Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Jun 4;16(6):712-24.
doi: 10.1016/j.stem.2015.04.004. Epub 2015 May 21.

Combined Single-Cell Functional and Gene Expression Analysis Resolves Heterogeneity within Stem Cell Populations

Nicola K Wilson  1 , David G Kent  1 , Florian Buettner  2 , Mona Shehata  3 , Iain C Macaulay  4 , Fernando J Calero-Nieto  1 , Manuel Sánchez Castillo  1 , Caroline A Oedekoven  1 , Evangelia Diamanti  1 , Reiner Schulte  5 , Chris P Ponting  6 , Thierry Voet  7 , Carlos Caldas  3 , John Stingl  3 , Anthony R Green  1 , Fabian J Theis  8 , Berthold Göttgens  9
Affiliations

Combined Single-Cell Functional and Gene Expression Analysis Resolves Heterogeneity within Stem Cell Populations

Nicola K Wilson et al. Cell Stem Cell. .

Abstract

Heterogeneity within the self-renewal durability of adult hematopoietic stem cells (HSCs) challenges our understanding of the molecular framework underlying HSC function. Gene expression studies have been hampered by the presence of multiple HSC subtypes and contaminating non-HSCs in bulk HSC populations. To gain deeper insight into the gene expression program of murine HSCs, we combined single-cell functional assays with flow cytometric index sorting and single-cell gene expression assays. Through bioinformatic integration of these datasets, we designed an unbiased sorting strategy that separates non-HSCs away from HSCs, and single-cell transplantation experiments using the enriched population were combined with RNA-seq data to identify key molecules that associate with long-term durable self-renewal, producing a single-cell molecular dataset that is linked to functional stem cell activity. Finally, we demonstrated the broader applicability of this approach for linking key molecules with defined cellular functions in another stem cell system.

PubMed Disclaimer

Figures

None
Graphical abstract
Figure 1
Figure 1
Single-Cell Expression Analysis Reveals an Overlapping Molecular Signature for Four Heterogeneous HSC Populations (A) Schematic of the hematopoietic tree. The cell types highlighted are populations that will be further investigated within this study; the colors and names remain constant throughout the text. The individual sorting strategies are also highlighted next to the appropriate cell population. HSC1 (dark blue, Lin−c-kit+Sca-1+CD34−Flt3−CD48−CD150+), HSC2 (pink, Lin−CD45+EPCR+CD48−CD150+), HSC3 (cyan, Lin−c-kit+Sca-1+CD34−Flt3−), HSC4 (orchid, Lin−c-kit+Sca-1+SP CD150+), HSC5 (seagreen, Lin−CD45+EPCR+CD48−CD150−), LMPP (yellow, Lin−c-kit+Sca-1+CD34+Flt3+), CMP (red, Lin−c-kit+Sca-1−CD34+FcγRlow), MEP (yellow-green, Lin−c-kit+Sca-1−CD34−FcγRlo) and GMP (orange, Lin−c-kit+Sca-1−CD34+FcγRhi). (B) Flow diagram of single-cell qRT-PCR. (C) Unsupervised hierarchical clustering of gene expression for all investigated cell populations. Colored bar (population) above heat map indicates the cell population (colors are the same as in A). Intensity of heat map is based on the ΔCt, black is highest expressed—dark blue is lowest, and gray is not detected. The distances of the population dendrogram are not proportional to the dissimilarity. See also Figure S1 and Table S1.
Figure 2
Figure 2
Multidimensional Analysis Can Further Resolve Cell Populations (A) t-SNE plot of all cells calculated from the 43 genes analyzed by Fluidigm. All HSCs are circles and all progenitors are diamonds. Axes are in arbitrary units. (B) Table of the published repopulation data used for the weighting program and schematic of the computational weighting program. (C) Schematic showing the definition of MolO cells. (D) t-SNE plot as in (A) with the MolO HSCs identified by the computational weighting highlighted in red. Axes are in arbitrary units. Table showing differentially regulated genes between MolO and NoMO populations. Red, genes upregulated in MolO population; blue, genes downregulated in MolO population. See also Figure S2.
Figure 3
Figure 3
Genome-wide Expression Pattern of 92 Single HSCs Reveals a Gene Signature for the MolO Population (A) RNA-seq analysis. (i) Identification of variable genes across all 92 cells. The genes highlighted in magenta have a coefficient of variation exceeding technical noise. The blue dots represent the distribution of the internal control ERCC spike-ins. (ii) PCA plot for the 92 cells analyzed by RNA-seq, showing the first and second components for all genes which were identified to be variably expressed. (iii) Principal component loading plot of scRNA-seq, indicating which genes also assayed by Fluidigm analysis and/or flow cytometry contribute to the separation of the cells along each component. (B) Schematic showing the principle of the classifier to determine the MolO HSCs from the scRNA-seq dataset. (C) PCA plot showing MolO score. (D) Table of signature genes differentially expressed in either NoMO or MolO cells following correction for multiple testing at a false discovery rate (FDR) of 0.1. Coloring relates to the GO term associated with the gene: red, cell cycle; yellow, negative regulation of cell proliferation. See also Figure S3 and Tables S2 and S3.
Figure 4
Figure 4
SLAM Scalo Cells Make Large Differentiated Clones Compared with SLAM Scahi Cells (A) The most discriminating sequence of surface markers resulted in the sorting strategy shown on the right, which first selects CD48−CD150+ cells and then partitions the Sca positive cell fraction into high (SLAM Scahi) and low (SLAM Scalo) levels. The negative Sca-1 population was set at less than 101, meaning all cells were Sca1+. (B) Schematic for single cell in vitro study where single HSCs were cultured in SCF and IL-11 for 10 days and analyzed by flow cytometry. (C) The bar graph shows the cumulative number of cells that reached the first, second, and third division on each of the first four days of culture. First division was determined by the presence of two cells, second by three or more cells, and third by five or more cells. Notably, the SLAM Scahi population entered division significantly later and also had fewer second and third division clones on days 2–4. (D) The pie charts depict the ratio of small (<500), medium (500–5,000), large (5,000–20,000), and very large (>20,000) clones formed from single SLAM Scalo (upper chart) and SLAM Scahi (lower chart). All clones formed by single SLAM Scalo cells were large or very large. (E) Clones were assessed by flow cytometry, and accurate clone sizes were determined using a standard number of fluorescent beads in each well and then back calculated to get total clone size. The clone size (left), percentage of lineage marker expression (middle), and percentage of KSL cells (right) are shown. Notably, SLAM Scahi clones are smaller and less differentiated. Error bars represent data ± SEM. See also Figure S4. ∗p < 0.05, ∗∗p < 0.01.
Figure 5
Figure 5
SLAM Scahi Cells Are Enriched for Long-Term Multilineage HSCs, and Their Single-Cell Transplantation Activity Links to a Distinct Molecular Profile (A) Donor chimerism (% donor/[% donor + % recipient]) in mice receiving either 10 SLAM Scahi or SLAM Scalo cells. Recipients of SLAM Scahi cells have significantly increased levels of donor chimerism. Error bars represent data ± SEM. (B) Individual recipient mice of ten SLAM Scahi or SLAM Scalo cells and the donor contribution to various lineages. Ratios are formed by taking the total cells of a particular lineage (e.g., GM) and calculating the donor contribution (e.g., Donor GM/(Donor + Recipient GM). GM contribution is red, B is blue, and T is green. Note that four of five recipients of SLAM Scalo cells have <1% GM contribution, whereas all five recipients of SLAM Scahi cells have robust myeloid contribution. (C) Donor chimerism (% donor/[% donor + % recipient]) in mice receiving 1 SLAM Scahi cell. Fifteen of 29 mice transplanted had donor chimerism of >1% and are displayed on this graph. Blue indicates beta subtype; red indicates alpha subtype; and green indicates gamma/delta subtypes. (D) Joint representation of sequenced cells and transplanted cells. In the t-SNE space, cells with a high predicted MolO score cluster together with repopulating cells; cells with a low predicted MolO score cluster with mostly non-repopulators. Transplanted cells are represented by squares. White indicates non-repopulators. Black indicates repopulators. Hatch pattern indicates gamma-HSCs and the 1% chimerism beta-HSC highlighted in the main text. Sequenced cells are represented by circles, and the predicted MolO score is shown. Axes are in arbitrary units. (E) The hidden differentiation factor recovered using scLVM was strongly correlated with EPCR expression. Cells with high EPCR expression and low differentiation factor also had a high predicted MolO score (colors as in D). Axes are in arbitrary units. (F) Donor chimerism (% donor/[% donor + % recipient]) in mice receiving 1 ESLAM Scahi cell. Twenty-six of 39 mice transplanted had donor chimerism of >1% and are displayed on this graph. Blue indicates beta subtype. Red indicates alpha subtype, and green indicates gamma/delta subtypes. The asterisk indicates an HSC that had <1% chimerism at 16 weeks, but >1% at 24 weeks. (G) Table of signature genes significantly associated with SuMO and non-SuMO cells. Overlapping genes with the MolO/NoMO gene list are highlighted in green. See also Figure S5 and Table S4.
See this image and copyright information in PMC

Comment in

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Adolfsson J., Borge O.J., Bryder D., Theilgaard-Mönch K., Astrand-Grundström I., Sitnicka E., Sasaki Y., Jacobsen S.E. Upregulation of Flt3 expression within the bone marrow Lin(-)Sca1(+)c-kit(+) stem cell compartment is accompanied by loss of self-renewal capacity. Immunity. 2001;15:659–669. - PubMed
    1. Adolfsson J., Månsson R., Buza-Vidas N., Hultquist A., Liuba K., Jensen C.T., Bryder D., Yang L., Borge O.J., Thoren L.A. Identification of Flt3+ lympho-myeloid stem cells lacking erythro-megakaryocytic potential a revised road map for adult blood lineage commitment. Cell. 2005;121:295–306. - PubMed
    1. Akashi K., Traver D., Miyamoto T., Weissman I.L. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature. 2000;404:193–197. - PubMed
    1. Amir el-A.D., Davis K.L., Tadmor M.D., Simonds E.F., Levine J.H., Bendall S.C., Shenfeld D.K., Krishnaswamy S., Nolan G.P., Pe’er D. viSNE enables visualization of high dimensional single-cell data and reveals phenotypic heterogeneity of leukemia. Nat Biotechnol. 2013;31:545–552. - PMC - PubMed
    1. Anders S., Pyl P.T., Huber W. HTSeq—A Python framework to work with high-throughput sequencing data. Bioinformatics. 2014;31:166–169. - PMC - PubMed

Publication types

Associated data