Mining the active proteome of Arabidopsis thaliana

Assigning functions to the >30,000 proteins encoded by the Arabidopsis genome is a challenging task of the Arabidopsis Functional Genomics Network. Although genome-wide technologies like proteomics and transcriptomics have generated a wealth of information that significantly accelerated gene annotation, protein activities are poorly predicted by transcript or protein levels as protein activities are post-translationally regulated. To directly display protein activities in Arabidopsis proteomes, we developed and applied activity-based protein profiling (ABPP). ABPP is based on the use of small molecule probes that react with the catalytic residues of distinct protein classes in an activity-dependent manner. Labeled proteins are separated and detected from proteins gels and purified and identified by mass spectrometry. Using probes of six different chemotypes we have displayed activities of 76 Arabidopsis proteins. These proteins represent over 10 different protein classes that contain over 250 Arabidopsis proteins, including cysteine, serine, and metalloproteases, lipases, acyltransferases, and the proteasome. We have developed methods for identification of in vivo labeled proteins using click chemistry and for in vivo imaging with fluorescent probes. In vivo labeling has revealed additional protein activities and unexpected subcellular activities of the proteasome. Labeling of extracts displayed several differential activities, e.g., of the proteasome during immune response and methylesterases during infection. These studies illustrate the power of ABPP to display the functional proteome and testify to a successful interdisciplinary collaboration involving chemical biology, organic chemistry, and proteomics.