Systematic immune cell dysregulation and molecular subtypes revealed by single-cell RNA-seq of subjects with type 1 diabetes.

BACKGROUND: Type 1 diabetes mellitus (T1DM) is a prototypic endocrine autoimmune disease resulting from an immune-mediated destruction of pancreatic insulin-secreting ß  cells. A comprehensive immune cell phenotype evaluation in T1DM has not been performed thus far at the single-cell level. METHODS: In this cross-sectional analysis, we generated a single-cell transcriptomic dataset of peripheral blood mononuclear cells (PBMCs) from 46 manifest T1DM (stage 3) cases and 31 matched controls. RESULTS: We surprisingly detected profound alterations in circulatory immune cells (1784 dysregulated genes in 13 immune cell types), far exceeding the count in the comparator systemic autoimmune disease SLE. Genes upregulated in T1DM were involved in WNT signaling, interferon signaling and migration of T/NK cells, antigen presentation by B cells, and monocyte activation. A significant fraction of these differentially expressed genes were also altered in T1DM pancreatic islets. We used the single-cell data to construct a T1DM metagene z-score (TMZ score) that distinguished cases and controls and classified patients into molecular subtypes. This score correlated with known prognostic immune markers of T1DM, as well as with drug response in clinical trials. CONCLUSIONS: Our study reveals a surprisingly strong systemic dimension at the level of immune cell network in T1DM, defines disease-relevant molecular subtypes, and has the potential to guide non-invasive test development and patient stratification.

scConsensus: combining supervised and unsupervised clustering for cell type identification in single-cell RNA sequencing data.

BACKGROUND: Clustering is a crucial step in the analysis of single-cell data. Clusters identified in an unsupervised manner are typically annotated to cell types based on differentially expressed genes. In contrast, supervised methods use a reference panel of labelled transcriptomes to guide both clustering and cell type identification. Supervised and unsupervised clustering approaches have their distinct advantages and limitations. Therefore, they can lead to different but often complementary clustering results. Hence, a consensus approach leveraging the merits of both clustering paradigms could result in a more accurate clustering and a more precise cell type annotation. RESULTS: We present SCCONSENSUS, an [Formula: see text] framework for generating a consensus clustering by (1) integrating results from both unsupervised and supervised approaches and (2) refining the consensus clusters using differentially expressed genes. The value of our approach is demonstrated on several existing single-cell RNA sequencing datasets, including data from sorted PBMC sub-populations. CONCLUSIONS: SCCONSENSUS combines the merits of unsupervised and supervised approaches to partition cells with better cluster separation and homogeneity, thereby increasing our confidence in detecting distinct cell types. SCCONSENSUS is implemented in [Formula: see text] and is freely available on GitHub at https://github.com/prabhakarlab/scConsensus .

Pathogenic Epigenetic Consequences of Genetic Alterations in IDH-Wild-Type Diffuse Astrocytic Gliomas.

Gliomas are classified by combining histopathologic and molecular features, including isocitrate dehydrogenase (IDH) status. Although IDH-wild-type diffuse astrocytic glioma (DAG) shows a more aggressive phenotype than IDH-mutant type, lack of knowledge regarding relevant molecular drivers for this type of tumor has hindered the development of therapeutic agents. Here, we examined human IDH-wild-type DAGs and a glioma mouse model with a mosaic analysis with double markers (MADM) system, which concurrently lacks p53 and NF1 and spontaneously develops tumors highly comparable with human IDH-wild-type DAG without characteristic molecular features of glioblastoma (DAG-nonMF). During tumor formation, enhancer of zeste homolog (EZH2) and the other polycomb repressive complex 2 (PRC2) components were upregulated even at an early stage of tumorigenesis, together with an increased number of genes with H3K27me3 or H3K27me3 and H3K4me3 bivalent modifications. Among the epigenetically dysregulated genes, frizzled-8 (Fzd8), which is known to be a cancer- and stem cell reprogramming-related gene, was gradually silenced during tumorigenesis. Genetic and pharmacologic inhibition of EZH2 in MADM mice showed reactivation of aberrant H3K27me3 target genes, including Fzd8, together with significant reduction of tumor size. Our study clarifies a pathogenic molecular pathway of IDH-wild-type DAG-nonMF that depends on EZH2 activity and provides a strong rationale for targeting EZH2 as a promising therapeutic approach for this type of glioma. SIGNIFICANCE: EZH2 is involved in the generation of IDH-wild-type diffuse astrocytic gliomas and is a potential therapeutic target for this type of glioma. GRAPHICAL ABSTRACT: http://cancerres.aacrjournals.org/content/canres/79/19/4814/F1.large.jpg.