Dissociation of solid tumor tissues with cold active protease for single-cell RNA-seq minimizes conserved collagenase-associated stress responses - PubMed (original) (raw)

doi: 10.1186/s13059-019-1830-0.

Kieran R Campbell 1 2 3, Allen W Zhang 1 4 5, Farhia Kabeer 1 6, Jamie L P Lim 7, Justina Biele 1, Peter Eirew 1, Daniel Lai 1, Andrew McPherson 1 7, Esther Kong 1, Cherie Bates 1, Kelly Borkowski 1, Matt Wiens 1, Brittany Hewitson 1, James Hopkins 1, Jenifer Pham 1, Nicholas Ceglia 4, Richard Moore 8, Andrew J Mungall 8, Jessica N McAlpine 9; CRUK IMAXT Grand Challenge Team; Sohrab P Shah 10 11 12, Samuel Aparicio 13 14

Affiliations

Dissociation of solid tumor tissues with cold active protease for single-cell RNA-seq minimizes conserved collagenase-associated stress responses

Ciara H O'Flanagan et al. Genome Biol. 2019.

Abstract

Background: Single-cell RNA sequencing (scRNA-seq) is a powerful tool for studying complex biological systems, such as tumor heterogeneity and tissue microenvironments. However, the sources of technical and biological variation in primary solid tumor tissues and patient-derived mouse xenografts for scRNA-seq are not well understood.

Results: We use low temperature (6 °C) protease and collagenase (37 °C) to identify the transcriptional signatures associated with tissue dissociation across a diverse scRNA-seq dataset comprising 155,165 cells from patient cancer tissues, patient-derived breast cancer xenografts, and cancer cell lines. We observe substantial variation in standard quality control metrics of cell viability across conditions and tissues. From the contrast between tissue protease dissociation at 37 °C or 6 °C, we observe that collagenase digestion results in a stress response. We derive a core gene set of 512 heat shock and stress response genes, including FOS and JUN, induced by collagenase (37 °C), which are minimized by dissociation with a cold active protease (6 °C). While induction of these genes was highly conserved across all cell types, cell type-specific responses to collagenase digestion were observed in patient tissues.

Conclusions: The method and conditions of tumor dissociation influence cell yield and transcriptome state and are both tissue- and cell-type dependent. Interpretation of stress pathway expression differences in cancer single-cell studies, including components of surface immune recognition such as MHC class I, may be especially confounded. We define a core set of 512 genes that can assist with the identification of such effects in dissociated scRNA-seq experiments.

Keywords: Breast cancer; Gene expression; Ovarian cancer; Quality control; RNA-seq; Single cell; Tissue dissociation; Tumor microenvironment.

PubMed Disclaimer

Conflict of interest statement

SPS and SA are founders, shareholders, and consultants of Contextual Genomics Inc. The other authors declare that they have no competing interests.

Figures

Fig. 1

Fig. 1

Overview of 48 single-cell experiments generated in this study. a Schematic showing the various substrates used to generate the 48 single-cell experiments in this dataset. b Descriptions of the cell status, substrate, cancer type, dissociation temperature, and tissue state of each sample in the dataset. c Substantial variability in three key QC metrics (number of genes detected, percentage of counts mapping to the mitochondrial genome, number of UMIs sequenced) across all experiments. d Embedding of all 48 single-cell experiments to a low-dimensional projection with uniform manifold approximation and projection [12]

Fig. 2

Fig. 2

Transcriptomic landscape of live, dead, and dying cells. a FACS analysis showing gating strategy for untreated, live cells (PI−/annexin V−) or TNFα-treated dying cells (PI/annexin V+) and dead cells (PI+/annexin V+). b PCA projection of the three cell conditions showing approximate segregation of cell status along the first principal component (PC1), with live and dying cells enriched at lower PC1 values and dead cells enriched at higher values. c PCA projection colored by the percentage mitochondrial genes (“% transcriptome mitochondrial”) shows significant increase along the PC1. d Dead cells exhibit significantly higher percentage of the transcriptome as mitochondrial compared to both live and dying cells. e Unsupervised clustering of the gene expression profiles clusters the cells into three groups, approximately tracking both PC1 of the data and the percentage of transcriptome mitochondrial. f The composition of each cluster demonstrates that cluster 1 is primarily composed of live cells and cluster 2 a mix of live, dying, and dead cells, while cluster 3 is composed mainly of dead cells. g The percentage of transcriptome mitochondrial is significantly different between the three clusters, with a step increase in proportion moving from cluster 1 to 2 and 2 to 3. h Cluster 2 significantly upregulates the MHC class I gene set, suggesting it represents stressed or pre-apoptotic cells. i Differential expression analysis of transcriptomically “healthy” cells within cluster 1 reveals residual differences between cells sorted as live and dead. j The distribution of absolute effect sizes (log fold change) of live vs. dead cells within cluster 1 (_x_-axis) compared to between clusters 1 and 2 (_y_-axis) demonstrates the residual effect on the transcriptome of being live/dead sorted is small compared to the inter-cluster expression variance

Fig. 3

Fig. 3

Dissociation with collagenase at 37 °C induces a distinct stress response in 23,731 cells from PDX samples that is minimized by dissociation at 6 °C. a The top 40 genes (by log fold change) from the 11,975 identified as significantly differentially expressed between cells digested at 6 °C and 37 °C. b UMAP plots of 23,731 cells colored by digestion temperature (top) then by normalized expression of three key stress response genes (FOS, JUNB, NR4A1) demonstrate a distinct concordance between temperature and induction of the stress gene signature. Expression values are log normalized counts winsorized to [0, 2) then scaled to [0, 1). c Pathway analysis of differentially expressed genes with the MSigDB hallmark gene sets highlights induction of genes involved in NF-κB signaling at 37 °C digestion with 46.5% of 200 genes annotated in the pathway being found in the 512 core gene set

Fig. 4

Fig. 4

Disentangling the effects of digestion time and digestion method on transcriptomic response. a Mean normalized expression of genes in the core gene set as a function of digestion time colored by digestion temperature. Digestion by collagenase causes upregulation of the gene set at all time points, with a subset showing further upregulation as digestion time increases. B Log fold changes of a 2-h vs. 30-min digestion for collagenase only as a function of log counts-per-million. c Log fold changes of a collagenase vs. cold protease digestion at 30-min digestion time as a function of log counts-per-million. d Log fold changes of a collagenase vs. cold protease digestion at 2-h digestion time as a function of log counts-per-million. e Log fold changes of a 2-h vs. 30-min digestion (collagenase only) compared to a collagenase vs. cold protease digestion at 2 h demonstrate a large overlap between genes affected (ρ = 0.8)

Fig. 5

Fig. 5

Conserved stress response to the collagenase dissociation method in breast and ovarian patient tissues. a Histology of ovarian (top) and breast (bottom) cancer patient samples highlighting the architecture of the tumor microenvironment. b FACS analysis of ovarian tumor tissue dissociated at 37 °C with collagenase or 6 °C with cold active protease and stained with markers for tumor cells (EpCAM), endothelial cells (CD31), fibroblasts (FAP), lymphocytes (CD45), B cells (CD19), NK cells (CD56), and T cells (CD8, CD3). c UMAP of combined scRNA-seq experiments of ovarian cancer (n = 2) and breast cancer (n = 3) patient tissues with cell type assignments according to known gene markers for each cell type. d The top 40 genes from the gene set derived in Fig. 3 as expressed in each cell type in breast and ovarian patient samples. Black circles around points denote significance at 5% FDR. e Pathway analysis of the differential expression results with the MSigDB hallmark gene sets for each cell type

Similar articles

Cited by

References

    1. Potter S. Steven. Single-cell RNA sequencing for the study of development, physiology and disease. Nature Reviews Nephrology. 2018;14(8):479–492. doi: 10.1038/s41581-018-0021-7. - DOI - PMC - PubMed
    1. Stegle O, Teichmann SA, Marioni JC. Computational and analytical challenges in single-cell transcriptomics. Nature Reviews Genetics. 2015;16(3):133. doi: 10.1038/nrg3833. - DOI - PubMed
    1. Raj A, van Oudenaarden A. Nature, nurture, or chance: stochastic gene expression and its consequences. Cell. 2008;135(2):216–226. doi: 10.1016/j.cell.2008.09.050. - DOI - PMC - PubMed
    1. Islam S, et al. Quantitative single-cell RNA-seq with unique molecular identifiers. Nature Methods. 2014;11(2):163. doi: 10.1038/nmeth.2772. - DOI - PubMed
    1. Volovitz I, et al. A non-aggressive, highly efficient, enzymatic method for dissociation of human brain-tumors and brain-tissues to viable single-cells. BMC Neuroscience. 2016;17(1):30. doi: 10.1186/s12868-016-0262-y. - DOI - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources