This paper introduces single-cell DNA–RNA sequencing (SDR-seq), a novel high-throughput method designed to simultaneously profile up to 480 genomic DNA loci and genes in thousands of single cells. This technology addresses the limitations of existing methods in precisely linking genetic variants (both coding and noncoding) to gene expression changes within their endogenous context, particularly in the vast noncoding regions often associated with disease. SDR-seq combines in situ reverse transcription with multiplexed PCR in droplets, enabling accurate determination of variant zygosity and associated gene expression changes with high sensitivity and reproducibility across various panel sizes and chromatin states, achieving significantly lower allelic dropout rates compared to many existing techniques. The method's capabilities are demonstrated by associating coding and noncoding variants with distinct gene expression in human induced pluripotent stem cells and by revealing elevated B cell receptor signaling and tumorigenic gene expression in primary B cell lymphoma cells with higher mutational burdens. SDR-seq offers a powerful, scalable platform for dissecting regulatory mechanisms driven by genetic variants, advancing understanding of gene expression regulation in health and disease.
Summary of "Functional phenotyping of genomic variants using joint multiomic single-cell DNA–RNA sequencing"
Introduction to the Challenge: Over 90% of disease-associated genetic variants are in noncoding regions, making their regulatory impact difficult to assess. Existing tools for linking genotypes to gene expression at single-cell resolution are often inefficient, lack endogenous context, or have low throughput and high allelic dropout rates.
Introducing SDR-seq (Single-Cell DNA–RNA Sequencing): SDR-seq is a targeted, droplet-based method that simultaneously measures RNA and genomic DNA (gDNA) targets in the same cell with high coverage and sensitivity.
In situ reverse transcription (RT) adds unique molecular identifiers (UMIs), sample barcodes (BCs), and capture sequences (CS) to cDNA.
Cells with cDNA and gDNA are loaded into the Tapestri machine.
Multiplexed PCR amplifies both gDNA and RNA targets within droplets using forward primers with CS overhangs and barcoding beads.
Distinct overhangs on reverse primers (R2N for gDNA, R2 for RNA) allow for separate next-generation sequencing (NGS) library generation, optimizing sequencing for full-length variant information on gDNA and transcript/BC information for RNA.
Proof-of-Principle (POP) Experiment (iPS Cells):
Tested with human induced pluripotent stem (iPS) cells, targeting 28 gDNA and 30 RNA targets.
Compared PFA and glyoxal fixatives; glyoxal showed increased RNA target detection and UMI coverage.
Achieved ~9,000 high-quality cells from a single run with >95% correct sample barcode detection.
Detected 82% of gDNA targets with high coverage in most cells and varying RNA expression.
Demonstrated higher measurement stability and correlation with bulk RNA-seq compared to 10x Genomics and ParseBio data.
Species-mixing experiment showed low cross-contamination for gDNA (<0.16%) and RNA (0.8–1.6%).
Scalability to Hundreds of Targets:
SDR-seq successfully scaled to panels of 120, 240, and 480 gDNA and RNA targets.
Over 80% of gDNA targets detected with high confidence across all panels, with minimal decrease for larger panels.
Detection and gene expression of shared targets were highly correlated between panels, showing robustness.
GDNA detection was independent of chromatin context (e.g., promoter-like, enhancer-like sequences).
Allelic dropout (ADO) rates for heterozygous variants were low (87–94% correctly called), significantly better than other high-throughput methods.
Detecting Gene Expression Changes (CRISPRi, Prime Editing, Base Editing):
CRISPRi Screen: Confidently detected strong gene expression changes, with 95% of TSS-targeting gRNAs causing significant target gene reduction. Also identified eQTLs impacting expression directly.
Prime Editing (PE) Screen: Despite limited editing efficiency, detected significant gene expression reductions for STOP codon controls (e.g., ATF4, MYH10) due to nonsense-mediated decay.
Base Editing (BE) Screen: Identified several eQTL variants with significant effects on target gene expression. Discovered that somatic mutations accumulating in iPS cells, particularly in the 3′ untranslated region of POU5F1, significantly altered gene expression.
Application in B Cell Lymphoma Samples:
Profiled primary tumor samples from individuals with follicular lymphoma and diffuse large B cell lymphoma.
Successfully clustered cells into B cells and non-B cells based on both RNA and variant data.
Identified tumor-specific variants and analyzed their impact on differentiation and gene expression.
BCL2 variants were enriched in the light zone (LZ) compared to the dark zone (DZ), and many immunoglobulin variable region genes showed enrichment in LZ.
Cells with higher mutational burden showed elevated B cell receptor signaling and tumorigenic gene expression, suggesting an evasion of apoptosis induced by unsuccessful antigen binding in the LZ.
Key Advantages and Future Directions: SDR-seq offers high throughput (hundreds of loci in thousands of cells), high sensitivity (low ADO), and cost-effectiveness. It is flexible for both engineered cell lines and primary tissues. Future applications include combination with targeted protein readout, DNA methylation, mitochondrial genome targeting for clonal tracing, and expanding to whole-transcriptome readouts.