Characterization of DnaJ single mutants
Arabidopsis plants lacking a DnaJ protein AtJ8 (At1 g80920), AtJ11 (At4 g36040) or AtJ20 (At4 g13830) did not exhibit significantly different phenotypes compared to wild-type (WT) except for slightly stunted growth of the j11 and j20 mutants (Figure 1A and 1B). Photochemical efficiency of photosystem II (PSII) (Fv/Fm ratio) was not different between the WT and the DnaJ mutants under growth light (GL) conditions, whereas, it decreased somewhat more drastically in the mutants after exposure of 6 h to high light (HL) (1000 μmol photons m-2 s-1), especially in j11 and j20 as compared to that in WT (Figure 1C). When plants were returned to GL conditions, the PSII photochemical efficiency recovered quickly and no differences were found between the WT and mutants (Figure 1C). The other mutant lines for the AtJ11 and AtJ20 proteins exhibited similar phenotypes as described above (Additional file 1).
Localization of the three DnaJ proteins
In order to examine the localization of the three small DnaJ proteins, an antiserum for each protein was raised in rabbits using specific synthetic peptides. Despite purification of the antisera, we did not get good reactions using leaf total protein extracts (data not shown). However, as shown in Figure 2, the protein extracts from intact chloroplasts gave a specific band in WT around 17 kD, 15 kD, and 20 kD when the AtJ8, AtJ11 and AtJ20 antisera, respectively, were used, and importantly, the specific band was missing from the respective DnaJ mutant. This indicates that chloroplasts are at least one of the compartments containing these small DnaJ proteins in Arabidopsis. It should be noted that the size of each DnaJ protein in chloroplasts is somewhat lower than the predicted molecular mass (18.3-, 17.8- and 23.4-kD for AtJ8, AtJ11 and AtJ20, respectively). This is apparently due to the processing of the preprotein after import to chloroplast. In fact, Orme et al. reported that AtJ11 is located in chloroplast stroma and the mature protein has a molecular mass of 14.3 kD [7].
Capacity of CO2assimilation
To analyse whether the DnaJ proteins are involved in acquiring the maximal CO2 fixation capacity, we measured both the light response and CO2 response curves of the DnaJ mutants and WT. The light response curves showed the maximum CO2 assimilation rate at 500 μmol photons m-2 s-1 which then decreased with increasing photosynthetic photon flux density (PPFD) in both WT and the DnaJ mutants (Figure 3A). Compared to WT, the DnaJ mutants possessed lower CO2 assimilation, especially the j20 mutant. Relatively, the assimilation of j8 was only slightly lower as compared to WT, showing that the AtJ8 protein is less related to the light-dependent regulation of CO2 fixation. Nonetheless, the CO2 response curves revealed lower CO2 assimilation in j8 as compared to that in WT (Figure 3B). The A-Ci curves based on intracellular CO2 concentration less than 300 μmol mol-1 demonstrated a lower Rubisco activity in all three DnaJ mutants as deduced from lower slope values of the curves as compared to WT, especially for j8 (Figure 3C). Although the amount of the Rubisco protein (large subunit and small subunit) did not obviously differ between WT and the DnaJ mutants, an immunoblot analysis of Rubisco Activase showed reduced amounts of this enzyme under light conditions in the DnaJ mutants as compared to WT (Figure 3D).
Stabilization of PSII dimers and supercomplexes under high light illumination
Since the absence of one of the DnaJ proteins, AtJ8, AtJ11 or AtJ20, pronouncedly affected the photosynthetic capacity of respective mutants, we next investigated whether the DnaJ proteins are involved in regulation of the stability of the photosynthetic pigment protein complexes in the thylakoid membrane. Based on Blue-native gel electrophoretic (BN-PAGE) separation of thylakoid protein complexes (Figure 4A), the amount of PSII-LHCII supercomplexes was less in the DnaJ mutants than in WT after 6 h HL illumination (1000 μmol photons m-2 s-1). Immunoblotting of the BN-gels with D1 antibody more clearly showed the decrease of PSII supercomplexes in the mutants after the HL treatment. Moreover, the amount of PSII dimers also significantly decreased in the DnaJ mutants upon the HL treatment, especially in j11 and j20 (Figure 4A). To get more insights into the function of the three DnaJ proteins in the maintenance of the PSII oligomers, a long-term treatment under HL was employed. As shown in Figure 4B, the PSII supercomplexes completely disappeared both from WT and the DnaJ mutants whereas the PSII dimers were much more stable in WT than in the DnaJ mutants in the course of the long-term HL treatment. As compared to WT, the DnaJ mutants j11 and j20 showed a total disappearance of PSII dimers already during 24 h of HL treatment (Figure 4B), and clearly more of CP43 proteins had released from PSII complexes at this time point as compared to WT or the j8 mutant. As the total amounts of the D1, D2, CP43, CF1 and NDH-H proteins were similar in WT and the three mutants even after the HL treatment (deduced from PAGE and immunoblotting - see Additional file 2), it can be concluded that the three DnaJ proteins do not participate in the biosynthesis of individual PSII core proteins, but only provide stability for the PSII protein complexes.
Energy distribution between PSI and PSII and phosphorylation of the PSII-LHCII proteins
The 77 K chlorophyll fluorescence emission ratio F733/F685 was recorded as an indication of energy distribution between the PSI and PSII complexes (Figure 5A). The ratio of F733/F685 was slightly lower in the DnaJ mutants than in WT both when measured from dark acclimated and from GL acclimated plants. After HL illumination no clear differences in F733/F685 ratio were found between the WT and mutants with one exception, the ratio was higher in j11 as compared to that in WT after 500 μmol photons m-2 s-1 HL illumination (Figure 5A). To evaluate whether the phosphorylation of PSII proteins is related to redistribution of energy in plants lacking the DnaJ proteins, the phosphorylation levels of the major PSII phosphoproteins D1, D2, CP43 and LHCII were determined by immunoblotting with the P-Thr antibody. As can be seen in Figure 5B, only extremely weak phosphorylation of LHCII (P-LHCII) was detected in darkness and P-LHCII strongly accumulated in light conditions. Higher intensity light (1000 μmol photons m-2 s-1) decreased the level of P-LHCII but did this less efficiently in the DnaJ mutants than in WT (Figure 5B). Interestingly, LHCII was phosphorylated to the same level in all strains under GL and moderate HL (500 μmol photons m-2 s-1), despite clear differences in the 77 K fluorescence ratio under these two light conditions.
As to PSII core protein phosphorylation, under GL conditions the j8 and j11 mutants exhibited more P-CP43, P-D1 and P-D2 proteins as compared to WT while the j20 had less (Figure 5B). Under HL conditions (both 500 and 1000 μmol photons m-2 s-1) the j11 and j20 mutants had a clearly higher level of PSII core protein phosphorylation. A long-term HL illumination (1000 μmol photons m-2 s-1) experiment showed that fluctuations in phosphorylation of both the PSII core and LHCII proteins were characteristic for WT during acclimation to this HL condition. The j8 mutant showed similar fluctuations, though not as drastic as in WT (Figure 5C). The j11 and j20 mutants, however, differed from the WT and j8, showing clearly delayed and less obvious drop in the phosphorylation level of both the PSII core and LHCII proteins, which in WT and j8 occurred after 6 h illumination at HL whereas in j11 and j20, a less distinctive drop in phosphorylation was recorded after 12 - 24 h illumination at HL. Moreover in all DnaJ mutants, j8, j11 and j20, long HL illumination resulted in more drastic phosphorylation of the Cas protein (Figure 5C), a typical stress response of plants [17].
Gene expression profiles
Based on somewhat similar effects on photosynthetic parameters of the knockout of any of the three small chloroplast targeted DnaJ proteins, it was of interest to analyse the gene expression profiles of these mutants. The expression of about 1,200 genes showed more than two-fold changes in WT by HL treatment, and among those genes one third were upregulated (Figure 6, Additional file 3). It was interesting to note that the gene expression profiles of the mutants showed similarities under both GL and HL conditions to the HL-treated WT, although the expression levels somewhat varied in each mutant (Figure 6). More than half of genes changing expression were found to be coregulated between the DnaJ mutants, and all three mutants shared 556 and 687 coregulated genes under GL and HL, respectively, indicating their very similar response between the DnaJ mutants (Figure 7A). In each mutant, the expression of roughly 700 genes had changed independently of the growth light condition (Figure 7B). It is also worth noting that the j11 and j20 mutants showed more divergent gene expression (920 and 1047 genes, respectively) at growth light from that in WT as compared to j8 (560 genes) whereas after HL treatment the reverse situation was recorded (Figure 7B). However, although the three DnaJ proteins are localized in the chloroplasts, most of the genes related to thylakoid membranes were not affected by lacking of one of the small DnaJ proteins (Additional file 4).
More interestingly, the DnaJ mutants showed stress-related regulation of several genes even at GL conditions. Expression of a number of genes related to transcription, translation and cellular signaling and to enzymes participating in the control of reactive oxygen species (ROS) and in redox regulation resembled that observed in WT upon transfer to HL (Additional file 4). Nevertheless, the DnaJ mutants also showed unique gene expression patterns from those induced in WT by HL treatment, including upregulation of several distinct genes encoding transcription factors, heat shock proteins, DnaJ proteins as well as antioxidant and redox proteins, among others (Additional file 4). Additionally, by using the MapMan tool, it was found that changes in expression of several genes related to distinct regulation pathways were quite similar in the DnaJ mutants at GL conditions to those recorded in the HL-treated WT (Additional file 5). Several clustered genes related to different functions, including hormone metabolism, stress response, redox regulation, transcriptional regulation, and protein degradation, were visualized and the results show that almost the same numbers of genes were regulated by HL in WT or by the lack of one DnaJ protein, AtJ8, AtJ11 or AtJ20 at GL conditions (Additional file 6). Particularly, the genes related to ubiquitin and ubiquitin E3 presented a high correlation between the HL stress response in WT and the DnaJ protein knockout (Additional file 6).
Oxidative stress tolerance in the DnaJ mutants
Based on the cues from microarray results, we next tested some oxidative stress responses of the DnaJ mutants. At first, the H2O2 levels in the leaves of the DnaJ mutants and WT were detected using DAB (diaminobenzidine) as a substrate. Notably, the staining intensity and accordingly the level of H2O2 in the DnaJ mutants was lower as compared to WT, especially in plants illuminated under HL for 6 h (Figure 8A). Since ascorbate peroxidases (APXs) and chloroplast peroxiredoxins (PRXs) associated with the water-water cycle, are generally known protectants of chloroplasts against oxidative damage, the contents of these H2O2-detoxifying enzymes were evaluated by immunoblotting. As shown in Figure 8B, the amounts of all these enzymes, including tAPX (thylakoid APX), sAPX (stroma APX), cAPX (cytoplasmic APX), and two PRXs, PrxE (peroxiredoxin E) and 2-Cys Prx (2-cysteine peroxiredoxin), were pronouncedly higher in the three DnaJ mutants as compared to WT no matter whether the plants were subjected to darkness, GL or HL conditions before measurements. Nevertheless, higher amounts of these enzymes were present in the light conditions, especially in HL. These results suggest that the higher amounts of H2O2-detoxifying enzymes contributed to the lower H2O2 levels in the mutants.
To investigate the tolerance of the DnaJ mutants to oxidative stress, 50 μM methyl viologen (MV) was supplied to plants followed by illumination at 1000 μmol photons m-2 s-1 HL for 6 h, and the cellular ion leakage of whole plant rosettes was determined. Plants untreated with MV showed no differences in ion leakage between the DnaJ mutants and WT. In WT plants the MV treatment strongly enhanced ion leakage levels, whereas in the j11 and j20 mutant plants the ion leakage was only slightly increased, indicating that the mutants had better resistance to MV-induced oxidative stress (Figure 8C, Additional file 1). Although j8 exhibited similar levels of ion leakage as WT in MV treated plants, the oxidation level of leaf total proteins isolated from the mutant was less severe than that in WT after the HL treatment (Figure 8D). In general, the DnaJ mutants showed less oxidation of leaf total proteins, particularly the Rubisco protein, as a response to environmental light intensity changes as compared to WT (Figure 8D).