CD24 tracks divergent pluripotent states in mouse and human cells

Abstract

Reprogramming is a dynamic process that can result in multiple pluripotent cell types emerging from divergent paths. Cell surface protein expression is a particularly desirable tool to categorize reprogramming and pluripotency as it enables robust quantification and enrichment of live cells. Here we use cell surface proteomics to interrogate mouse cell reprogramming dynamics and discover CD24 as a marker that tracks the emergence of reprogramming-responsive cells, while enabling the analysis and enrichment of transgene-dependent (F-class) and -independent (traditional) induced pluripotent stem cells (iPSCs) at later stages. Furthermore, CD24 can be used to delineate epiblast stem cells (EpiSCs) from embryonic stem cells (ESCs) in mouse pluripotent culture. Importantly, regulated CD24 expression is conserved in human pluripotent stem cells (PSCs), tracking the conversion of human ESCs to more naive-like PSC states. Thus, CD24 is a conserved marker for tracking divergent states in both reprogramming and standard pluripotent culture.

Figure 1. Surface proteome analysis during reprogramming identifies CD24 as a differentially expressed surface marker.

(a) Principal component analysis of the surface proteome (including only the subset of surface proteins that were also present in the global proteome screen to improve rigour of quantitative analysis) of reprogramming secondary MEF 1B cells derived from tetraploid complementation, showing divergent routes of F-class and ESC-like iPSCs. (b) Summary of sampling time course and DOX treatment protocols. (c) Representative flow cytometry plots of CD24 versus SSEA1 expression during reprogramming of secondary MEF 1B cells derived from tetraploid complementation, revealing emerging CD24^high/SSEA1+ (CD24H) and CD24^low/SSEA1+ (CD24L) subpopulations. ESC control is included for comparison. Flow plots are representative from three biological replicates.

Figure 2. CD24H and CD24L subpopulations correspond to transgene-dependent (F-class) and ESC-like iPSCs, respectively.

(a) Summary of approach to characterize CD24H and CD24L populations. (b) Percentage of CD24H and CD24L cells in DOXH, DOXL− and DOXH− culture time courses. Data bars show mean±s.d. (n=3 biological replicates). (c) Representative phase contrast images (n=3 biological replicates) of emerging colonies in the three DOX treatments, including ESC control. Scale bar, 125 μm. (d) Summary of sorting strategy used to separate CD24H and CD24L subpopulations from D30 culture. Representative phase contrast images (n=3 technical replicates) of sorted CD24H and CD24L cells in DOXH and DOX− conditions. Scale bar, 125 μm. (e) Effect of long-term passaging on levels of CD24H and CD24L cells in D30-sorted DOXH CD24H, DOXL− CD24L and DOXH− CD24L cells. ESCs are included as a control. Data bars show mean±s.d. (n=3 technical replicates). (f) EdU staining of DOXH CD24H, DOXL− CD24L and DOXH− CD24L cells in DOXH and DOX− conditions to assess DOX dependence of proliferation. Data bars show mean±s.d. (n=3 technical replicates). (g) Expression levels of various pluripotent and F-class-specific genes in CD24H and CD24L cells with unsupervised hierarchical clustering, including F-class and ESC cell controls, normalized to Gapdh and MEF cells.

Figure 3. CD24 and CD40 exhibit differential expression on mouse ESCs and EpiSCs.

(a) Flow cytometry analysis of CD24/SSEA1/CD40 expression in mouse ESCs and EpiSCs. Representative flow plots (n=3 technical replicates) of stained mouse ESC and control embryo-derived EpiSCs. (b) Gene expression of CD24^high/CD40+ cells derived from R1 ESC culture compared with control embryo-derived EpiSCs. The gene panel is composed of pluripotency genes as well as some early differentiation markers. Values are normalized to sorted CD24^low/SSEA1+ (ESC-like) cells sorted from R1 ESC culture and Gapdh. Data bars show mean±s.d. (n=3 technical replicates).

Figure 4. CD24 delineates 'primed' and 'naive' pluripotent states in human cells.

(a) Flow cytometry analysis of CD24 expression in ‘primed' and ‘naive' human ESC lines following passages 6 and 10 in naive conditions. Flow plots are representative from three technical replicates. (b) CD24/Tra-1-60 staining of primed hES2 and naive-induced hES2 and H9 hESCs at indicated number of passages in naive conditions. Flow plots are representative from three technical replicates. (c) Gene expression analysis of CD24^high/Tra-1-60+ and CD24^low/Tra-1-60+ cells sorted from primed and naive culture at indicated passage number with unsupervised hierarchical clustering. Expression data are normalized to GAPDH and primed hES2 cells. (d) Gene expression analysis comparing unsorted, CD24H-sorted and CD24L-sorted hES2 cells. Expression data are normalized to GAPDH and primed hES2 cells. Statistical significance is assessed by a Student's t-test (heteroscedastic, two-sided). **P<0.05, *P<0.1. Data bars show mean±s.d. (n=3 technical replicates).

Figure 5. Summary of CD24 as a marker for three pluripotent states.

(a) Overview of CD24 as a tool for tracking transgene induction in early reprogramming, demarcating divergent reprogramming to F-class and ESC-like iPSC states in late reprogramming, and identifying multiple pluripotent states in transgene-independent pluripotent culture. (b) Summary of approach for identifying mouse F-class iPSCs, ESC-like iPSCs and EpiSCs via CD24/SSEA1/CD40 staining as well as human primed and naive hESCs via CD24/Tra-1-60 staining.