Biochemical analysis of HDAC6-mediated Influenza A virus uncoating and its targeting by small artificial proteins

HDAC6 is a member of the histone deacetylases (HDACs) family and has a unique structure, with tandem catalytic domains and a conserved ubiquitin (Ub)-binding zinc finger (ZnF) domain. While the deacetylase activity is important for cell motility, stress response, cancer progression, among others, HDAC6 function also depends on its ZnF domain which engages Ub in many cases. For example, it is important for the formation of cellular granules, such as aggresome and stress granules. In particular, the aggresome/HDAC6 pathway can be hijacked by exogenous pathogens, like influenza A virus (IAV), to facilitate the infection: the HDAC6 ZnF domain is indispensable for virion uncoating, but the molecular details are still missing.
The goal of this thesis has been (i) to gain additional details about the cellular components recruited by the HDAC6 ZnF domain and involved in the viral uncoating and (ii) to use small artificial proteins to interfere with this recruitment.
Analysis by immunoprecipitations and biochemical assays allowed me to show that the recruitment of Ub by the HDAC6 ZnF is essential to allow the formation of a complex involving components of the actomyosin system, in particular myosin 10 and actin. Unanchored Ub chains form a bridge between HDAC6 and myosin 10. This also has an impact on the recruitment of another motor system, dynein: recruitment of dynein by HDAC6 is increased when Ub cannot be recruited. These biochemical data were combined with mathematical analysis of the system by our colleagues at D-BSSE, to model the uncoating of IAV. This allowed developing a tug-of-war model of uncoating which realistically depicts the biological observations. Besides, results from clinical samples could be obtained; in particular, two IAV strains, H1N1 and H3N2, show a different dependency on the HDAC6/aggresome pathway. We could show that this relates to their M1 matrix protein having different affinities (H1N1 > H3N2) for interaction with HDAC6. This information could also be integrated into the model and allowed to further refine it (Arctibasova*, Wang* et al., submitted).
Above we demonstrated the importance of HDAC6 ZnF-Ub interaction during IAV infection, we decided to target this protein-protein interaction, as it might be valuable therapeutically. To achieve this, we screened out multiple DARPins (Designed Ankyrin Repeat Proteins) which bind specifically to HDAC6 ZnF. One of the DARPins, F10, could be shown to efficiently block the HDAC6-Ub interaction by in vitro pull-down and split-GFP assays in cells. X-ray crystallographic data (with a resolution of 2.55 Å) showed that F10 occupies the pocket where Ub engages. We established cell lines for expression and conditional degradation of DARpin F10; this allowed us to show that IAV infection is greatly inhibited by F10 expression. Investigation of IAV M1 protein release by microscopy convinced us that the uncoating step is interrupted. Furthermore, we showed that another RNA virus, ZIKA, is also impaired by DARPin F10. Other cellular pathways, like aggresome and stress granule formation, are inhibited as well (Wang et al., submitted). Considering that stress granule and aggresome are both potentially pathological granules in neurodegenerative diseases (e.g. Amyotrophic lateral sclerosis and Parkinson diseases), we have shown that the HDAC6 ZnF domain is a targetable site for drug discovery, with potential for patients who are suffering from severe virus infection or as neurodegenerative diseases.