Murine and human very small embryonic-like cells: A perspective



In 2006, very small embryonic-like (VSEL) stem cells were described as a pluripotent population of prospectively isolated stem cells in adult murine bone marrow (mBM) and human umbilical cord blood (hUCB). While rigorous proof of pluripotency is still lacking, murine VSEL cells have been shown to overlap with an independently identified population of neural crest derived mesenchymal stem cells (MSC). The presence of primitive mesenchymal precursors within the VSEL cell population may partially explain the findings that have led to the concept of an “embryonic-like” stem cell in mBM. However, our own studies on human VSEL cells revealed very little similarity between murine VSEL cells and their reportedly equivalent population in hUCB. On the contrary, our data strongly suggest that human VSEL cells are an aberrant and inactive population that cannot expand in vitro and has neither embryonic nor adult stem cell like properties. Here we critically re-examine the data supporting stemness and pluripotency of murine and human VSEL cells, respectively. © 2012 International Society for Advancement of Cytometry


The original description in 2006 of a cell population with embryonic stem cell-like characteristics in mouse bone marrow and human umbilical cord blood (hUCB) joins a number of reports over the years that have raised hopes of a non-embryonic source of pluripotent stem cells for autologous regenerative therapies (1, 2). However, unlike previous reports of extended regenerative potential arising under extended mesenchymal stem cell (MSC)-like culture conditions (3–5), the very small embryonic-like (VSEL) stem cells could be isolated prospectively from primary tissue using defined surface markers. Although the availability of well defined phenotypic markers should increase the reproducibility of identification and isolation between labs and make the detailed characterization of stem cell potential more straightforward, six years after the original publication there has been no independent confirmation of the data and no functional proof of the putative pluripotent nature of these cells (6). On the contrary, the evidence for pluripotency of adult cells has been repeatedly called into question (7–10). Therefore, it is time for a critical re-examination of the VSEL hypothesis and the data supporting stemness and pluripotency of the human and murine VSEL populations.

Murine VSEL Cells

Murine VSEL cells were first described as a CD45LinSca-1+ population in adult murine bone marrow (mBM) that showed high transcript levels of pluripotency genes, a nuclear localization of both Oct 4 and Nanog proteins, a demethylated Oct4 promotor, an overall euchromatic nuclear organization, a quiescent status that is maintained by a unique methylation pattern of certain imprinting genes, and an in vitro differentiation potential into neuronal, pancreatic, and cardiomyocyte-like cells representative of ectoderm, endoderm, and mesoderm, respectively (2, 11). The putative pluripotency of murine VSEL cells remains based on these preliminary findings, there having been in the meantime no demonstration of teratoma formation, tissue contribution following blastocyst injection or multilineage differential at a clonal level. Thus, the gold standards of pluripotent stem cell function have still not been demonstrated, and the mere detection of pluripotency-associated markers at mRNA and protein level must be regarded as an insufficient basis from which to claim pluripotency, especially as the assays used to detect the markers are prone to artifacts such as false-positives (7, 8, 12). While the case for pluripotency and quiescence of VSEL cells has not been convincingly made, murine VSEL cells have indeed been shown to have proliferation capacity and differentiation potential in vitro (2, 13). Furthermore, murine VSEL cells are able to form bone tissue in vivo, and may contribute to heart regeneration (14, 15), although the same cells failed in hematopoietic reconstitution assays and in teratoma formation assays (2, 11). Most recently, a population similar to murine VSEL cells has been identified in rat bone marrow and has been shown both to expand and differentiate in vitro and to contribute to heart regeneration in vivo (16). These data are not definitive proof of a specific adult stem cell function and are not compatible with the pluripotent and quiescent status claimed for murine VSEL, but they would be consistent with a bone marrow VSEL population containing mesenchymal stem or progenitor cells with in vivo regenerative potential.

Interestingly, supporting evidence in this respect has been provided by an independent Japanese group with the identification of a mesenchymal precursor overlapping with the CD45LinSca-1+ VSEL cell population (Fig. 1) (17). The CD45Ter119Sca-1+PDGFR-α+ (PαS) cells give rise to multipotent mesenchymal stromal cells in vitro, and are able to engraft in bone marrow when transplanted i.v. into irradiated mice. Furthermore, it was shown that at least some of the PαS cells originate from the neural crest and can give rise to neuronal cell types in vitro, suggesting that there may be a mixed mesodermal/ectodermal potential in the PαS population (18). Consistent with this, it has been shown that Nestin, a marker previously associated with neuronal stem cells, specifies the entire mesenchymal potential of the CD45 bone marrow population (19, 20). The fact that mesenchymal stem cells with very similar properties have been independently isolated from PαS cells and from CD45Nestin+ cells suggests a degree of overlap also between these populations.

Figure 1.

Heterogeneity of the murine VSEL cell population. The gating of VSEL cells shows that ∼60% of the VSEL cells express PDGF-Rα, thus resembling the PαS phenotype. A: Gating of Lin mBM cells (CD5, CD45R, CD11b, Ly-6G/C, 7-4, Ter-119; lineage cell depletion kit, Miltenyi Biotec). B: Lin cells: Gating of CD45-Sca-1+ VSEL cells. C: LinCD45+ cells: gating of c-kit+Sca-1+ hematopoietic stem and progenitor cells (KSL). D: Expression of PDGF-Rα in KSL cells and VSEL cells. Further experimental details are provided as Supporting Information.

In summary, these examples demonstrate: (1) that mBM VSEL cells do contain multipotent mesenchymal stem cells, (2) that the mBM VSEL population is heterogeneous and currently insufficiently characterized, and (3) that an extra-mesenchymal differentiation potential of the VSEL population (into the neuronal lineage in this case) can be explained without invoking a pluripotent stem cell. The set of recently identified markers that allow a prospective isolation of mesenchymal stem cells from mBM should also help to clarify the true nature of the murine VSEL population.

Human UCB VSEL Cell Phenotype

Clearly, the degree to which the perceived therapeutic potential of VSEL or related populations can be realized depends entirely on the presence of similar cells in human tissues. It was therefore encouraging that the description of VSEL in murine tissues was followed rapidly by a report from the same group of VSEL cells in hUCB (1). These data were fundamental to the definition of human VSEL cells as CD45LinCXCR4+ cells. The CD133 and CD34 markers were also described as being expressed by VSEL cells, but the data provided was not sufficient to clarify the relationship between the subsets of CXCR4+, CD133+, and CD34+ cells within the CD45Lin cell population. The transcript levels of the selected pluripotency genes including Oct4 and Nanog were indistinguishable between the CD45LinCXCR4+ and CD45LinCD133+ populations, but there was no indication that CD133 should be preferred in any way to CXCR4 as a positive marker for VSEL cells. Indeed, the CD45LinCXCR4+ cells were the only ones to be further characterized by the expression of Oct4 and Nanog on the protein level and by electron microscopy.

Following this initial definition of the human VSEL phenotype, alternative strategies for the isolation of VSEL from human UCB via magnetic enrichment of CD133+ cells have been published (21, 22). However, none of these reports provide functional indications of VSEL cell-like activity, let alone evidence of pluripotency, that would support superiority of CD133 over CD34 or CXCR4 for VSEL isolation. Until functional comparisons have been performed between the original CD45LinCXCR4+ and other populations such as the CD45LinCD133+, it seems sensible to base any comparison of data between different groups on the originally and more comprehensively described CD45LinCXCR4+ VSEL population.

CD45LINCXCR4+ VSEL Cells from Human UCB Lack Stem Cell Activity

With the aim of defining the true therapeutic potential of human VSEL cells, we have carried out an extensive characterization of purified populations of CD45LinCXCR4+ VSEL cells from hUCB (10). Consistent with the original description of these cells, our purified populations have a high density, are significantly smaller than hematopoietic stem cells, and appear to possess hardly any cytoplasm. However, a comprehensive flow cytometric screen for 22 surface markers typical of embryonic and adult stem cells revealed negligible expression on VSEL cells. Contrary to our expectations, purified VSEL cells also failed to expand in vitro under a range of culture conditions commonly used in embryonic or adult stem cell culture. Since our purification protocol excluded neither CD133+ nor CD34+ cells from the CD45LinCXCR4+ VSEL population, any stem cell activity arising from a CD133+ or CD34+ subpopulation should have been detected in our cell culture experiments.

In order to further extend our analysis and in the hope of gaining some indication of how to better culture and expand VSEL cells in vitro, we conducted a microarray analysis. The transcriptional profile of highly purified hUCB VSEL cells contradicted the previous report by Ratajczak's group and showed the VSEL cells to be clearly distinct both from the mature hematopoietic lineages, and from well-defined populations of pluripotent and adult stem cells (1). We further conducted a karyotypic analysis that revealed an aneuploid karyotype in the majority of purified VSEL cells by fluorescence in situ hybridization. Although aneuploidy has long been associated with cancer, it has recently been observed in cultured pluripotent and neuronal stem cells as well as normal neuronal progenitors and primary cells from blastocysts, so that the tendency to generate aneuploid cells may be a normal feature of regenerative systems (23–26). Even so, the aneuploid products themselves are unlikely to make a long-term contribution to regeneration, and it is unsurprising that these cells are unable to expand in vitro.


There can be little doubt that the VSEL cells isolated from mBM contain very promising stem cell functions, including the capacity to generate mesenchymal stem cells with high efficiency (Table 1). However, in the continuing absence of functional data there is no basis in our view to assume pluripotency. It is to be hoped that MSC similar to those present in the mBM VSEL population are present in human tissues. However, our studies on the reportedly equivalent “VSEL” population derived from hUCB strongly suggest that these are an aberrant and inactive population that cannot expand in vitro and has neither an embryonic nor an adult stem cell like phenotype. There has been no confirmation by an independent group to date to support that human UCB-derived VSEL cells possess the functional hallmarks of stem cells, such as high proliferative capacity, in vitro differentiation potential, or the ability to self-renew in vivo. In the absence of such evidence and in the light of our own data, we currently see no justification for use of the term “stem cell” in reference to this population.

Table 1. Phenotype and differentiation potential of prospectively isolated VSEL and related adult stem cells
OriginPhenotypeExpansion in vitroDifferentiation potentialPluripotency markersElectron micrographyRef.
mBMCD45LinSca-1+YesNeuronal, pancreatic and  cardiomyocyte like cellsmRNA, proteinYes(2)
mBMCD45Ter119 Sca-1+PDGFRα+YesAdipocyte, chondrocyte, osteoblast,  endothelial cell; neuronal like cellsnot testedNo(17, 18)
mBMCD45nestin+YesAdipocyte, chondrocyte, osteoblastnot testedNo(19)
rat BMSSEA-1+(CD45Lin)YesCardiomyocyte, endothelial cellmRNA, proteinYes(16)
hUCBCD45LinCXCR4+NoNonemRNA, proteinYes(1)


The authors declare the following conflicts of interest: Ralitza Danova-Alt was employed by Vita34 AG during the course of this project. Dietmar Egger is employed by Vita 34 AG. Rüdiger Alt received co-funding and cord blood samples from Vita 34 AG while working on this project at the Universität Leipzig and is now employed by Vita 34 AG.