Multisite Assessment of Methods for Cell Preservation Upstream of Single-Cell RNA Sequencing

Single-cell RNA sequencing (scRNA-seq) resolves cell types and molecular phenotypes within heterogeneous specimens but typically requires fresh, high-quality single-cell suspensions processed immediately to preserve transcriptional profiles. This constraint complicates samples with long preparation times and prevents collection at remote sites lacking single-cell instrumentation. Several commercial assays now enable preservation at the point of collection through fixation or cryopreservation, allowing processing to occur months later. The Association of Biomolecular Research Facilities' DNA Sequencing and Genomics and Bioinformatics Research Groups undertook a cross-platform, multisite study to assess the performance and reproducibility of three such platforms: 10x Genomics FLEX, Parse Biosciences Evercode WT v2, and Honeycomb Bio HIVE.

Total leukocytes and peripheral blood mononuclear cells (PBMCs) were isolated from a single healthy individual, with EasySep reagent used for red blood cell depletion of the leukocyte fraction. Cells were characterized by a 21-color flow cytometry panel to provide a reference, and the remaining material was fixed or cryopreserved according to each platform's protocol. Preserved leukocyte samples were prepared in parallel by two technicians ("A" and "B" replicates) and distributed to multiple ABRF member core facilities for downstream processing, while fresh leukocytes processed with the 10x 3' v3.1 (3pGEX) chemistry served as a reference. Libraries were sequenced at a central site, and performance was evaluated across standard scRNA-seq quality control metrics, gene and transcript detection sensitivity, cell-type discovery and annotation, differential expression, and correlation analyses.

Data from all platforms integrated effectively and produced concordant results for cell-type annotation and relative abundance, with cell-type proportions broadly consistent with the flow cytometry reference. However, platform-specific expression signatures were evident for a subset of genes, and cross-site reproducibility varied between methods, with the FLEX workflow showing greater susceptibility to technical variation introduced during on-site sample processing. Preservation-based methods (FLEX and HIVE) showed better retention of fragile granulocyte populations than fresh samples processed with the 10x 3pGEX workflow.

Improvements to preservation methods are changing how single-cell research is conducted by decoupling sample collection from downstream processing. Our investigation into the performance and reproducibility of each platform provides a resource to help investigators and core facilities select the most appropriate single-cell preservation workflow given their sample type, cell populations of interest, sample collection logistics, and laboratory infrastructure constraints.