Stem cells are promising tools for studying the mechanisms of development and regeneration and for use in cell therapy of various disorders. Unequivocal characterization of stem and progenitor cells is of imminent relevance for their identification after enrichment procedures and before their use in research or for therapy. If we apply stem cells for therapy in regenerative medicine, we have to regard them as “pharmaceutics” meaning that their unequivocal characterization by a set of specific functional or phenotypic markers is crucial. Particularly, we need to be certain that the characterized cells will fulfill the functions in the organism or patient that they are expected to. Stem cells are infrequent already in cord blood but even less frequent if derived from other sources of the organism so that their pre-enrichment is unavoidable.
Besides by their typical tissue-specific localization, stem cell lineages are frequently identified by expression of specific stemness marker proteins and other stem-cell specific epitopes, which are not expressed by somatic cells. However, because of overlapping expression patterns between stem cell lineages, a stem cell can often not be classified by detection of a single marker protein. The increased commercial availability of monoclonal antibodies coupled to fluorochromes with excitation wavelengths proximal to multiple laser lines of a state-of-the-art flow cytometer provides the bases for multi-parameter analysis of over 6–8 markers at the same time (1). Subsequently, multi-color detection by flow cytometry has been turned into an ideal tool for the identification and purification of stem cells with overlapping patterns of phenotypic markers. For instance, analysis of a combination of lineage-positive and -negative marker proteins allows differentiation between hematopoietic and endothelial stem cells by simultaneous polychromatic detection of the expression of various antigens.
Recently, we have summarized phenotypic characteristics of various stem cell types derived from peripheral blood, the eye, tumors, among others (2). This compilation was based on six publications of stem cell phenotypes from the same focus issue. Now the group of Donnenberg (3) characterized different stem and progenitor cells that they derived from human fatty tissue. The authors provide a detailed of specific phenotypic stem cell characteristics and compare them with adipose stromal cells. Furthermore, Porretti et al. (4) used multi-color flow cytometry for identifying rare stem cell lineages in human liver from donors and patients with liver disease. Using this assay, liver endothelial progenitors were identified for the first time in addition to the detection of mesenchymal (MSC) and epithelial stem cells (EpSC). The authors determined that frequencies of stem and progenitor cells change in disease conditions and this observation is particular important, because there is for instance, little knowledge available about MSCs, as these cells are important for the progress of fibrosis in disease, but also participate in liver regeneration.
On the basis of these new data, we updated our previous phenotypic table by a set of seven new cell types isolated from additional organs and include a set of additional markers not present in the earlier overview (2). Therefore, we asked all authors to revisit the table and upgrade their part either by own data or based on additional literature, as necessary. This resulted in the upgraded Table 1 showing phenotypes of very small embryonic-like stem cells (vSELs) (5, 6), neural stem cells (NSCs) (7–16), hematopoietic stem cells (HSCs) from two organs (4, 17), MSCs (4, 18), EpSC (4), limbal epithelial stem cells (LSCs) (19), endothelial progenitor cells (EPCs) from different organs (3, 4, 20–25), supra-adventitial adipose stromal cells (SA-ASCs) (3, 22–25), adipose pericytes (pericyte) (3, 22–25), and finally cancer stem cells (CSCs) (26).
|Organ/tissue||Various||Bone marrow||Adult liver||Adult liver||Adult liver||Eye||Peripheral blood||Adipose tissue||Adult liver||Adipose tissue||Adipose tissue||Tumor|
|References||I (5,6)||II (7–16)||III (17)||IV (4)||V (4,18)||VI (4)||VII (19)||VIII (20,21)||IX (3,22–25)||X (4)||XI (3,22–25)||XII (3,22–25)||XIII (26)|
|VEGF-R2 (KDR, Flk-1)||−||−||−||+||+||+|
|Hoechst side populationf||+/−||+||+||+||+|
|Size||3.63 μmg, 6.58 μmg||10.0–16.0 μmg|
By closely looking at this compilation, several gaps become immediately evident. First of all, the phenotypic markers analyzed vary substantially. With the exception of the expression of very few markers, such as, CD34, CD45, CD105, and CD133 that have been determined virtually on all cell types, reports on the majority of phenotypes are rather patchy. Therefore, it is unclear how the expressions of these markers applies for the other stem cells types and if they would be important to unequivocally tell one type from the other. Secondly, we can now compare phenotypes of HSC as well as EPC-derived from different sources with one another. With respect to EPCs, there is a substantial overlap in the range of markers analyzed and the expressions are in agreement. Still there are some discrepancies. Phenotypic analysis of stem-cell specific markers revealed that liver EPCs characterized by the lineage marker KDR and the expression of CD146 were either CD45+ or CD45−, being in agreement with observations from peripheral blood and adipose endothelial cells that CD45 immunoreactivity cannot be used for identification of EPCs. Similarly, expression levels for CD90 and CD146 seem to differ between EPCs from liver and fatty tissue (see Table 1). HSCs from bone marrow (murine) and liver (human), characterized by expression of CD45, differ in their expression of CD34 and CD133 showing again that the detection of these latter two markers is insufficient for HSC identification.
It is impossible to deduce from these data that the meaning of differences between phenotypes of (virtually) the same stem cell type and many questions still remain open. Are these discrepancies due to organ specific differences of stem cells otherwise similar in function? Are they from common origin and organ specific environment induces the different expression patterns? Or, do the phenotypes differ because of different isolation and enrichment procedures? As an example, EPCs from peripheral blood are isolated by density gradient centrifugation, for other sources, the first step of separation is the digestion in combination with (fatty tissue) or without mechanical treatment (liver) and then EPCs are purified by cell sorting. Most critically, if we speak of a certain stem or progenitor cell type are we indeed dealing with identical cells or with different subsets?
First of all, much attention needs to be put on the correct setting of gates for flow cytometry analysis. Data analysis can be challenging in stem cell biology as there is often no clear distinction between positive and negative populations (27). In addition, expression levels of stem cell markers are also related with the method of cell isolation (28). For instance, protease treatment for cell dissociation interferes with the expression of the epitopes measured in following titration assays. Among various proteases tested, Liberase-1 had less impact on CD133 expression, whereas papain treatment strongly affected CD24 and CD133 expression (28). Moreover, a variation in the expression pattern has also been observed for bone-marrow stem cells obtained from different donors (29). Most surprisingly, some of these stem cell populations expressed nestin, Enolase2, and microtubule associated protein 1b (MAP1b), which are characteristic for cells undergoing neural differentiation. In fact, neural stem and progenitor cells forming neurospheres were shown to differ in their multipotency. Neural precursor cells were identified by CD133, CD15, CD24, A2B5, and PSA-NCAM expression. However, the highest mean levels of CD133 and CD15 were detected in neural progenitor cells, whereas multipotent NSC expressed less of these marker proteins (28).
In summary, a careful evaluation of natural phenotypic variations within stem cell lineages and optimization of isolation, culture conditions and flow cytometric analysis needs to be performed, thereby avoiding the measurements of artifacts due to experimental conditions and misinterpretation of these data. The paradigm that so named neural genes are only expressed in cells of neuroectodermal origin with destination to differentiate into a neuron-specific phenotype also seems not be anymore valid, because such genes are also expressed by stem cells originating from a different germ layer, which naturally do not participate in neurogenesis. Variations of marker expression in the same stem cell lineage obtained from various tissues may indicate differences in their pluripotency. It is also possible that stem cell marker expression pattern and levels depend on respective cell cycle phases. Without any doubts, further investigations into mechanisms and functions of these stem cell epitopes will be necessary for understanding stem cell differentiation and using these cells successfully in cell therapy.