Cell type-specific novel long non-coding RNA and circular RNA in the BLUEPRINT hematopoietic transcriptomes atlas

Abstract

Transcriptional profiling of hematopoietic cell subpopulations has helped characterize the developmental stages of the hematopoietic system and the molecular bases of malignant and non-malignant blood diseases for the past three decades. Previously, only the genes targeted by expression microarrays could be profiled genome wide. High-throughput RNA sequencing (RNA-seq), however, encompasses a broader repertoire of RNA molecules, without restriction to previously annotated genes. We analysed the BLUEPRINT consortium RNA-seq data for mature hematopoietic cell types. The data comprised 90 total RNA-seq samples, each composed of one of 27 cell types, and 32 small RNA-seq samples, each composed of one of 11 cell types. We estimated gene and isoform expression levels for each cell type using existing annotations from Ensembl. We then used guided transcriptome assembly to discover unannotated transcripts. We identified hundreds of novel non-coding RNA genes and showed that the majority have cell type dependent expression. We also characterized the expression of circular RNAs and found that these are also cell type specific. These analyses refine the active transcriptional landscape of mature hematopoietic cells, highlight abundant genes and transcriptional isoforms for each blood cell type, and provide a valuable resource for researchers of hematological development and diseases. Finally, we made the data accessible via a web-based interface: https://blueprint.haem.cam.ac.uk/bloodatlas/.

Identification of autosomal cis expression quantitative trait methylation (cis eQTMs) in children’s blood

“…In consequence, the comparison between unadjusted and adjusted models for cellular composition can provide information about potential cell type-specific eQTMs. Our results seem to support our hypothesis, as eQTMs exclusive to the main unadjusted model, and thus potential cell type-specific, were composed by CpGs and TC with higher cell type specificity (higher F-statistic) in blood cell methylation and expression sorted studies [30,31]. Of note, a high F-statistic can either mean high methylation/expression differences in one particular blood cell type or intermediate methylation/expression differences across several cell types.…”

“…Finally, we checked our hypothesis that CpGs and TCs in eQTMs uniquely identified in the main model (unadjusted cell-specific eQTMs) are blood cell type-specific by contrasting them with data from sorted blood cell types. For this, we retrieved DNA methylation levels from six sorted blood cell types by Reinius and colleagues [30] and gene expression levels from twelve sorted blood cell types from the Blueprint project [31]. We used the log 10 of the F-statistic of a linear regression (see Material and Methods) as a measure of cell type specificity (higher F-statistic, higher specificity), as described elsewhere [32].…”

“…DNA methylation and gene expression are cell type-specific [25,30,31,36]. To explore this in our bulk data, we hypothesized that causally related CpG-TC pairs would not be affected by adjustment for cellular composition if their association takes place across cell types, while if it was cell type-specific, it would be attenuated.…”

“…Higher F-statistics are assumed to be indicative of higher cell type specificity. DNA methylation and gene expression in sorted blood cell types were retrieved from [30] and the Blueprint project [31] (https://blueprint.haem.cam.ac.uk/bloodatlas/), respectively. Once gene/CpG cell type specificity was calculated and log 10 transformed (log 10 F-statistic), we tested its association with the eQTM category (model-shared or unadjusted cell-specific eQTMs) by fitting linear regressions.…”

Novel manifestations of immune dysregulation and granule defects in gray platelet syndrome

“…Of these, cathelicidin antimicrobial peptide (Camp), cysteine-rich secretory protein 3 (Crisp3), and neutrophil gelatinase-associated lipocalin (Lcn2) were all differentially less abundant in GPS neutrophils, and all localize to the neutrophilspecific granules. 45 Cross-referencing the discriminatory plasma proteins with the BLUEPRINT consortium gene expression data 51 revealed that 14 of the discriminatory plasma proteins are not known to be expressed in hematopoietic cells (Figure 4B; supplemental Figure 4). We then analyzed the gene expression of these 14 proteins in the Genotype-Tissue Expression database, 52,53 which showed that 9 of these nonhematopoietic proteins are predominantly synthesized by liver-residing cells.…”