An atlas of transcribed enhancers across helper T cell diversity for decoding human diseases

Transcribed enhancer maps can reveal nuclear interactions underpinning each cell type and connect specific cell types to diseases. Using a 5′ single-cell RNA sequencing approach, we defined transcription start sites of enhancer RNAs and other classes of coding and non-coding RNAs in human CD4+ T cells, revealing cellular heterogeneity and differentiation trajectories. Integration of these datasets with single-cell chromatin profiles showed that active enhancers with bidirectional RNA transcription are highly cell type–specific, and disease heritability is strongly enriched in these enhancers. The resulting cell type–resolved multimodal atlas of bidirectionally transcribed enhancers, which we linked with promoters using fine-scale chromatin contact maps, enabled us to systematically interpret genetic variants associated with a range of immune-mediated diseases.

Methods
All experiments using human samples were approved by the ethical review committee of RIKEN [approval no. H30-9(13)]. Written informed consent was obtained from all donors. CD4+ T cells were isolated by the immunomagnetic negative selection method. Stained CD4+ T cells were sorted using a FACSAria IIu Cell Sorter (BD Biosciences). Human CD4+ T cells and FACS-sorted heterogenous populations were processed with a Chromium Next GEM Single Cell 5′ kit (10x Genomics). Libraries were sequenced on an Illumina NovaSeq 6000 sequencing platform using 2 × 150 bp paired-end sequencing. Multiome assay (10x Genomics) was performed according to the manufacturer’s instructions. Multiome libraries were pooled and sequenced as above with 10 cycles for i7 index and 24 cycles for i5 index. Micro-C libraries were generated using a Dovetail Micro-C Kit (Cantata Bio, Cat#21006) and were sequenced on an Illumina NovaSeq 6000 platform using 2 × 150 bp paired-end sequencing. Chromium scRNA-seq, snRNA-seq, and CITE-seq data were processed using Cell Ranger Software version 5.0.1 (10x Genomics) and R package Seurat version 5 (4.9.9.9067). Multiome data were processed by Cell Ranger ARC version 2.0.0 (10x Genomics), Seurat version 5 (4.9.9.9067), and Signac version 1.10.0. scRNA-seq, snRNA-seq, and Multiome 3′ snRNA-seq data were integrated using canonical correlation analysis. snATAC peaks were identified from fragment files of each cluster using MACS2 version 2.2.6 with default settings as implemented in Signac version 1.10.0. For ReapTEC, paired-end reads were mapped again using STAR (STARsolo) to obtain reads with unencoded G, which was tagged as a soft-clipped G by STARsolo. Reads were deduplicated, and those with the barcodes of each cell type were extracted. A count file was generated for each transcription start site (TSS) using the “bamToBed” function in BEDTools version 2.30.0. TSS peaks were generated by merging TSSs located within 10 bp of each other. To identify btcEnhs, TSS peak pairs were detected using scripts provided at https://github.com/anderssonrobin/enhancers/blob/master/scripts/bidir_enhancers with minor modifications. Micro-C data were processed with the dovetail_tools pipeline (Cantata Bio). Chromatin loop contacts were identified by the HiCCUPS algorithm using the Juicer Tools package version 2.20.0 and the scale-space representation algorithm using the Mustache package. Loops were called at a 1-kb resolution with SCALE-normalized contact matrices for HiCCUPS and with ICE-normalized contact matrices for Mustache, and were filtered for an FDR < 0.05.