Paper: Stereo-cell: Spatial enhanced-resolution single-cell sequencing with high-density DNA nanoball-patterned arrays
https://www.science.org/doi/10.1126/science.adr0475
Abstract:
INTRODUCTION
Single-cell sequencing has transformed our ability to study cellular diversity and molecular function. However, most existing methods require trade-offs between throughput, sensitivity, and compatibility with diverse cell types. Limitations in capture uniformity and compatibility with different cell sizes or experimental modalities restrict their utility in applications such as unbiased cell detection, analysis of in situ microenvironments or cell-cell interactions, and integration with spatial or multimodal assays.
RATIONALE
We hypothesized that high-density DNA nanoball (DNB)鈥損atterned arrays, with their planar architecture and nanometer-scale resolution, could enable a single-cell sequencing strategy based on in situ transcript capture without the need for compartmentalized encapsulation. This design would support direct profiling of small to large cells, extracellular vesicles, and microstructures. To this end, we developed Stereo-cell, a spatially resolved, high-throughput single-cell transcriptomic platform built on DNB arrays. This system enables scalable, unbiased capture of cells across a wide input range and supports high-fidelity transcriptome profiling while also allowing integration with multiplexed imaging and multiomics workflows.
RESULTS
Stereo-cell accurately and efficiently captured transcriptomes from a broad range of cell types, with input sizes spanning from as few as 200 to nearly 1 million cells per chip. The platform demonstrated high reproducibility and robust performance across different cell loading densities, with consistent transcript detection and minimal doublet rates, supported by deep learning鈥揵ased cell segmentation and imaging-based doublet elimination. In benchmarking studies that used human peripheral blood mononuclear cells (PBMCs), Stereo-cell generated gene expression profiles closely matching those obtained from established droplet-based methods while recovering cell-type proportions that aligned more faithfully with flow cytometry reference data. Notably, Stereo-cell identified rare immune populations, such as hematopoietic stem and progenitor cells, from large-scale samples. Integration with multiplex immunofluorescence and oligo-barcoded antibodies enabled concurrent profiling of mRNA and protein markers, revealing distinct immune activation states and uncovering phenotypic diversity not apparent from RNA alone. Stereo-cell further enabled in situ profiling of cultured fibroblasts, capturing microenvironmental cues, extracellular vesicle鈥搇ike structures based on distinct gene signatures and imaging data, and fibroblast responses to stimulation. Stereo-cell succeeded in sequencing structurally complex and oversized samples, including mouse oocytes and multinucleated skeletal myofibers, while preserving transcript localization at subcellular resolution. RNA staining validation and spatial gene module analysis confirmed Stereo-cell鈥檚 capability to resolve intracellular transcript architecture and microenvironmental organization.
CONCLUSION
Stereo-cell establishes an integrative framework for high-throughput single-cell transcriptomics with spatial and multimodal capabilities. This platform overcomes key limitations of existing methods, improving scalability, flexibility, and precision in profiling cell-cell interactions, microenvironments, and subcellular gene expression. These improvements underscore the expansive potential of Stereo-cell across a spectrum of applications.