Umbilical Cord Versus Bone Marrow-Derived Mesenchymal Stromal Cells

S ince the placenta is a postnatal tissue and discarded as medical waste, harvesting stem cells from this organ represents a noninvasive and ethically conductive procedure. Perinatal stem cells isolated from amnion, chorion, umbilical cord, and cord blood are increasingly viewed as reliable sources of mesenchymal stromal cells (MSCs) alternative to bone marrow-derived ones (BM-MSCs), which are currently the most commonly used in clinical applications [1 –5].

Perinatal stem cells are a bridge between embryonic stem cells (ESCs) and adult stem cells (such as BM-MSCs). They share many characteristics of both cells [1,6]. Considering the structural complexity of the term “placenta,” we have focused our attention on umbilical cord stem cells (UCSCs). Like BM-MSCs, UCSCs possess the fibroblast-like morphology, nonhematopoietic cell surface phenotypes, low immunogenicity, and multipotent differentiation ability [3,5 –8]. However, there are many differences between UCSCs and BM-MSCs. First, without the ethical cloud, stem cells are easily harvested from the UC, and the cells have a higher frequency of proliferation and colony-forming units (CFU) formation than BM-MSCs [9,10]; senescent BM-MSCs were recorded earlier than UCSCs during subculturing [11]. Second, beyond MSCs markers, several ESCs markers were present in UCSCs, but not to the same extent in BM-MSCs. UCSCs expressed TRA-1-60, TRA-1-81, SSEA-1, SSEA-3, SSEA-4, Oct-4, alkaline phosphatase (ALP), DNMT3B, and GABRB3 [12,13], but showed low expression levels of genes associated with teratomas formation [14]. These expression patterns contribute to justify the observed multipotency of UCSCs at the molecular level, which may readily cross germ layer boundaries in the differentiation process [15,16]. Third, UCSCs possessed the differentiation to adipogenic, osteogenic, and chondrogenic lineages [17,18]. When incubated in an adipogenic medium and stained with Oil Red O, UCSCs readily differentiated into multilocular adipocyte-like cells [10,19 –21], while a unilocular lipid droplet was generally seen in the mature adipocytes or BM-MSCs after induction. Hence, the adipogenic capacity of UCSCs was lower than that of BM-MSCs. Also, the lower osteogenesis ability of UCSCs was documented, suggesting that BM-MSC comparatively possess a better osteogenic potential [22,23], while UCSCs seemed to be more primitive because they share common genes with ESCs [23]. Interestingly, UCSCs were shown to be nontumorigenic, which suggested that UCSCs are safe for potential clinical application [24]. Although there are many in vivo studies to determine the therapeutic potential, only 2 illustrated the transplantation of UCSCs in human clinical application with safe and beneficial results [25,26]. It is clear that many more studies are essential and necessary. Long-term follow-up is absolutely needed to validate the feasibility of UCSC-based therapy. In summary, according to the minimal criteria of the International Society for Cellular Therapy (ISCT) [27], UCSCs generally, but not strictly, fulfill the definition of MSCs, as a primitive stem cell population increasingly used for extensive preclinical tests and clinical applications.

In a recent article published in this journal by Bosch et al. [28], UC-MSCs were isolated and the morphology was fibroblastic. Although UC-MSCs exhibited a similar expression profile of cell surface proteins compared with BM-MSCs, the cells failed to differentiate into adipo-, osteo-, and chondrogenic lineages [28]. The authors therefore concluded that this cell population should not be regarded as MSCs. As mentioned above, it is known that UCSCs possess a lower adipogenic and osteogenic differentiation ability with respect to BM-MSCs, but in this study no typical mesenchymal differentiation potential was found. In our opinion, the following reasons should be considered to provide a better interpretation of the results, in light of the published articles supporting the UC-MSCs differentiation capacity.

First, the case for adipogenic differentiation showed that in the authors' hands the extent of differentiation of BM-MSCs was up to 6.54% (assessed by the measurement of areas of lipid vacuoles). This is a strikingly low efficiency, obtained for a cell type that is supposed to constitute the reference for all other MSCs. Therefore, the possibility that the induction medium failed to deliver the expected lipid vacuole yield cannot be excluded. This also repeated in lower adipogenic efficiency MSCs such as UC-MSCs, could perhaps give rise to the absence of differentiation claimed by the authors. In contrast, clear demonstration of adipogenic differentiation of UC-MSCs has been previously demonstrated, along with the acquisition of expression of 2 key genes of the adipocyte phenotype such as adiponectin and leptin [17].

Second, the authors claimed that UC-MSCs were not able to differentiate into chondrocytes for the chondrogenic differentiation experiments. Interestingly, pellets from UC-MSCs were Alcian Blue-positive as those derived from the other analyzed stem cells, but the authors based their conclusion solely on the expression of SOX-9. Since the authors correctly stated that markers expression was needed to confirm staining, we add that only 1 marker may not be sufficient to this aim. In fact, to assess chondrogenic differentiation, collagen type II, a cartilage oligomeric matrix protein, aggrecan and fibromodulin expression are widely used together with SOX-9 to assess proper differentiation [18]. Additionally, UC-MSCs were demonstrated to differentiate into chondrocytes both via basic and multistep protocols even with better results when compared to BM-MSCs [18,29,30]. The reliability of SOX-9 in the experimental setting showed by the authors appears to be limited, since in most lines (including BM-MSCs) the induction protocol failed to upregulate the expression of SOX-9 from baseline levels of undifferentiated cells. Therefore, we must again ask if SOX-9 is not responsible for the morphological evidence of chondrogenic differentiation provided by Alcian blue and Safranin O stainings, then what is responsible? Maybe an immunohistochemistry (IHC) for collagen type II could clarify this point.

Third, the authors again claimed the lack of proper differentiation of UC-MSCs regarding osteogenic differentiation, despite positive (but weak) Alizarin Red S staining. We would like to note that while the tested genes were mostly negative for UC-MSCs, the same were not clearly upregulated in BM-MSCs, despite the intense staining observed. Therefore, morphological data were not supported by an expression analysis. We and others just demonstrated that UC-MSCs differentiated toward osteoblasts and did express de novo both osteonectin and periostin, which were absent in undifferentiated cells. Furthermore, they displayed the ALP activity [17].

Fourth, the authors have demonstrated that different isolation protocols did not result in any difference regarding the generation efficiency and biological features of UC-MSCs. Nevertheless, MSCs were successfully harvested and the differentiation potential capacity was illustrated by histological staining using a similar method [20]. As the dominant component in the UC matrix, Wharton's Jelly (WJ) is an abundant source of MSCs and the differentiation capacity has been clearly reported [3] and extensively underlined. According to the isolation method mentioned in this study, the harvested cells may also come from cord subregions other than WJ, a point that probably should be discussed. Indeed, there currently is a lack of comprehensive comparisons between cells harvested from the different components of the UC to understand if they represent different populations, as the literature suggests [31,32]. A question that should be raised is whether the cells obtained from Bosch et al. [28] represent a new cell population and different from WJ-MSCs.

Lastly, the culture medium may play a significant role in this difference. It was suggested that different basal media affect the in vitro expansion, adipogenesis, and osteogenesis of WJ-MSCs [13]. Therefore, whether the choice of the basal media is associated with the absence of multiple differentiation potential in this study remains to be determined. The authors used an uncommon medium to subculture UC-MSCs. This is strikingly evident when compared with Dulbecco's modified Eagle medium (DMEM), which they used for the other cells investigated and which the vast majority of authors worldwide use for UC-MSCs.

The discrepancies regarding the differentiation potential observed and discussed herein should bring us to consider the major limitation of the in vitro culture procedures. No matter what the source of MSCs being evaluated, differentiation is commonly induced by cocktails containing a selection of powerful, stimulating factors; different protocols and laboratories could give rise to discordant data of therefore difficult interpretation. Furthermore, in vitro cultures represent neither the physiological in vivo condition nor the injured tissue microenvironment, the latter being even more complex to reproduce. Therefore, we could acknowledge from the Bosch et al. study [28] that the discrepancies lead us to consider the limitation of the in vitro studies and even more so the crucial role of in vivo multilineage differentiation studies as the definitive/ultimate assay to define the multipotency of MSCs independently of their source.

Perinatal stem cells have generated noteworthy interests as promising graft cells for tissue regeneration. According to minimal criteria suggested by ISCT, the majority of cell populations isolated from the term placenta are MSC-like cells, but the biological characteristics vary between cells isolated from different regions. Therefore, many investigators, including these authors, performed research to determine the differences. The expression patterns of HOX genes, CD56 and CD146 could broaden our understanding of the difference between perinatal stem cells and BM-MSCs. Nevertheless, it is still difficult to perform a direct comparison of various study outcomes when different isolation, culturing, and differentiation procedures are used. Thus, it is increasingly necessary to standardize the preparation and differentiation procedures for promising perinatal stem cells to better define and properly characterize these cells and accelerate the advances in the use of these cells for preclinical studies and in the development of clinical applications.