UC Berkeley

Degradation of proteins is an essential biological process. As is the case for protein synthesis, degradation equally represents an intricate and tightly regulated process which is coordinated by a large macromolecular assembly. Compartmentalized proteases have been an evolutionary solution to the need of the cell to selectively remove proteins and are seen in all membrane-bound organisms, from the ClpXP complex in bacteria to the 26S proteasome in eukaryotes. An astounding 80% of all proteins are believed to be shipped into this degradation machine. In eukaryotes, this fate is largely determined by a protein’s post or co-translational modification with the small signaling protein ubiquitin by a network of ubiquitin ligases. Together, protease and signal form the cell’s ubiquitin-proteasome system (UPS). Protein clearance can be as specific as the removal of a single copy protein in a sea of thousands or as broad as remodeling the entire proteome. The UPS has critical roles in every aspect of cell biology from the regulation of gene transcription to quality control of translation, and protein import across membranes. Despite this, methods to profile the pool of substrates of the proteasome have not developed as robustly as methods to profile the entire pool of translated mRNAs as has been done with the proteasome’s counterpart, the ribosome.

The second chapter of this thesis describes the theory and application of genetic code expansion for the introduction of chemical tools to trap substrates in the act of translocation into the proteasome. Taking a structure-guided approach in addition to leveraging a suite of chemical biology tools available, I describe the construction of a system for the incorporation of a light-activated crosslinking unnatural amino acid (UAA) into recombinantly expressed proteasomes. Systematic testing of UAA-containing variants allowed me to find a position within the substrate pore of the AAA+ motor of the proteasome base subcomplex that resulted in proteasomes which were capable of still engaging and degrading a substrate. Using crosslinking assays, I could show that these proteasomes are crosslinking competent in a UV-dependent manner and that they cross-link specifically to substrates with a bona fide initiation region. Using this biochemical information, a system for the incorporation of UAA in live cells was developed and shown to also result in proteasome crosslinking to substrate.

Having set up systems for recombinant and live-cell crosslinking and found a permissible crosslinking position on the proteasome motor, In the third chapter of this work, I focus on then optimizing a workflow for the purification of proteasome-substrate crosslinks and their preparation and identification through mass spectrometry (MS). By using a denaturing IMAC approach, I was able to purify proteasome-substrate crosslinks that I could then analyze using mass spectrometry. My results identified over 900 putative proteasome substrates whose roles in the cell are predictably varied. I show that in log-phase growth cells the proteasome is turning over a large pool of cytosolic enzymes as its primary substrates. Amongst these protein substrates are a large pool of metabolic enzymes, translation associated proteins, either subunits or assembly factors of the ribosome, and stress granule components. Taken together this work demonstrates the power of the use of genetic code expansion as a tool for probing complex biological systems. It also represents the first use of a crosslinker to trap substrates to an actively translocating molecular motor.