Search

COVID-19 patients commonly present with neurological signs of central nervous system (CNS) and/or peripheral nervous system dysfunction. However, which neural cells are permissive to infection by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been controversial. Here, we show that midbrain dopamine (DA) neurons derived from human pluripotent stem cells (hPSCs) are selectively permissive to SARS-CoV-2 infection both in vitro and upon transplantation in vivo, and that SARS-CoV-2 infection triggers a DA neuron inflammatory and cellular senescence response. A high-throughput screen in hPSC-derived DA neurons identified several FDA approved drugs, including riluzole, metformin, and imatinib, that can rescue the cellular senescence phenotype and prevent SARS-CoV-2 infection. RNA-seq analysis of human ventral midbrain tissue from COVID-19 patients, using formalin-fixed paraffin-embedded autopsy samples, confirmed the induction of an inflammatory and cellular senescence signature and identified low levels of SARS-CoV-2 transcripts. Our findings demonstrate that hPSC-derived DA neurons can serve as a disease model to study neuronal susceptibility to SARS-CoV-2 and to identify candidate neuroprotective drugs for COVID-19 patients. The susceptibility of hPSC-derived DA neurons to SARS-CoV-2 and the observed inflammatory and senescence transcriptional responses suggest the need for careful, long-term monitoring of neurological problems in COVID-19 patients.

The aim of this study was to test whether bilinear and nonlinear effective connectivity (EC) measures of working memory fMRI data can differentiate between patients with schizophrenia (SZ) and healthy controls (HC). We applied bilinear and nonlinear Dynamic Causal Modeling (DCM) for the analysis of verbal working memory in 16 SZ and 21 HC. The connection strengths with nonlinear modulation between the dorsolateral prefrontal cortex (DLPFC) and the ventral tegmental area/substantia nigra (VTA/SN) were evaluated. We used Bayesian Model Selection at the group and family levels to compare the optimal bilinear and nonlinear models. Bayesian Model Averaging was used to assess the connection strengths with nonlinear modulation. The DCM analyses revealed that SZ and HC used different bilinear networks despite comparable behavioral performance. In addition, the connection strengths with nonlinear modulation between the DLPFC and the VTA/SN area showed differences between SZ and HC. The adoption of different functional networks in SZ and HC indicated neurobiological alterations underlying working memory performance, including different connection strengths with nonlinear modulation between the DLPFC and the VTA/SN area. These novel findings may increase our understanding of connectivity in working memory in schizophrenia.