Drs. Henrique Ribeiro and Reinhard Laubenbacher co-authored a research article with Achyudhan Kutuva, Felipe Lima, Matthew Schaller, and Borna Mehrad just published in the Bulletin of Mathematical Biology.
Abstract:
The mechanistic target of rapamycin complex 1 (mTORC1) has been implicated in coronavirus pathogenesis, yet its precise role in shaping antiviral defenses in pneumocytes remains unresolved. This study combines ex vivo human lung tissue assays with a literature-curated, logic-based (Boolean) signaling model to study how mTORC1 influences SARS-CoV-2 replication and type-I interferon (IFN) responses. Our in-vitro data showed that pharmacologic inhibition of mTORC1 with sirolimus was associated with reduced viral replication and increased IFN- expression across donor samples. To interpret these data, we constructed a network integrating inflammation, stress, apoptosis, and RIG-I–IFN signaling, and evaluated four mechanistic hypotheses for explaining our experiments in which mTORC1 either promotes replication, inhibits IFN, both, or neither. The model was constructed according to the best evidence in the literature and evaluated against 56 independent protein/phosphoprotein readouts from 26 studies not used for model construction; the model achieved high qualitative accuracy overall (F1 > 0.75) and on live-virus experiments (F1 > 0.9). We found that the hypotheses where mTORC1 promotes viral replication, inhibits IFN, or both were able to qualitatively reproduce the empirical data, i.e., mTORC1 inhibition by sirolimus leads to lower viral replication and higher IFN expression. We queried the robustness of those predictions through a systematic edge knock-in/knockout screen generating 6,328 network variants per scenario (25,312 total). We found that the scenario where mTORC1 enhances viral replication is the more plausible, due to parsimony (this mechanism is also observed in other viral infections), higher robustness (more variants still reproduce the result), and stronger inhibition that is more likely to be observed in biological experiments with low statistical power. This scenario supports the idea that the mTORC1 effects on IFN expression are indirectly mediated by viral replication itself.
