Comment on Wang et al, page 4598
Human embryonic stem cells are a valuable resource for research and cell replacement therapy but are notoriously cumbersome to culture. Bhatia and colleagues show that an increased dose of basic fibroblast growth factor eliminates the need for feeder layer coculture.
Human embryonic stem cells (hESCs) represent an important tool for research and a potential resource for cell-replacement therapies, but they are difficult to cultivate. Typically, hESCs are cocultured with a supportive feeder cell layer of murine fibroblasts that provides factors that maintain self-renewal divisions and inhibit the otherwise spontaneous tendency of the cells to differentiate. The cumbersome and finicky conditions for growing hESCs have hindered wider exploitation of these cells by the research community, and significant advances in cell manufacturing and bioprocess engineering are needed before these cells are ready for clinical applications. In this issue of Blood, Bhatia and colleagues have taken an important first step toward simplifying hESC culture by showing that high doses of basic fibroblast growth factor (bFGF) are adequate to maintain hESCs under feeder-free and serum-free growth conditions.
Mouse and human ESCs share many important properties, yet for years scientists have been perplexed by the different cytokine requirements for propagating the 2 cell types. Mouse ESCs can be grown in the absence of feeder cells if the media is supplemented with leukemia inhibitory factor (LIF), which maintains symmetrical self-renewal divisions and acts as a potent inhibitor of ESC differentiation by activating signal transducers and activators of transcription 3 (STAT3) signal transduction. In contrast, LIF does not prevent differentiation of hESCs, despite the fact that the LIF receptor and the STAT3 signaling pathway are intact.1,2 Human ESCs can be propagated in the absence of feeders as long as they are plated on a preformed extracellular matrix (provided by matrigel [BD Biosciences] or laminin), provided with media conditioned by growth on feeder fibroblasts, and supplemented with bFGF (typically at doses <10 ng/mL3 ). The advantages of feeder-free culture have prompted enormous interest in identifying the “LIF-like” components in fibroblast conditioned medium (CM). Surprisingly, Bhatia and colleagues have now shown that simply increasing the dose of bFGF added to a simple serum-free culture media allows researchers to maintain hESCs in the absence of feeders or CM. Cells cultured in this manner retain pluripotency and can be differentiated into hemogenic-endothelial precursors, thereby facilitating in vitro studies of human hematopoietic development. Furthermore, Bhatia et al show that incubation of CM from feeders with a neutralizing antibody against bFGF abrogates the capacity of CM to support hESCs, suggesting that bFGF is an essential factor produced by feeder cells, although direct quantification of the amount of bFGF in CM would have helped determine whether all of the critical activities of CM can be accounted for by bFGF. Recently, the observation that high-dose bFGF (40 ng/mL) can sustain hESCs has been independently reported by 2 other groups, thereby corroborating this important insight.4,5
Although bFGF appears to function as an LIF equivalent in hESC culture, there are probably other active constituents of CM. Thomson's group5 report that feeders may serve in part to neutralize or antagonize illdefined inducers of differentiation that are contributed by serum or various commercially available cocktails of “serum-replacement.” Other elements currently employed in hESC culture also remain a mystery, especially the complex and variable composition of matrigel (a basement matrix prepared from a murine sarcoma tumor line) and the proprietary constituents of various serum replacements. The derivation of new hESC lines free of contamination with animal products, to minimize risks associated with animal pathogen transmission, is a worthy goal for hESC research. Conditions for growth of hESCs in entirely chemically defined media have yet to be discovered, but identification of bFGF as an essential component is an important first step. ▪
