Quantitative Deubiquitinase (DUB) Interactomics to Dissect the Biology of Cancer Associated DUBs

Abstract

Ubiquitination is a prominent reversible post-translational modification, which plays crucial roles in regulating numerous cellular events. Conversely, deubiquitination is the removal of ubiquitin/polyubiquitin chains from ubiquitinated proteins. Deubiquitination is carried out by the class of enzymes called deubiquitinases (DUBs). These DUBs have vital roles in the regulation of protein ubiquitination and have a profound influence on the fate of proteins and their activity. Dysregulation of DUBs has been associated with the development of many human diseases, including various cancers. Therefore, DUBs are increasingly becoming attractive therapeutic targets in recent years. For instance, the USP14 DUB inhibitor VLX1570 entered clinical trials (now suspended) for the treatment of multiple myeloma (Want et al., 2016). Though DUBs are implicated in cancers, the precise cellular roles of many DUBs remain poorly understood.

USP35, a poorly studied DUB was identified with genomic amplification in breast and ovarian cancers (Boehringer Ingelheim, Vienna, Austria). Its amplification was correlated with overexpression and poor prognosis (Chin et al., 2007). siRNAs knockdown of USP35, in USP35 amplified cancer cells resulted in reduced cell proliferation. Therefore, USP35 was suggested to be an attractive therapeutic target for cancer treatment. During the start of my project, the biological function of USP35 was unknown. Therefore, in the present study, I explored the functions of USP35 with interaction proteomics, bioinformatics and cell biology-based approaches. Interaction proteomic analysis of USP35 with immunoprecipitation (IP) of endogenous USP35 yielded few USP35 interactors and many non-specific IgG-interacting proteins. Intriguingly, we found evidence for the existence of at least two major isoforms of USP35 with distinct N-termini. The canonical full-length USP35 (iso1) is the major isoform and a shorter isoform with an N-terminal deletion, USP35270-1018 (iso2) is also present.

In order to investigate the isoform-specific function of USP35, I applied the recently developed BioID method and established a bioinformatics pipeline to capture the isoform-specific interactome of USP35. The high-confidence proximity interacting proteins (HPIPs) obtained from the BioID method revealed that each isoform of USP35 interacts with a distinct set of proteins localized to different cellular compartments. We find that USP35 iso1 interacts with proteins that mostly localize to the centrosome and cytosol, whereas USP35 iso2 interacts with proteins localized to the ER and membrane. The distinct subcellular localization of the two USP35 isoforms was validated with immunofluorescence microscopy. Additionally, BioID interaction proteomic analysis revealed that both isoforms of USP35 predominantly interact with many pro- and anti-apoptotic proteins. Therefore, I investigated if USP35 has any role in apoptosis. Our analysis revealed USP35 iso1 is an anti-apoptotic deubiquitinase that inhibits TRAIL (Tumor Necrosis-Factor Related Apoptosis- inducing Ligand) and staurosporine-induced apoptosis. In contrast, USP35 iso2 is an endoplasmic reticulum (ER) associated protein and has a pro-apoptotic function, as ectopic expression of USP35 iso2 resulted in ER stress, and lead to cell death. Our study highlights the presence of at least two different isoforms of USP35 that have opposite roles in the regulation of apoptosis.

Next, I sought to expand my analysis and systematically investigate the interactome of 32 DUBs that are majorly amplified, mutated or deleted in various cancers. Notably, several of them are poorly studied. I used the catalytically dead version of the different DUBs to increase DUB-substrate/interacting proteins while the biotin ligase (BirA*) biotinylates interacting proteins. Using this approach, I identified many reported interacting proteins suggests that this is a robust approach. Importantly, we identified several novel interacting proteins for many DUBs, further study of which could reveal hitherto unknown cellular function. With Gene Ontology (GO) analysis, I determined the putative subcellular localization and biological processes in which they are involved. I validated the identified interacting proteins for OTUD6B, which has revealed a role for OTUD6B in centrosome and ciliary regulatory function.

Knowing the identity of the substrate of a given DUB is critical to understand the biology of the DUB. However, identification of DUB substrates is challenging due to transient interactions between DUB and its substrate. In some cases, the substrates are rapidly degraded making it difficult to capture the DUB substrate interaction. Therefore, I proposed a strategy to identify substrates of USP25, whose substrates targeted for proteasomal degradation. To identify putative substrates of USP25, I combined BioID analysis with proteasome inhibition. The underlying rationale for this approach was that biotinylation of substrates would be enhanced following proteasome inhibition. I validated the identified novel putative substrates using USP25 KO HEK293 cells generated with CRISPR/Cas9 genome editing. Consequently, the proposed approach could be effectively applied for the identification of substrates and interacting proteins of DUBs that are targeted for proteasomal degradation.

Lastly, I have discussed the merits and limitations of the DUB BioID interactome approach employed in the current study. I have proposed approaches to validate the BioID identified putative substrates and interacting proteins. I have also proposed strategies to identify the ubiquitination site on the substrates and interacting proteins, and also to incorporate intensity-based analysis to improve confidence on the BioID method identified HPIPs.

Date of Award	2019
Original language	English
Awarding Institution	University of Dundee
Supervisor	Yogesh Kulathu (Supervisor)