Proteomic characterization of the get3b mutant
For the GET3B locus, three T-DNA insertion lines are publicly available (Xing et al., 2017). We decided to analyze the get3b-2 mutant (from here on get3b ) with an insertion in the fourth intron of the allele. The line was homozygous for the mutation and had no transcript nor protein detectable by RT-PCR or western blot, respectively (Figure 1A-D).
To analyze the overall effect of the get3b mutation on chloroplast function, we examined get3b chloroplasts on a proteomic level. By using wild-type and get3b plastids, a label-free technique could be employed in a quantitative manner as previously described (Vermeulen et al., 2008). For this, we isolated plastids of the respective lines from 21-day-old A. thaliana seedlings and subjected these to LC-MS/MS.
The examination revealed a total of 999 proteinaceous hits. After subtraction of contaminants, 203 were classified as significantly depleted and 105 as enriched in get3b (Figure 1E). Intriguingly, not only nuclear encoded proteins, but also plastome encoded proteins were significantly affected. These covered a wide range of biological processes and were manually curated and categorized according to their cellular functions (Figure 1F, G). The majority of the significantly depleted hits (33.9 %) belonged to the category other metabolic pathways , which included components of the AA synthesis machinery as well as the post-translational modification machinery. Apart from that,chaperone/assembly factors (10.3 %), the translational machinery (9.8 % ribosomes plus 5.9 % transcription/translation ), the photosynthetic machinery (9.8 % light reaction plus 2.9 %dark reaction ) were mostly affected. Constituents of starch synthesis (5.4 %), the plastid redox regulation andtranslocation (3.4 % each), tRNA (2.9 %), proteases (2.4 %), Fe-S cluster synthesis (1.4 %), fatty acids andmembranes (0.9 % each) were also depleted to a significant extent in get3b (Figure 1F). Interestingly, one entire paralogous set of nuclear encoded proteins of the oxygen evolving complex (OEC; Allahverdiyeva et al., 2013) was reduced to a significant level. Furthermore, several components of the electron transport chain were also decreased. Additionally, soluble proteins involved in the Benson-Calvin cycle like several subunits of ribulose-1,5-bisphosphat-carboxylase/-oxygenase (RuBisCo) and their assembly factors were also negatively affected as well as STIC2, a factor involved in stromal targeting networks (Bédard et al., 2017). Strikingly, GET3B was detectable even though it was significantly depleted (Supplemental table S1), suggesting that the T-DNA insertion might lead to a knock-down rather than a knock-out of the respective gene.
The majority of the significantly enriched hits were attributed to thelight reaction of photosynthesis (39 %), followed by other metabolic pathways (21.9 %) and components of translocation machineries (12.4 %). Additionally, starch synthesis (8.5 %),proteases (3.8 %), constituents involved intranscription/translation (2.8 %), chaperone/assembly factors, redox regulation, Fe-S cluster biogenesis andfatty acid synthesis (1.9 % each) as well as elements involved in membrane modulation (0.9 %) were elevated to a significant extent (Figure 1G). Strikingly, nearly all LHCPs of PSII and a few of PSI were enriched in get3b , while the respective cpSRP targeting machinery was unaffected (Supplemental table S1).
The comparative proteomic data displayed that the loss or reduction of GET3B has a pervasive effect on the proteome of plastids. Interestingly, the photosynthetic apparatus seemed to be affected the most.