Interaction between PVY HC-Pro and the NtCF1β-subunit reduces the amount of chloroplast ATP synthase in virus-infected tobacco

The photosynthetic rate of virus-infected plants is always reduced. However, the molecular mechanism underlying this phenomenon remains unclear. The helper component-proteinase (HC-Pro) of Potato virus Y (PVY) was found in the chloroplasts of PVY-infected tobacco, indicating some new function of HC-Pro in the chloroplasts. We generated HC-Pro transgenic plants with a transit peptide to target the protein to chloroplast. The HC-Pro transgenic tobacco showed a decreased photosynthetic rate by 25% at the light intensity of 600 μmol m−2 s−1. Using a yeast two-hybrid screening assay to search for chloroplast proteins interacting with HC-Pro, we identified that PVY HC-Pro can interact with the chloroplast ATP synthase NtCF1β-subunit. This interaction was confirmed by GST pull-down and co-immunoprecipitation assays. HC-Pro didn’t interfere with the activity of assembled ATP synthase in vitro. The HC-Pro/NtCF1β-subunit interaction might affect the assembly of ATP synthase complex. Quantitative western blot and immunogold labeling of the ATP synthase indicated that the amount of ATP synthase complex was decreased in both the HC-Pro transgenic and the PVY-infected tobacco. These results demonstrate that HC-Pro plays an important role in reducing the photosynthetic rate of PVY-infected plants, which is a completely new role of HC-Pro besides its multiple known functions.


Results
The photosynthetic rate of PVY-infected plants was reduced. PVY HC-Pro was reported to be present in the chloroplasts of virus-infected plants 46 and we confirmed this result in our present work. Wild-type tobacco was inoculated with PVY viruses and intact chloroplasts of PVY-infected plants were isolated. HC-Pro was detected at 12 to 14 days post inoculation (dpi) in the chloroplasts of PVY-infected plants with an anti-HC-Pro antibody raised against the peptide DYRRQPGVSRKCTSSKDGN in rabbit. The purity of isolated chloroplast was examined by western-blot detection of the widely expressed cytoplasmic protein β -Actin. The β -Actin was detected in the total leaf proteins, but not in the purified chloroplasts (Fig. 1A). The results demonstrated that there was no contamination of cytoplasmic proteins in our purified chloroplasts. Meanwhile, we determined the photosynthetic rate in wild-type, mock-inoculated and PVY-infected plants at 12 dpi at the light intensity of 600 μ mol m −2 s −1 . The results showed that the photosynthetic rate of PVY-infected plants was reduced by 58% compared with that of wild-type plants, and the mock-inoculated plants presented a similar photosynthetic rate to the wild-type plants (Fig. 1B).
The photosynthetic rate of HC-Pro transgenic plants was reduced. We hypothesized that HC-Pro might interact with the host components involved in photosynthesis. To test this hypothesis, we generated transgenic plants retargeting HC-Pro to the chloroplasts to focus on the function of HC-Pro in chloroplasts. The transit peptide of the ribulose bisphosphate carboxylase small subunit (rbcs) was fused to the N-terminus of HC-Pro to target the protein to chloroplasts. The subcellular location of HC-Pro in the protoplasts of Arabidopsis demonstrated that the transit peptide functioned efficiently to lead HC-Pro into the chloroplasts ( Fig. 2A). Thus, the transit peptide was used to generate transgenic plants with HC-Pro retargeted to chloroplasts. Meanwhile, HC-Pro was fused to a GFP tag at its C-terminus to form a HC-Pro-GFP fusion. Plants transformed with the HC-Pro-GFP fusion retargeted to chloroplasts were referred to as HC-Pro transgenic plants. Intact chloroplasts of HC-Pro transgenic plants were isolated and the accumulation of HC-Pro in the chloroplasts of transgenic tobacco was detected with an anti-GFP antibody. The purity of isolated chloroplast was examined by western-blot detection of the widely expressed cytoplasm protein β -Actin (Fig. 2B). We also generated transgenic plants with the empty vector that contained only the transit peptide and the GFP tag, which was referred to as empty-vector transgenic plants, as a control for HC-Pro transgenic plants in addition to the wild-type control. The photosynthetic rates of the wild-type, empty-vector transgenic and the HC-Pro transgenic plants were then determined. The photosynthetic rate of the HC-Pro transgenic plants was reduced by 25% on average compared to that of the wild-type plants at the light intensity of 600 μ mol m −2 s −1 , while plants transformed with the empty vector presented a similar photosynthetic rate as the wild-type plants (Fig. 2C).
PVY HC-Pro and the NtCF 1 β-subunit interacted in yeast cells. The  with PVY HC-Pro. We identified HC-Pro-interacting proteins from a Nicotiana tabacum cDNA library constructed by our lab using the tobacco leaves 38 . Among the obtained positive clones, one encoded a polypeptide containing 208 amino acids that shared 100% identity with the tobacco chloroplast CF 1 β -subunit. The full-length gene sequence of the tobacco CF 1 β -subunit (NtCF 1 β-subunit) was cloned into the pMD18-T vector (Takara) and verified by sequencing. Based on sequence alignment, the cloned sequence exhibited 100% identity with a tobacco chloroplast DNA sequence (GenBank accession no: Z00044.2) on file in GenBank.

The interaction between HC-Pro and the NtCF 1 β-subunit was further verified both in vitro and in vivo.
To verify the interaction between HC-Pro and the NtCF 1 β -subunit, a pull-down assay was employed. The HC-Pro and NtCF 1 β -subunit proteins were expressed in E. coli and then we purified them. The NtCF 1 β -subunit was fused to a GST tag, and HC-Pro was expressed with a 6 × His tag. The GST-tagged NtCF 1 β -subunit was bound to Glutathione Sepharose 4B and incubated with His-HC-Pro. After washing the beads, His-HC-Pro was detected with an anti-His antibody. HC-Pro was pulled down by NtCF 1 β -subunit but not the GST tag alone (Fig. 3B). The pull-down assay confirmed the interaction between HC-Pro and the NtCF 1 β -subunit in vitro.
To confirm the interaction between HC-Pro and NtCF 1 β -subunit in vivo, intact chloroplasts from wild-type, empty-vector transgenic and HC-Pro transgenic plants were isolated, and a co-immunoprecipitation assay was performed using the chloroplast proteins. Because PVY HC-Pro was fused to a GFP tag at its C-terminus in the transgenic plants, an anti-GFP antibody was used to precipitate the HC-Pro/NtCF 1 β -subunit complex. The results showed that the NtCF 1 β -subunit co-immunoprecipitated with HC-Pro in the chloroplasts of transgenic plants in vivo but not in the The photosynthetic rate of PVY-infected plants was significantly reduced by 58% compared with that of the wild-type plants (Student's t-test, **P < 0.01). There was no difference between the wild-type and mock-inoculated plants (Student's t-test, P > 0.05). Three biological repeats were done.
chloroplasts of both the wild-type and empty-vector transgenic plants (Fig. 3C), thus demonstrating the interaction between HC-Pro and the NtCF 1 β -subunit in chloroplasts of living cells.
The full-length HC-Pro and the NtCF 1 β -subunit mutants were co-transformed into yeast cells. In the yeast two-hybrid system, the NtCF 1 β -subunit2/HC-Pro transformant was able to grow on PVY HC-Pro had no effect on the enzymatic activity of ATP synthase. To further elucidate the biological significance of the interaction between HC-Pro and the NtCF 1 β -subunit, we purified PVY HC-Pro and investigated its effect on the enzymatic activity of ATP synthase. The CF 1 complex of ATP synthase from spinach was isolated and its ability to hydrolyze ATP was examined. Purified HC-Pro was added to the CF 1 complex to test its influence on the enzyme activity of ATP synthase. The hydrolytic activity of ATP synthase was not disturbed by purified HC-Pro (Fig. 5A). There were no significant differences between the activity of CF 1 alone and CF 1 with the HC-Pro protein (Student's t-test, P > 0.05).
Active thylakoid membranes were extracted from spinach and the ATP synthetic activity was examined. Purified HC-Pro was added to the membrane extract to examine its influence on the synthetic activity of ATP synthase. Again, no disturbance in the synthetic activity was observed (Fig. 5B). There were no significant differences between the synthetic activity of ATP synthase alone and ATP synthase with the HC-Pro protein (Student's t-test, P > 0.05).

The content of ATP synthase was reduced in HC-Pro transgnic plants. The in vitro biochemical
reaction suggested that HC-Pro did not interfere with the enzymatic activity of chloroplast ATP synthase. Because CF 1 is a complex with several subunits, we assumed that the interaction between HC-Pro and the NtCF 1 β -subunit might affect the assembly of the CF 1 complex and could thus reduce the amount of ATP synthase in HC-Pro transgenic plants. To test this hypothesis, both quantitative western blot and immunogold labeling methods were used to determine the amount of ATP synthase.
First, we calculated the number of ATP synthase molecules in wild-type and HC-Pro transgenic plants. Typical immunogold labeling TEM images of wild-type plants and HC-Pro transgenic plants were shown (Fig. 6A). The amount of ATP synthase was found to be reduced by 16% on average in HC-Pro transgenic tobacco compared with the wild-type plants (Student's t-test, *P < 0.05, **P < 0.01) (Fig. 6B). Meanwhile, intact chloroplasts of the wild-type, empty-vector transgenic and HC-Pro transgenic plants were isolated and chloroplast proteins were extracted. The content of ATP synthase was determined by quantitative western blot using an antibody against the essential α -subunit of CF 1 (AtpA). The large subunit of Rubisco (RbcL) was used as a loading control to demonstrate equal input in all lanes. The amount of ATP synthase in HC-Pro transgenic plants was reduced by 14% compared to the wild-type plants and the amount of ATP synthase in empty-vector transgenic plants was nearly the same as the wild-type plants (Fig. 6C). The levels of the AtpA and RbcL proteins were quantified using ImageJ software.
Scientific RepoRts | 5:15605 | DOi: 10.1038/srep15605 NtCF 1 β -subunit co-localized in the chloroplasts of PVY-infected tobacco (Fig. 7A). And we also calculated the number of ATP synthase molecules in mock and PVY-inoculated tobacco. Typical immunogold labeling TEM images were presented (Fig. 7B). The amount of ATP synthase was observed to be significantly reduced, by 39% on average, in PVY-infected tobacco compared with the mock-inoculated plants (Student's t-test, **P < 0.01) (Fig. 7C). Intact chloroplasts of the wild-type, mock-inoculate and PVY-infected plants were isolated and the content of ATP synthase was also determined using an antibody against the essential α -subunit of CF 1 (AtpA). The large subunit of Rubisco (RbcL) was used as a loading control to demonstrate equal input in all lanes. The amount of ATP synthase in PVY-infected plants was diminished by 41% compared to the wild-type plants and the amount of ATP synthase in mock-inoculated plants was nearly the same as the wild-type plants (Fig. 7D). The levels of the AtpA and RbcL proteins were quantified using ImageJ software.
The above results showed that the interaction of HC-Pro and the NtCF 1 β -subunit did not interfere with the enzymatic activity of ATP synthase but instead decreased the amount of ATP synthase in both HC-Pro transgenic plants and PVY-infected plants. Our results showed a new function of PVY HC-Pro in viral attacks on the photosynthetic system of plants.

Discussion
Viral infection is a complicated process that involves both viral components and host factors. As demonstrated in the integrated molecular model of plant-virus interactions presented by Elena et al. 47   to wild-type plants, as was observed in PVY-infected plants. The interaction between HC-Pro and the NtCF 1 β -subunit led to a decreased number of chloroplast ATP synthase complexes in both the HC-Pro transgenic and PVY-infected tobacco. Although the HC-Pro/NtCF 1 β -subunit interaction did not affect the enzymatic activity of ATP synthase, the abundance of chloroplast ATP synthase was greatly reduced in both PVY-infected and HC-Pro transgenic plants. As a result, the total ATP synthase activity was decreased and the energy supply for photosynthesis may be strongly disturbed, leading to the reduced photosynthesis in PVY-infected plants. ATP synthase activity has been reported as a potentially limiting  The number of ATP synthase molecules per area of the chloroplast was employed for comparison of the differences between mock-inoculated and PVY-infected plants. Ten sections were used for three biological replicates. Calculations were performed for over 15 chloroplasts with each sample. The quantity of ATP synthase was found to be significantly reduced, by 39% on average, in the PVY-infected plants (Student's t-test, **P < 0.01). (D) Quantitative western-blot of ATP synthase. Intact chloroplasts were isolated and chloroplast proteins were loaded for western-blot analysis. ATP synthase abundance was determined from the contents of the essential α -subunit of CF 1 . Anti-AtpA was used to detect the CF 1 α -subunit. Estimates of AtpA protein levels relative to the large subunit of Rubisco (RbcL) in the mockinoculated plants (Mock) and the PVY-infected plants (PVY) compared with those in wild-type plants (WT) were shown at the bottom. factor of photosynthesis 50,51 . Thus, the photosynthetic capacity of the PVY-infected plants was likely disrupted due to this interaction. The molecular approach applied in this study illuminated the phenomenon observed in PVY-infected plants and added a new function of HC-Pro besides its known multiple functions. The reduced photosynthetic rate is a comprehensive result of multiple host-virus interactions. The interaction between HC-Pro and the NtCF 1 β -subunit identified in this study is one of the causes of the decreased photosynthetic rate observed in potyvirus-infected plants. The mechanism underlying the impairment of photosynthesis in virus-infected plants has been explored for many decades. The disrupted activity of photosystem II (PSII) was previously thought to be the main cause of this reduced photosynthetic activity 4,7,52,53 . Our results presented here demonstrated some disturbance of photosynthetic rate correlated with ATP synthesis, thus deepening our understanding of viral attack on the photosynthesis system.
The chloroplast F-ATPase stromal region CF 1 contains five subunits, α , β , γ , δ and ε , that are assembled in a 3:3:1:1:1 stoichiometry 54 . Here, we have shown that PVY HC-Pro can interact with the NtCF 1 β -subunit. However, HC-Pro does not bind directly to the ATP-binding site on the β -subunit. Rather, it binds to the N-terminus (residues 1-95) of the β -subunit. In vitro biochemical assays indicated that this interaction did not affect the enzymatic activity of chloroplast ATP synthase. The known inhibitors of ATP synthase are all small molecules or short peptides, such as singlet oxygen 55 , tentoxin 56 and melittin 57 . The large HC-Pro protein would not fit into the inhibition sites for this enzyme, and this explains why HC-Pro did not affect the enzymatic activity of ATP synthase. On the other hand, the correct assembly of ATP synthase is the basis for its structure and function as a complete enzyme. HC-Pro interacts with the N-terminus of the β -subunit, which is important for its assembly into the CF 1 complex 20 . Thus, the interaction between HC-Pro and the CF 1 β -subunit could disturb the assembly of ATP synthase and the level of ATP synthase would be reduced due to this interaction in PVY-infected tobacco.
Virus-plant interactions represent an interesting topic involving many different stages and changes that take place both in plants and viruses. The interactions between viral proteins and plant factors are key elements of viral infections, viral movements and the development of associated symptoms. The 11 PVY proteins have been shown to interact with different host proteins during different stages of viral infection [58][59][60][61] . We verified the interaction between HC-Pro and the NtCF 1 β -subunit here. And the biological significance of this interaction was further demonstrated. HC-Pro has no effect on the enzyme activity of assembled ATP synthase, instead, the amounts of ATP synthase in both the HC-Pro transgenic and PVY-infected tobaccco were significantly reduced. HC-Pro may interfere with the assembly of NtCF 1 β -subunit into the whole enzyme, leading to the reduced amount of ATP synthase and finally caused the reduction of photosynthetic rate. The findings presented here regarding the interaction of HC-Pro with the β -subunit of chloroplast ATP synthase have contributed to our understanding of one of the causes of the reduced photosynthetic rate observed in PVY-infected plants.

Methods
Plant material and growth conditions. Wild-type Nicotiana tabacum tobacco was inoculated with PVY virus according to Mojca Milavec 62 with a few modifications. Gloved fingers were wetted with sap from the PVY-infected tobacco, and inoculation was carried out by gently rubbing the bottom leaves with tripolite. Sap from healthy tobacco was used as a mock inoculation control. The inoculated plants (PVY-inoculated) and corresponding control plants (wild-type and mock-inoculated) were grown in a constant incubator at 22 °C under a 16-h-light/8-h-dark photoperiod at the light intensity of 50 μ mol m −2 s −1 . PVY-infected plants at 12 days post-inoculation (dpi), together with control plants, were employed for future analysis.
Wild-type and HC-Pro transgenic Nicotiana tabacum plants were grown under standard greenhouse conditions at 25 °C under a 16-h-light/8-h-dark photoperiod at the light intensity of 100 μ mol m −2 s −1 . N. tabacum was transformed using Agrobacterium tumefaciens strain EHA105 through the leaf disc method. The transit peptide from the ribulose bisphosphate carboxylase small subunit (rbcs) was fused to the N-terminus of HC-Pro (the fusion was referred to as TP-HC-Pro) to enable its entry into the chloroplasts. The TP-HC-Pro gene was cloned into pCAMBIA-1302 vector through the restriction site NcoI/SpeI to generate the HC-Pro transgenic plants. The positive transformants were employed for future analysis.

Subcellular localization of HC-Pro in Arabidopsis protoplasts.
To test the efficiency of the transit peptide of rbcs, the TP-HC-Pro gene was cloned into the modified pE3025 vector (which contained GFP instead of RFP) using the SacI/KpnI site to form the pE3025-TP-HC-Pro vector to analyse the subcellular localization of HC-Pro. Rosette leaves of 4-week-old Arabidopsis plants grown under short-day conditions were used for isolation of protoplasts. The pE3025-TP-HC-Pro vector was transformed into Arabidopsis protoplasts according to a previously described protocol 63 . GFP fluorescence in the protoplasts was visualized using a confocal laser scanning microscope (LSM510, Carl Zeiss).

Isolation of chloroplasts and extraction of chloroplast proteins.
Intact chloroplasts were isolated using a chloroplast isolation kit (Sigma-Aldrich) according to the manufacturer's instructions. Chloroplasts were collected after centrifugation at 4 °C and 1000 × g. Intact chloroplasts were separated from the broken chloroplasts by centrifugation on top of a 40/80% Percoll gradient. The band formed by intact chloroplasts at the interface between the 40% and 80% Percoll layers was collected and used for the following experiments. Chloroplast proteins were extracted with lysis buffer A (50 mM Tris-Cl, pH 8.0; 150 mM NaCl; 0.5% Triton; 1× Protease Inhibitor Cocktail; 1 mM PMSF) and diluted with the same volume of buffer B (50 mM Tris-Cl, pH 8.0; 150 mM NaCl; 1× Protease Inhibitor Cocktail; 1 mM PMSF). The lysate was centrifuged at 4 °C and 5000 × g for 10 min to separate the membranes and the supernatant. The chloroplast proteins in the supernatant were used for future analysis.
Measurement of photosynthetic rates. The photosynthetic rates of wild-type, mock-inoculated and PVY-infected tobacco plants were measured using the LI-6400 XT Portable Photosynthesis System according to the user manual. Five values were collected at the light intensity of 600 μ mol m −2 s −1 for each plant. Three biological repeats were done.
The photosynthetic rates of wild-type, empty-vector transgenic and HC-Pro transgenic plants were measured using the same method except that values were collected at different light intensity from 50 to 600 μ mol m −2 s −1 to make the light intensity curve. Three independent lines were used for the measurement of photosynthetic rates, and three biological repeats were conducted for each line.

Identification of the interaction between HC-Pro and NtCF 1 β-subunit in yeast.
We used a yeast two-hybrid system to search for HC-Pro-interacting proteins in an N. tabacum cDNA library according to the user manual for the BD Matchmaker Library Construction and Screening Kit (BD Biosciences). The cDNA library was constructed by our lab previously using the mature leaves of N. tabacum L. cv. Xanthi NN about four weeks' old 38 . The tobacco plants were grown under standard greenhouse conditions at 25 °C under a 16-h-light/8-h-dark photoperiod. HC-Pro was used as a bait to screen the library. In the obtained positive clones, the chloroplast CF 1 β -subunit (NtCF 1 β -subunit) was found to interact with HC-Pro. The full-length coding sequences of the NtCF 1 β-subunit was cloned into pGADT7 via the EcoRI/XhoI sites to form the pGADT7-NtCF 1 β -subunit plasmid. The reconstructed plasmid was co-transformed with the pGBKT7-HC-Pro vector described previously 36 , into S. cerevisiae AH109 cells to verify their interaction in yeast. Appropriate negative controls were produced at the same time.
GST pull-down assay in vitro. The full-length coding sequence of HC-Pro was cloned into the pET-30a vector via the EcoRI/NotI sites to form the pET-30a-HC-Pro vector, and the full-length coding sequence of NtCF 1 β-subunit was cloned into the pGEX-4T-1 vector via the EcoRI/SalI sites to form the pGEX-4T-1-NtCF 1 β -subunit plasmid. The proteins were induced at 16°C for 12-14 hours in E. coli BL21 cells. HC-Pro was expressed with a 6 × His tag (His-HC-Pro), and the NtCF 1 β -subunit was expressed with a GST tag (GST-NtCF 1 β ). The purified GST-NtCF 1 β protein was bound to Glutathione Sepharose 4B (GE Healthcare) for 1 hour at 4 °C and then washed with lysis buffer (150 mM NaCl, 0.1 M Tris pH 8.0, 10% glycerol) three times. His-HC-Pro was added to the GST-NtCF 1 β -bound Sepharose and incubated at 4 °C for 2 additional hours. The Sepharose was then washed with the same lysis buffer three times. Samples were eluted from the Sepharose with the elution buffer (20 mM reduced glutathione in 50 mM Tris-Cl, pH 8.0) and prepared with 2× SDS loading buffer. The pulled down His-HC-Pro was detected with an anti-His antibody.
Co-immunoprecipitation assay with the chloroplast proteins in vivo. The chloroplasts of wild type, empty-vector transgenic and HC-Pro transgenic plants were isolated using the same method mentioned above. The chloroplast proteins were also extracted as mentioned. The chloroplast proteins in the supernatant were incubated with anti-GFP antibody overnight at 4 °C. The immunocomplex was captured with washed Protein A Agarose (Millipore) at 4 °C for 2 hours. The agarose beads were collected and washed with ice-cold PBS buffer three times. Samples were prepared with 1× SDS loading buffer and detected using a CF 1 β -subunit antibody (Product AS05085, Agrisera).
Detection of the enzymatic activity of ATP synthase. To determine the hydrolytic activity of ATP synthase, the F 1 component of chloroplast ATP synthase was isolated from spinach according to the method described by Richter 64 , and ATP was added as the reaction substrate. Following incubation at 37 °C for 5 minutes, the reaction was terminated with trichloroacetic acid (TCA), and the amount of Pi produced was determined by spectrophotometry at a wavelength of 740 nm.
To perform photophosphorylation assays, thylakoid membranes were extracted from spinach. An assay mixture containing 100 mM tricine-NaOH (pH 8.0), 10 mM NaCl, 100 μ M PMS, 10 mM MgCl 2 , Scientific RepoRts | 5:15605 | DOi: 10.1038/srep15605 6 mM ADP and 2 mM KH 2 PO 4 was added to the thylakoid membrane extract, and the reaction was initiated by illuminating the membrane with a halogen lamp in a tank of CuSO 4 for 2 minutes. The reaction was terminated by turning off the light and then quickly adding 1 ml of 0.5 M TCA. The amount of Pi remaining in the reaction system was again determined by spectrophotometry at a wavelength of 740 nm.
Immunogold labeling of HC-Pro, NtCF 1 -β subunit and ATP synthase. Samples were prepared from tobacco leaf tissue fixed in 2% paraformaldehyde and 1% glutaraldehyde. Following serial dehydration with ethanol, the samples were embedded in LR White resin as described previously 65 . Thin sections were cut using a Leica UC6i microtome. The grids were blocked with 1% BSA in PBS buffer for 2 hours. Then the grids were incubated with the primary antibody diluted in blocking buffer for 3 hours. After washing with PBS buffer three times for 10 minutes each, the grids were incubated with the gold-labeled secondary antibody diluted in PBS buffer for 1 hour. The same washing step was repeated. To enhance the contrast of the samples, the grids were stained with uranyl acetate for 15 minutes and then washed with milli-Q water three times for 10 minutes each.
The primary antibody used to count the number of ATP synthase was against the whole ATP synthase bought from Agrisera (Product AS08370). The secondary antibody was conjugated to 10-nm gold particles to label the ATP synthase. The anti-HC-Pro antibody was used to label HC-Pro in PVY-infected tobacco. The corresponding secondary antibody was conjugated to 10-nm gold particles. The 5-nm gold particle labeled secondary antibody was used to indicate the NtCF 1 -β subunit. The samples were viewed using a JEM-123O transmission electron microscope. For the counting of the number of ATP synthase, the chloroplast plan area and the number of gold particles were calculated using iTEM software, and the data were processed to indicate the particle number per area of the chloroplast.

Quantitative western-blot analysis. Chloroplast protein concentration was determined by
Bradford method. Twenty microgram protein were equal loaded. Protein samples were separated on a 12% SDS page and electroblotted to nitrocellulose membrane. The membrane was blocked for one hour with 5% nonfat milk in TBST and probed with anti-AtpA antibody against the essential α -subunit of CF 1 to estimate the amount of ATP synthase according to Rott et al. 50 . The large subunit of Rubisco (RbcL) was used as a loading control to demonstrate equal input in all lanes. The levels of the AtpA and RbcL proteins were quantified using ImageJ software. The amounts of ATP synthase relative to RbcL in each lane were caculated and the differences among the samples were used to demonstrate a change of ATP synthase level.