Broad-Spectrum Proteome Editing with an Engineered Bacterial Ubiquitin Ligase Mimic

Abstract

Manipulation of the ubiquitin-proteasome pathway to achieve targeted silencing of cellular proteins has emerged as a reliable and customizable strategy for remodeling the mammalian proteome. One such approach involves engineering bifunctional proteins called ubiquibodies that are comprised of a synthetic binding protein fused to an E3 ubiquitin ligase, thus enabling post-translational ubiquitination and degradation of a target protein independent of its function. Here, we have designed a panel of new ubiquibodies based on E3 ubiquitin ligase mimics from bacterial pathogens that are capable of effectively interfacing with the mammalian proteasomal degradation machinery for selective removal of proteins of interest. One of these, the Shigella flexneri effector protein IpaH9.8 fused to a fibronectin type III ( FN3 ) monobody that specifically recognizes green fluorescent protein (GFP), was observed to potently eliminate GFP and its spectral derivatives as well as 15 different FP-tagged mammalian proteins that varied in size (27–179 kDa) and subcellular localization (cytoplasm, nucleus, membrane-associated, and transmembrane). To demonstrate therapeutically relevant delivery of ubiquibodies, we leveraged a bioinspired molecular assembly method whereby synthetic mRNA encoding the GFP-specific ubiquibody was coassembled with poly A binding proteins and packaged into nanosized complexes using biocompatible, structurally defined polypolypeptides bearing cationic amine side groups. The resulting nanoplexes delivered ubiquibody mRNA in a manner that caused efficient target depletion in cultured mammalian cells stably expressing GFP as well as in transgenic mice expressing GFP ubiquitously. Overall, our results suggest that IpaH9.8-based ubiquibodies are a highly modular proteome editing technology with the potential for pharmacologically modulating disease-causing proteins.

Proteome editing using engineered proteins that hijack cellular quality control machinery

“…A distinct advantage of uAbs is their highly modular architecture which enables target selection to be rewired by simply swapping the synthetic protein‐binding domain. For example, targeted proteolysis has been achieved for a diverse array of protein substrates including eukaryotic proteins (ASC, HRAS/KRAS, Lck, and SHP2,), intraneuronal bacterial proteins ( Clostridium botulinum neurotoxin [BoNT] proteases), fluorescent proteins (FPs), and dozens of FP‐tagged proteins that range in size from 27 to 179 kDa and localize in different subcellular compartments including the cytoplasm, nucleus, and cell membrane . Moreover, by incorporating synthetic binding proteins that recognize particular protein states (e.g., active vs. inactive conformation, mutant vs. wild‐type, posttranslationally modified, and so forth), it becomes possible to deplete certain protein subpopulations while sparing others .…”

“…For example, targeted proteolysis has been achieved for a diverse array of protein substrates including eukaryotic proteins (ASC, HRAS/KRAS, Lck, and SHP2,), intraneuronal bacterial proteins ( Clostridium botulinum neurotoxin [BoNT] proteases), fluorescent proteins (FPs), and dozens of FP‐tagged proteins that range in size from 27 to 179 kDa and localize in different subcellular compartments including the cytoplasm, nucleus, and cell membrane . Moreover, by incorporating synthetic binding proteins that recognize particular protein states (e.g., active vs. inactive conformation, mutant vs. wild‐type, posttranslationally modified, and so forth), it becomes possible to deplete certain protein subpopulations while sparing others . Also, in contrast to PROTACs where ligands for new substrates are hard to come by, expanding the number of targets for uAbs including those deemed undruggable is made relatively straightforward by library‐based screening methods such as phage display, ribosome display, or yeast surface display …”

“…Indeed, a number of different E3 domains have been developed for uAb‐mediated knockdown including stand‐alone E3s such as CHIP, E3 substrate adaptors/receptors that associate with larger E3 ligase complexes, and a large collection of E3 ligases of bacterial origin including the Shigella flexneri IpaH family members . When compared side‐by‐side, it is clear that the choice of ubiquitin ligase can have a profound effect on silencing efficiency with IpaH family enzymes proving to be top performers in the uAb context . This area warrants further investigation as it seems likely that other candidates from the exceptionally large pool of naturally occurring E3 ligases of eukaryotic, bacterial, and viral origin might function as even more potent degraders in the uAb context.…”

bioPROTACs establish RAS as a degradable target and provide novel RAS biology insights

“…Indeed, active bioPROTACs have been generated with fusions between E3s and nanobodies, monobodies, alpha-reps, DARPins, and peptides 16,17 . The choice of E3 is also flexible, with functional bioPROTACs having been engineered from both human and bacterial sequences 16,20 .…”

“…First, the natural turn-over of RAS proteins was reported to be proteasome-dependent and regulated by the E3 ligases LTZR1 [22][23][24] and βTrCP 25 . Second, the G12C covalent modifier and bioPROTAC approaches have been successful for degrading GFP-KRAS 20,21 . Third, bioPROTAC equivalents consisting of the endogenous RAS-bindingdomain (RBD) fused to either VIF or CHIP E3 ligases have resulted in modest KRAS degradation 26,27 .…”