Genetic and functional differences between multipotent neural and pluripotent embryonic stem cells

Abstract

Stem cells (SCs) are functionally defined by their abilities to self-renew and generate differentiated cells. Although much effort has been focused on defining the common characteristics among various types of SCs, the genetic and functional differences between multipotent and pluripotent SCs have garnered less attention. We report a direct genetic and functional comparison of molecularly defined and clonally related populations of neural SCs (NSCs) and embryonic SCs (ESCs), using the Sox2 promoter for isolation of purified populations by fluorescence-activated cell sorting. A stringent expression profile comparison of promoter-defined NSCs and ESCs revealed a striking dissimilarity, and subsequent chimera analyses confirmed the fundamental differences in cellular potency between these populations. This direct comparison elucidates the molecular basis for the functional differences in pluripotent ESCs and multipotent NSCs.

Fig. 1.

P/Sox2-EGFP identifies a subset of Sox2-expressing cells that are NSCs. (A) Structure of the transgene. (B and C) The 1D2 ESC clone exhibits telencephalic-restricted expression at E14, and the localization of P/Sox2-EGFP expression is in the cell-dense neuroepithelium surrounding the lateral ventricles but not in differentiating cells identified by βIII-tubulin (red). (D) FACS isolation of P/Sox2-EGFP high, low, and negative populations from short-term cultures derived from E14 fetuses. The total proportion of expressing cells represents ≈8% of the sorted population that was further subdivided based on high and low expression levels. (E) Quantitation of Sox2 mRNA in populations with varying fluorescence intensity by real-time quantitative PCR, relative to the control Hprt message, exhibits correlation with fluorescence intensity. Notably, the P/Sox2-EGFP^- fraction still contains cells expressing significant levels of Sox2 mRNA, indicating that the transgene marks only a subset of cells that express the Sox2 gene. (F) Selection of P/Sox2-EGFP-expressing cells provides a 20-fold enrichment in neurosphere-initiating activity over nonexpressing cells from cultured populations. (G) All clones derived from single P/Sox2-EGFP cells are multipotent and differentiate into neurons, astrocytes, and oligodendrocytes, as indicated by βIII-tubulin (red), GFAP (purple), and RIP (green) immunoreactivity.

Fig. 2.

Expression profiling experimental design. Three populations of cells isolated by using FACS for P/Sox2-EGFP expression are compared. The three populations are of clonal origin because they all originate from the 1D2 clone of P/Sox2-GIP ESCs by virtue of isolating the NSCs from E14 chimeric fetuses. Three replicates for each population were derived from separate flasks of cells and/or different pools of chimeric embryos. The Affymetrix platform was used for generating expression profiles beginning with 5 μg of total RNA for each replicate. Three independent methods were used for data analysis, and the genes found to be differentially expressed in multiple methods are considered to be different between populations with a very high degree of confidence. Some of the expected genes found to be enriched in ESCs (red), NSCf (green), and NSCc (blue) are listed, as well as genes that are expressed at equivalent levels in all three populations (black).

Fig. 3.

Only P/Sox2-EGFP-defined ESCs are pluripotent. P/Sox2-defined ESCs and NSCs were injected into blastocysts followed by analysis at E3.5 and E9.5 time points. (A and D) There is a similar frequency of adherence to the ICM for ESCs and NSCs injected into blastocysts. Only ESCs participate in development to form highly chimeric embryos at E9.5 (B and C), whereas the P/Sox2-EGFP cells derived from E14 fetuses do not exhibit chimerism to any tissues (E and F). (G) The numbers of blastocysts injected and E9.5 embryos recovered are presented for P/Sox2-defined NSCs and ESCs. The frequency of recovering an E9.5 embryo is equivalent for NSC and ESC injections, indicating that no toxicity is imparted by the injected NSCs. Therefore, the lack of chimerism in the 76 NSC-injected embryos functionally identifies these P/Sox2-EGFP⁺ cells distinctly as multipotent SCs.

Fig. 4.

RT-PCR confirmation of presence/absence in ESC and NSC populations. Groups of SC class-specific transcription factors of high confidence (2.0 or greater fold change) were assayed by RT-PCR as a more sensitive method for detection of rare transcripts. Most of the ESC-specific genes were verifiably absent in NSCs. Only two of four NSC-specific genes were found absent from ESCs. Gapdh and β-actin gene signals are approximately the same, indicating an equivalent input for both cell types.