ABSTRACT
This study describes our proposed model design for SMM4H 2021 shared tasks. We fine-tune the language model of RoBERTa transformers and their connecting classifier to complete the classification tasks of tweets for adverse pregnancy outcomes (Task 4) and potential COVID-19 cases (Task 5). The evaluation metric is F1-score of the positive class for both tasks. For Task 4, our best score of 0.93 exceeded the median score of 0.925. For Task 5, our best of 0.75 exceeded the median score of 0.745. © 2021 Association for Computational Linguistics.
ABSTRACT
Disulfiram is an FDA-approved drug that has been used to treat alcoholism and has demonstrated a wide range of anti-cancer, anti-bacterial, and anti-viral effects. In the global COVID-19 pandemic, there is an urgent need for effective therapeutics and vaccine development. According to recent studies, disulfiram can act as a potent SARS-CoV-2 replication inhibitor by targeting multiple SARS-CoV-2 non-structural proteins to inhibit viral polyprotein cleavage and RNA replication. Currently, disulfiram is under evaluation in phase II clinical trials to treat COVID-19. With more and more variants of the SARS-CoV-2 worldwide, it becomes critical to know whether disulfiram can also inhibit viral entry into host cells for various variants and replication inhibition. Here, molecular and cellular biology assays demonstrated that disulfiram could interrupt viral spike protein binding with its receptor ACE2. By using the viral pseudo-particles (Vpps) of SARS-CoV-2, disulfiram also showed the potent activity to block viral entry in a cell-based assay against Vpps of different SARS-CoV-2 variants. We further established a live virus model system to support the anti-viral entry activity of disulfiram with the SARS-CoV-2 virus. Molecular docking revealed how disulfiram hindered the binding between the ACE2 and wild-type or mutated spike proteins. Thus, our results indicate that disulfiram has the capability to block viral entry activity of different SARS-CoV-2 variants. Together with its known anti-replication of SARS-CoV-2, disulfiram may serve as an effective therapy against different SARS-CoV-2 variants.
ABSTRACT
Transmembrane serine protease (TMPRSS2) plays an oncogenic role in prostate cancer as the fusion gene with ERG, and has also been demonstrated to be essential for the cellular entry of severe acute respiratory syndrome coronaviruses (SARS-CoV). Thus, targeting TMPRSS2 is a promising strategy for therapies against both prostate cancer and coronavirus infection. Although Nafamostat and Camostat have been identified as TMPRSS2 inhibitors, severe side effects such as cerebral hemorrhage, anaphylactoid reaction, and cardiac arrest shock greatly hamper their clinical use. Therefore, more potent and safer drugs against this serine protease should be further developed. In this study, we developed a fluorescence resonance energy transfer (FRET)-based platform for effectively screening of inhibitors against TMPRSS2 protease activity. The disruption of FRET between green and red fluorescent proteins conjugated with the substrate peptide, which corresponds to the cleavage site of SARS-CoV-2 Spike protein, was measured to determine the enzymatic activity of TMPRSS2. Through an initiate pilot screening with around 100 compounds, Flupirtine, a selective neuronal potassium channel opener, was identified as a potential TMPRSS2 inhibitor from an FDA-approved drug library by using this screening platform, and showed inhibitory effect on the TMPRSS-dependent infection of SARS-CoV-2 Spike-pseudotyped lentiviral particles. This study describes a platform proven effective for rapidly screening of TMPRSS2 inhibitors, and suggests that Flupirtine may be worthy of further consideration of repurposing to treat COVID-19 patients.