Matthew K. Howard

Deep mutational scanning (DMS) measures the effects of thousands of genetic variants in a protein simultaneously. The small sample size renders classical statistical methods ineffective. For example, p-values cannot be correctly calibrated when treating variants independently. We propose Rosace, a Bayesian framework for analyzing growth-based DMS data. Rosace leverages amino acid position information to increase power and control the false discovery rate by sharing information across parameters via shrinkage. We also developed Rosette for simulating the distributional properties of DMS. We show that Rosace is robust to the violation of model assumptions and is more powerful than existing tools.

Parkinson's disease (PD) is closely linked to α-synuclein (α-syn) misfolding and accumulation in Lewy bodies. The PDZ serine protease HTRA1 degrades fibrillar tau, which is associated with Alzheimer's disease, and inactivating mutations to mitochondrial HTRA2 are implicated in PD. Here, we report that HTRA1 inhibits aggregation of α-syn as well as FUS and TDP-43, which are implicated in amyotrophic lateral sclerosis (ALS) and frontotemporal dementia. The protease domain of HTRA1 is necessary and sufficient for inhibiting aggregation, yet this activity is proteolytically-independent. Further, HTRA1 disaggregates preformed α-syn fibrils, rendering them incapable of seeding aggregation of endogenous α-syn, while reducing HTRA1 expression promotes α-syn seeding. HTRA1 remodels α-syn fibrils by targeting the NAC domain, the key domain catalyzing α-syn amyloidogenesis. Finally, HTRA1 detoxifies α-syn fibrils and prevents formation of hyperphosphorylated α-syn accumulations in primary neurons. Our findings suggest that HTRA1 may be a therapeutic target for a range of neurodegenerative disorders.

Procollagenase 3 can be activated by interaction with and cleavage by the cell-associated membrane type 1 metalloproteinase (MT1 MMP; MMP 14). It has also been shown to bind to a specific receptor, and is subsequently internalized via the low-density lipoprotein-related receptor by osteoblast cell lines. The receptor was identified as a recycling glycoprotein of the macrophage mannose receptor family, Endo180. In order to ascertain whether there is a relationship between Endo180 binding and procollagenase 3 activation, we have compared procollagenase 3 activation by an HT1080 fibrosarcoma cell line overexpressing MT1 MMP, without and with overexpression of Endo180. No difference in procollagenase 3 activation was observed, and neither was the enzyme bound to the cells or internalized. In contrast, the osteoblast cell lines, MG63 and UMR-106, both bound and internalized procollagenase 3. However, immunolocalization studies showed that the Endo180 abundantly expressed by these cells did not co-localize with the procollagenase 3. In further biochemical studies we confirmed that procollagenase 3 did not bind to Endo180, using both ligand- blotting and immunoprecipitation techniques. We conclude that Endo180 is unlikely to be a receptor for collagenase 3 in relation to either its activation or cell binding and internalization, and that other interaction partners must be sought.