To provide proof-of-concept, we have established RBPP in plant science and used this to demonstrate that this can display differential labeling in the presence/absence of ATP (Mooney et al., manuscript in prep). We have also established CLMS in plant science and identified hundreds of crosslinked peptides and are using AlphaFold Multimer to confirm the detected close proximities in silico (Mahadevan et al., manuscript in prep). We have established robust infection assays for the three bacterial strains on Nicotiana benthamiana and have isolated large quantities of apoplastic fluids from the three infections and the mock control at the start of infection (t=0) in n=4 replicates and two days after infection (2dpi) in n=6 replicates. We performed chemical proteomics using ABPP, RBPP and CLMS on the t=0 samples and analysed the results. The chemical proteomics experiments on the t=0 samples did not yield the large number of labeling that we anticipated. ABPP displayed only 12 active hydrolases but only one protease activity was suppressed in the presence of the pathogen. Likewise, RBPP displayed only 1718 unique labeling sites in the plant proteome. Interestingly, one site in a glycosidase is consistently more reactive in the presence of the pathogen, indicating that this glycosidase is manipulated at t=0. Furthermore, CLMS revealed 32 interlinks between two different protein, including six plant-pathogen interactions. These experiments showed that, although the between-sample variation is low for the t=0 samples, the protein concentration is relatively low and we did not monitor the hundreds of proteins that we aimed at. Protein concentrations are much higher in the 2dpi samples and we are currently labeling these samples for proteomic analysis with sufficient replicates (n=6). Meanwhile, we have used AlphaFold Multimer (AFM) to predict interactions between 48 hydrolases that are suppressed upon infection with one of the three strains (pv tomato, PtoDC3000, Sueldo et al., 2024) with 211 PtoDC3000-produced proteins that were detected in the apoplast of infected plants. This screen identified >500 robust protein complexes, of which 30 complexes seem to involve a PtoDC3000 protein that might inhibit the plant hydrolase. We have produced some of these candidate inhibitors from PtoDC3000 in E. coli and expressed their presumed target hydrolases by agroinfiltration and tested if the PtoDC3000 proteins are indeed hydrolase inhibitors. Preliminary data indicate that at least some of these PtoDC3000 proteins are indeed novel inhibitors, for instance of glycosidases BGAL2 and PR3.
