Distinct miRNA Signatures and Networks Discern Fetal from Adult Erythroid Differentiation and Primary from Immortalized Erythroid Cells.

MicroRNAs (miRNAs) are small non-coding RNAs crucial for post-transcriptional and translational regulation of cellular and developmental pathways. The study of miRNAs in erythropoiesis elucidates underlying regulatory mechanisms and facilitates related diagnostic and therapy development. Here, we used DNA Nanoball (DNB) small RNA sequencing to comprehensively characterize miRNAs in human erythroid cell cultures. Based on primary human peripheral-blood-derived CD34+ (hCD34+) cells and two influential erythroid cell lines with adult and fetal hemoglobin expression patterns, HUDEP-2 and HUDEP-1, respectively, our study links differential miRNA expression to erythroid differentiation, cell type, and hemoglobin expression profile. Sequencing results validated by reverse-transcription quantitative PCR (RT-qPCR) of selected miRNAs indicate shared differentiation signatures in primary and immortalized cells, characterized by reduced overall miRNA expression and reciprocal expression increases for individual lineage-specific miRNAs in late-stage erythropoiesis. Despite the high similarity of same-stage hCD34+ and HUDEP-2 cells, differential expression of several miRNAs highlighted informative discrepancies between both cell types. Moreover, a comparison between HUDEP-2 and HUDEP-1 cells displayed changes in miRNAs, transcription factors (TFs), target genes, and pathways associated with globin switching. In resulting TF-miRNA co-regulatory networks, major therapeutically relevant regulators of globin expression were targeted by many co-expressed miRNAs, outlining intricate combinatorial miRNA regulation of globin expression in erythroid cells.

Knowledge-based classification of fine-grained immune cell types in single-cell RNA-Seq data.

Single-cell RNA sequencing (scRNA-Seq) is an emerging strategy for characterizing immune cell populations. Compared to flow or mass cytometry, scRNA-Seq could potentially identify cell types and activation states that lack precise cell surface markers. However, scRNA-Seq is currently limited due to the need to manually classify each immune cell from its transcriptional profile. While recently developed algorithms accurately annotate coarse cell types (e.g. T cells versus macrophages), making fine distinctions (e.g. CD8+ effector memory T cells) remains a difficult challenge. To address this, we developed a machine learning classifier called ImmClassifier that leverages a hierarchical ontology of cell type. We demonstrate that its predictions are highly concordant with flow-based markers from CITE-seq and outperforms other tools (+15% recall, +14% precision) in distinguishing fine-grained cell types with comparable performance on coarse ones. Thus, ImmClassifier can be used to explore more deeply the heterogeneity of the immune system in scRNA-Seq experiments.

Understanding heterogeneity of fetal hemoglobin induction through comparative analysis of F and A erythroblasts.

Reversing the developmental switch from fetal hemoglobin (HbF, α2γ2) to adult hemoglobin (HbA, α2ß2) is an important therapeutic approach in sickle cell disease (SCD) and ß-thalassemia. In healthy individuals, SCD patients, and patients treated with pharmacologic HbF inducers, HbF is present only in a subset of red blood cells known as F cells. Despite more than 50 years of observations, the cause for this heterocellular HbF expression pattern, even among genetically identical cells, remains unknown. Adult F cells might represent a reversion of a given cell to a fetal-like epigenetic and transcriptional state. Alternatively, isolated transcriptional or posttranscriptional events at the γ-globin genes might underlie heterocellularity. Here, we set out to understand the heterogeneity of HbF activation by developing techniques to purify and profile differentiation stage-matched late erythroblast F cells and non-F cells (A cells) from the human HUDEP2 erythroid cell line and primary human erythroid cultures. Transcriptional and proteomic profiling of these cells demonstrated very few differences between F and A cells at the RNA level either under baseline conditions or after treatment with HbF inducers hydroxyurea or pomalidomide. Surprisingly, we did not find differences in expression of any known HbF regulators, including BCL11A or LRF, that would account for HbF activation. Our analysis shows that F erythroblasts are not significantly different from non-HbF-expressing cells and that the primary differences likely occur at the transcriptional level at the ß-globin locus.