Advancing Pluripotent Stem Cell Culture: It Is a Matter of Setting the Standard

T he techniques for generation of human pluripotent stem cells (hPSCs) are nowadays procedures that most laboratories master, and the reagents needed can be obtained from commercial sources relying on standardized and quality controlled processes [1,2]. Thus, creating an hPSC line no longer represents a major obstacle to enter into this research field. In addition, if there is a preference to work on already established cell lines, there is a wide access to various cell lines available from academic and industrial institutions. Furthermore, a range of specific culture media and support matrices for propagation and expansion of the undifferentiated cells are also available [3]. Thus, by combining feeder-free culturing systems and taking advantage of enzymatic or nonenzymatic passaging, researchers can now grow hPSCs with relative ease [4]. Over the years, much effort has been dedicated to the development and agreement on a set of standard tests that should be used to characterize and classify hPSCs [5]. These tests were initially defined using human embryonic stem cells (hESCs) and have later been expanded to also include induced PSCs. Such tests include gene and protein expression analysis of pluripotency markers (e.g., Oct3/4, Nanog, Sox2, and Lin28), alkaline phosphate activity, analysis of genomic integrity, and evaluation of differentiation capacity in vitro and in vivo (teratoma analysis). Together, the tests have proven useful to be able to demonstrate the pluripotent phenotype of the undifferentiated cells at the time of their procurement. However, it is important to emphasize that a continuous monitoring of the hPSCs is needed during extended in vitro culture since genomic aberrations or other types of culture adaptations can create a drift in the cultures over time [6]. Spending precious time and resources to implement and uphold rigorous quality routines in the hPSC laboratories is sometimes viewed as less exciting and scientifically less rewarding and is therefore at risk to be down-prioritized in the day-to-day activities. Nevertheless, for all applications of hPSCs, it is obvious that a standardized source of high-quality hPSCs is critical. Self-renewing cultures of the hPSCs most often also contain fractions of differentiated cells that may affect the pluripotent state and the down-stream performance of the cultures. Unfortunately, the vast majority of published articles using hPSCs contain very limited data, which describe the detailed properties of the starting cultures used in the investigations. This sometimes makes it quite difficult for a reader to interpret and assess the findings and conclusions made in the articles. Specifically, reproducing published directed differentiation protocols is a challenging exercise, since even subtle differences between the starting hPSC cultures between laboratories may have a substantial impact on how the cells respond to the differentiation signals [7,8].

Variation between individual hPSCs is often attributed to their having different genetic backgrounds, but there are many other parameters that can influence the behavior and characteristics of the cells. Typically, investigators point to the fact that there are variations in the culture systems used, media formulations, and epigenetic differences between the cells lines. Nevertheless, there is seldom experimental evidence to support that these factors are indeed the explanation for the discrepancies in the data obtained. The standard set of markers and functional tests that are used to characterize the undifferentiated cells are certainly informative and useful for a basic evaluation of the hPSC cultures [5]. However, with regard to being able to predict the hPSC's capacity to differentiate, either spontaneously or in a guided manner to one or a few specific cell types using a directed differentiation approach, a more detailed analysis of the hPSC culture is required.

Recent studies have illustrated that, although the hPSCs display a pluripotent gene profile, the cells may be primed to respond to differentiation cues in a substantially differential manner. In this regard, Blauwkamp et al. elegantly demonstrated that cultures of undifferentiated hESCs were composed of subfractions of cells expressing high or low Wnt activity [8]. Interestingly, the hESC populations displayed comparable expression levels of the typical pluripotency markers OCT4, SSEA-4, and TRA-1-60, indicating that these marker expressions were relatively blunt and imprecise in this setting. The authors separated Wnt^high from Wnt^low hESCs and showed differences between these populations with regard to clonogenic potential, epigenetic status, and differentiation propensity. The Wnt^high hESCs seemed to be primed to differentiate toward mesoderm/endoderm, and expressed higher levels of early differentiation markers, such as Brachyury, Goosecoid, Sox 17, and CXCR4. On the other hand, the Wnt^low hESCs expressed higher levels of Pax6, a neuroectodermal marker. The gene expression profiles correlated also with lineage-specific differentiation, and the Wnt^high hESCs differentiated more efficiently to mesoderm and endoderm derivatives and the Wnt^low hESCs were more prone to generate ectoderm. Nevertheless, the Wnt^high and Wnt^low hESCs generated teratomas when injected into immunocompromized mice demonstrating their maintained pluripotency despite the different levels of Wnt activity. Interestingly, reculturing pure Wnt^high hESCs resulted in a heterogeneous population containing both the Wnt^high and Wnt^low hESCs within 1 passage, indicating a high degree of plasticity of the state of the Wnt activity levels in the cells. This study is an excellent example that illustrates that there is a great need to investigate the heterogeneous nature of hPSC cultures in much more detail. Taken together, these data may help to begin to explain why different laboratories can obtain widely different results when employing the same differentiation protocol on seemingly similar and pluripotent hPSCs.

The complexity of the molecular network regulating pluripotency was also recently illustrated by Wang and colleagues who investigated in detail the roles of OCT4, SOX2, and NANOG in lineage specification of hESCs [7]. Of particular note, and of relevance for the present discussion, the authors demonstrated that the level of OCT4 expression significantly affected the response of the cells toward BMP4 treatment. Specifically, high levels of OCT4 in combination with BMP4 resulted in mesendoderm differentiation, while low levels of OCT4 in combination with BMP4 resulted in extraembryonic lineage differentiation. As indicated above, standard cultures of hPSCs are not homogenous and cells displaying different levels of OCT4 are most likely present to various degrees in a dynamic manner during extended in vitro culturing and passaging. This is important to consider, especially when subjecting the cells to various differentiation regimes. Accordingly, specific subfractions of cells may respond substantially different to the induction signals, and thus, offer an explanation for the variability in results obtained when subjecting the cells to directed differentiation protocols.