Mutant plants exhibiting the high chlorophyll fluorescence phenotype were identified under UV light as described (Meurer et al., 1996b). For blue native PAGE and affinity chromatography experiments, plants were grown on soil (Floraton-I; Floragard) in a growth chamber at a PFD of 50 to 70 µmol·m–2·s–1 and a temperature of 21 to 23°C. Plants were illuminated for 8 h (short-day conditions) or 16 h (long-day conditions) per day, respectively.
Spectroscopic Analyses
Chlorophyll fluorescence induction and P700 absorption measurements were performed as described by Meurer et al. (1996b) with the following modifications. Saturating light pulses had a length of 1 s, and actinic light had a PFD of 200 µmol·m–2·s–1. For the P700 absorbance kinetics, the pulse amplitude–modified fluorometer was modified with the dual-wavelength emitter detector unit ED-P700DW-E (Walz).
SDS-PAGE and Immunoblotting
Crude leaf proteins (Meurer et al., 1996b) from wild-type and mutant plants were separated by SDS-PAGE according to Schägger and von Jagow (1987) and transferred onto nitrocellulose membranes (PH79, 0.1 µm; Schleicher and Schüll). Blotted proteins were immunodecorated with specific antibodies as listed by Meurer et al. (1996b) and visualized by chemiluminescence (ECL; Amersham Biosciences). The fusion protein HCF173-TAP was detected by peroxidase-antiperoxidase soluble complex (Sigma-Aldrich). The protein HCF173-tHA could be identified by two monoclonal antibodies (HA.11 [BabCO] or anti-HA high affinity antibody peroxidase conjugate [Roche]).
In Vivo Labeling of Chloroplast Proteins and Immunoprecipitation
The in vivo labeling and immunoprecipitation reactions were performed as described by Lennartz et al. (2001) with the following modifications. Instead of 5 µL of specific antibody, 10 µL were used for immunoprecipitation to ensure quantitative immunoprecipitation. To solubilize membrane proteins for immunoprecipitation, 2% (w/v) SDS was used. The antibody-antigen-protein A complex was dissociated by incubation in gel loading buffer (100 mM Na2CO3, 10% [w/v] sucrose, 50 mM DTE, and 2% SDS). After SDS-PAGE, radiolabeled proteins were analyzed by autoradiography.
RNA Isolation and Gel Blot Analysis
Isolation of total leaf RNA was performed according to Westhoff et al. (1991). For small amounts of RNA or RNA from seedlings, the RNeasy plant mini kit (Qiagen) was used. Poly(A)+ RNA was obtained by incubation of total leaf RNA from wild-type and mutant plants with OligoTex slurry (Qiagen).
Electrophoresis, gel blot analysis, and hybridization were performed as described before (Westhoff et al., 1991). Hybridization was done with specific RNA probes listed by Westhoff and Herrmann (1988) and Meurer et al. (1996b). Gene-specific probes for At1g16610, At1g16630, and At1g16720 were obtained by first-strand cDNA synthesis with total leaf RNA from wild-type plants as template and subsequent amplification by PCR with the following primer pairs: 16610-H2 (5'-CTCGTTCCAGTTCCAGATCTCG-3') and 16610-R1 (5'-CGATGGGGAGAATCAGGTGATCGC-3'); 16630-H2 (5'-CTTGGTGTGGCGTATCTGTTGTTGG-3') and 16630-R1 (5'-GGCTACTCCGGTTTGTTCTTGCTGG-3'); 16720-H1 (5'-CTACTGTGCGACTGCTCGGTCTACG-3') and 16720-R1 (5'-CTCCCGAACCAGTACACGATACCAG-3').
Polysome Purification
Polysome fractions were prepared as described previously (Barkan, 1993). To disrupt polysomes, EDTA was added to the extraction buffer according to Lu and Hanson (1996).
Measurement of mRNA Stability after Inhibition of Transcription with Actinomycin D
Arabidopsis thaliana wild-type and mutant plants were harvested under water and incubated for 0, 2, 6, 12, 18, and 24 h in 200 µM actinomycin D (Sigma-Aldrich). Prior to this treatment, plants were preincubated in actinomycin D for 30 min to allow proper distribution of the antibiotic. Plants were preilluminated for 4 h before actinomycin D treatment and were kept in the light during incubation. Plants incubated with water served as a control. Total RNA was isolated and quantified after gel blot analysis and hybridization with radiolabeled probes specific for psbA and rbcL. PsbA and rbcL mRNA amounts were normalized using the 16S rRNA levels as internal standard.
Mapping of the hcf173 Mutation
A mapping population was obtained by crossing wild-type plants of the ecotype Landsberg erecta (Ler) with heterozygous hcf173 plants, ecotype Columbia (Col). Plants of the F2 and the F3 generation were examined. The linkage group of the hcf173 mutation was determined with a set of microsatellite and CAPS markers (Konieczny and Ausubel, 1993; Bell and Ecker, 1994). For fine-mapping, the mapping population was examined with markers listed in The Arabidopsis Information Resource database and markers designed with the information of the Cereon Arabidopsis Polymorphism Collection (www.arabidopsis.org.cereon; Jander et al., 2002). The region of the mutation locus was narrowed down between the two markers SNP711/918 (SNP711/918-H, 5'-CATACAGAACACATACAGCAGCCCCG-3'; SNP711/918-R, 5'-AGGATGGTTCTGCTTCATTCCTGAGC-3'; polymorphism, SspI restriction site in Col) and SNP872 (SNP872-H, 5'-CTTGAGGCTTTGCGAGATGGTGAGC-3'; SNP872-R, 5'-ACATGTGAGCACACTCAAAGGGAGG-3'; polymorphism, RsaI restriction site in Ler).
Cloning of HCF173 cDNA
Full-length cDNA of At1g16720 was obtained by the use of the SMART RACE cDNA amplification kit (Clontech). Poly(A)+ RNA was used as template in 5'-rapid amplification of cDNA ends (RACE) and 3'-RACE reactions producing overlapping cDNA fragments. Full-length cDNA was constructed via a KpnI restriction site in the overlapping regions. The complete cDNA was subcloned into the vector pCR2.1-TOPO (Invitrogen) and into the binary Ti-vector pPEX001 (kindly provided by B. Reiss, MPIZ Cologne) and renamed pPEXAtHCF173-1. Nucleic acid and protein sequences were analyzed with the programs MacMolly Tetra 3.8 (Softgene) and ClustalW (Thompson et al., 1994). For localization predictions, the internet tools ChloroP (Emanuelsson et al., 1999) and WolfPSORT (Horton et al., 2006) were used. The BLAST server (Altschul et al., 1994), the conserved domain database (Marchler-Bauer et al., 2005), the Pfam database (Bateman et al., 2004), and the SMART database (Schultz et al., 1998; Letunic et al., 2004) were used for protein domain predictions.
Complementation of the Mutant hcf173
For complementation of the hcf173 phenotype, three different constructs were produced. Construct At-HCF173-2 consisted of the HCF173 cDNA (beginning with the start codon), which was amplified by PCR with the primers 16720-5'-SmaI-H (5'-TCATCCCGGGATGGTGGGTAGTATTGTTGG-3'; the SmaI restriction site is underlined), 16720-3'-BamHI-R (5'-CACGGATCCATGCGTATCAGAGCTTGAG-3'; the BamHI restriction site is underlined), and pPEXAtHCF173-1 as template.
The construct At-HCF173-TAP consisted of the HCF173 cDNA beginning with the start codon fused to the nucleic acid sequence for a C-terminal TAP tag (Puig et al., 2001). The TAP tag comprises a calmodulin binding domain, a TEV protease cleavage site, and two IgG binding domains of the Staphylococcus aureus protein A. The TAP tag fragment was obtained by restriction of the pBS1479 vector (Puig et al., 2001) with the enzymes NcoI and BamHI. The HCF173 fragment was amplified by PCR with the primers 16720-5'-SmaI-H and 16720-3'-NcoI-R (5'-CTGACCATGGTGTGTTCTTCTCAAGTACAG-3'; the NcoI restriction site is underlined) and pPEXAtHCF173-1 as template.
The fusion gene At-HCF173-tHA was composed of the HCF173 cDNA beginning with the start codon and the nucleic acid sequence for a C-terminal triple HA tag. The HA tag is an epitope tag, which is derived from the influenza hemagglutinin protein. The triple HA fragment was amplified with the oligonucleotides 3xHANcoI-H (5'-GACTGCCATGGTACCCATATGACGTTCCAGAC-3'; the NcoI restriction site is underlined) and 3xHABamHI-R (5'-CAGTGGATCCTCAAGCGTAGTCAGGTACGTCG-3'; the BamHI restriction site is underlined). The plasmid Spa1g3xHA-pBS served as the template, which was kindly provided by U. Hoecker (University Cologne, Germany). This vector contains the triple HA sequence according to Sato and Wada (1997). The HCF173 fragment was obtained as described for At-HCF173-TAP.
All three cDNA constructs were cloned into the binary Ti vector pPEX001 via the SmaI and BamHI restriction sites (pPEXAtHCF173-2, pPEXAtHCF173-TAP, and pPEXAtHCF173-tHA) and thereby fused to the 35S promoter of Cauliflower mosaic virus. Subsequently, they were transferred to heterozygous hcf173 plants by Agrobacterium tumefaciens–mediated transformation according to the floral dip method (Clough and Bent, 1998). Complemented transgenic plants were named HCF173-c, HCF173-cTAP, and HCF173-ctHA.
Immunolocalization Studies
For a total leaf protein extract, HCF173-ctHA and wild-type plants were homogenized in protein isolation buffer (50 mM Tris/HCl, pH 8.0, 10 mM EDTA, 2 mM EGTA, and 10 mM DTE). Cell debris were separated by Miracloth filtration.
For chloroplast isolation, leaves of 3- to 4-week-old Arabidopsis wild-type, HCF173-cTAP, and HCF173-ctHA plants were homogenized in homogenization buffer (330 mM sorbitol, 50 mM HEPES/KOH, pH 7.8, 1 mM MgCl2, 2 mM EDTA, 0.1% [w/v] sodium ascorbate, and 0.25% BSA) with a razor blade on ice. The suspension was centrifugated at 4°C and immediately stopped when a speed of 5900g was reached. The pellet was resuspended in homogenization buffer. Chloroplasts were obtained by Percoll gradient centrifugation and lysed in ice-cold lysis buffer (10 mM HEPES/KOH, pH 7.8, 10 mM MgCl2, and 25 mM KCl) for 10 min. Membrane and stroma proteins were separated by 30 min of centrifugation at 17,000g at 4°C. Total leaf, total chloroplast, chloroplast membrane, and stroma proteins of all fractions were separated by SDS-PAGE and subsequently subjected to immunoblot analysis.
Strength of the membrane association of HCF173-tHA was tested by resuspending and sonicating pelleted membranes of lysed chloroplasts in lysis buffer or lysis buffer supplemented with 200 mM NaCl or 250 mM Na2CO3, respectively. Membranes were pelleted again for 10 min at 17,000g and 4°C. Proteins of the supernatants were obtained by precipitation with 15% (w/v) trichloroacetic acid.
For chloroplast membrane flotation, chloroplasts were lysed in 600 mM sucrose, 10 mM HEPES/KOH, pH 7.8, and 2 mM EDTA and ruptured with 15 strokes in a glass homogenizer. Lysed chloroplasts (500 µL) equivalent to 125 µg of chlorophyll were mixed with 2.6 M sucrose, 10 mM HEPES/KOH, pH 7.8, and 2 mM EDTA to a final sucrose concentration of 1.8 M and placed on the bottom of a centrifugation tube. The chloroplast fraction was overlaid with 2.05 mL of 1.2 M sucrose in 10 mM HEPES/K |
TOP
ABSTRACT
INTRODUCTION