Biology subjects

Misharin, A.

Publications and source records attributed to Misharin, A..

5 recordsLinked to original sources

bioRxiv2022

Biologically informed deep learning to infer gene program activity in single cells

The increasing availability of large-scale single-cell datasets has enabled the detailed description of cell states across multiple biological conditions and perturbations. In parallel, recent advances in unsupervised machine learning, particularly in transfer learning, have enabled fast and scalable mapping of these new single-cell datasets onto reference atlases. The resulting large-scale machine learning models however often have millions of parameters, rendering interpretation of the newly mapped datasets challenging. Here, we propose expiMap, a deep learning model that enables interpretable reference mapping using biologically understandable entities, such as curated sets of genes and gene programs. The key concept is the substitution of the uninterpretable nodes in an autoencoders bottleneck by labeled nodes mapping to interpretable lists of genes, such as gene ontologies, biological pathways, or curated gene sets, for which activities are learned as constraints during reconstruction. This is enabled by the incorporation of predefined gene programs into the reference model, and at the same time allowing the model to learn de novo new programs and refine existing programs during reference mapping. We show that the model retains similar integration performance as existing methods while providing a biologically interpretable framework for understanding cellular behavior. We demonstrate the capabilities of expiMap by applying it to 15 datasets encompassing five different tissues and species. The interpretable nature of the mapping revealed unreported associations between interferon signaling via the RIG-I/MDA5 and GPCRs pathways, with differential behavior in CD8+ T cells and CD14+ monocytes in severe COVID-19, as well as the role of annexins in the cellular communications between lymphoid and myeloid compartments for explaining patient response to the applied drugs. Finally, expiMap enabled the direct comparison of a diverse set of pancreatic beta cells from multiple studies where we observed a strong, previously unreported correlation between the unfolded protein response and asparagine N-linked glycosylation. Altogether, expiMap enables the interpretable mapping of single cell transcriptome data sets across cohorts, disease states and other perturbations.

bioinformatics↗

bioRxiv2021

Lung injury induces alveolar type 2 cell hypertrophy and polyploidy with implications for repair and regeneration

Epithelial polyploidization post-injury is a conserved phenomenon, recently shown to improve barrier restoration during wound healing. Whether lung injury can induce alveolar epithelial polyploidy is not known. We show that bleomycin injury induces AT2 cell hypertrophy and polyploidy. AT2 polyploidization is also seen in short term ex vivo cultures, where AT2-to-AT1 trans-differentiation is associated with substantial binucleation due to failed cytokinesis. Both hypertrophic and polyploid features of AT2 cells can be attenuated by inhibiting the integrated stress response (ISR) using the small molecule ISRIB. These data suggest that AT2 hypertrophic growth and polyploidization may be a feature of alveolar epithelial injury. As AT2 cells serve as facultative progenitors for the distal lung epithelium, a propensity for injury-induced binucleation has implications for AT2 self-renewal and regenerative potential upon reinjury, which may benefit from targeting the ISR.

cell biology↗

bioRxiv2021

Tissue-resident, extravascular CD64-Ly6C- population forms a critical barrier for inflammation in the synovium

Monocytes are one of the most abundant immune cells infiltrating the inflamed organs. However, the majority of studies on monocytes focus on circulating cells, rather than those in the tissue. Here, we identify and characterize an intravascular (i.v.) and extravascular (e.v.) synovial population (Syn Ly6C- cells) which lack cell surface markers of classical monocytes (Ly6C and CD62) or tissue macrophages (CD64 and Tim4), are transcriptionally distinct and conserved in RA patients. e.v. Syn Ly6C- cells are independent of NR4A1 and CCR2, long-lived and embryonically derived while the i.v. Syn Ly6C- cells are dependent on NR4A1, short lived and derived from circulating NCM. e.v. Syn Ly6C- cells undergo increased proliferation and reverse diapedesis dependent on LFA1 in response to arthrogenic stimuli and are required for the development of RA-like disease. These findings uncover a new facet of mononuclear cell biology and are imperative to understanding tissue-resident myeloid cell function in RA.

immunology↗

bioRxiv2020

Influenza-induced activation of recruited alveolar macrophages during the early inflammatory phase drives lung injury and lethality

Severe respiratory virus infections initiate a robust host immune response that contributes to disease severity. Immunomodulatory strategies that limit virus-initiated inflammation are of critical importance. In this study, we compared the host response to influenza A virus (IAV) infection in susceptible animals (wild-type, WT) to resilient mice (Vimentin-/- mice). We identified distinct gene expression patterns in recruited monocyte-derived alveolar macrophages (MoAMs) associated with three phases (Infiltrating, Early Inflammatory, Late Inflammatory) that evolve in sequence over the course of IAV infection. We report a core set of pro-inflammatory genes involved in the WT MoAM Early Inflammatory response that is suppressed in Vim-/- MoAMs. Moreover, we identify CEBPB, Jun-AP1, and IRF transcriptions factors as regulators of this attenuated inflammatory response. We performed causal experiments using bone-marrow chimeras to credential that Vim-/- MoAMs with suppressed pro-inflammatory genes confer protection from influenza-induced mortality in WT susceptible mice. Taken together, these data support the notion that vimentin plays a causal role in determining the pro-inflammatory function of recruited MoAMs and drives IAV-induced lung injury.

immunology↗

bioRxiv2017

Arginase 1 Is Negatively Regulated By β-Catenin Signaling In The Lung

We previously demonstrated that mice lacking the Wnt co-receptor, Lrp5, had attenuated pulmonary fibrosis in the bleomycin model. We found that Arginase 1 (Arg1), an enzyme that converts L-arginine to urea and ornithine, was markedly elevated in Lrp5-/- lungs compared with wild-type mice after bleomycin injury. We show that this induction is not apparently due to the expression of Th2 cytokines, IL-4 and IL-13, but instead is due to Wnt/{beta}-catenin signaling, which negatively regulates Arg1 expression in lung macrophages. Although Arg1 expression in macrophages has been used to define an alternatively activated phenotype, flow cytometry analysis of alveolar and interstitial macrophage sub-populations in Lrp5-/- lungs 14 days after bleomycin injury revealed no clear evidence of skewing from a classical to an alternatively activated phenotype. Upregulation of Arg1 expression and arginase activity might diminish lung arginine levels with consequent alterations in collagen or cytokine production. However, dietary supplementation of bleomycin-treated Lrp5-/- mice with the Arg1 substrate, L-arginine, failed to alter lung collagen content or cytokine levels 21 days after bleomycin injury. These findings demonstrate that Arg1 is negatively regulated by {beta}-catenin signaling in macrophages, raising the possibility that Wnt signaling directs alterations in immune cell metabolism that may be relevant to lung repair after injury.

cell biology↗