Pluripotency, defined as the ability of a cell to generate all cell types in the adult organism, is a transient feature of early embryonic development. Two distinct pluripotent cell types can be isolated from embryos and cultured in vitro1,2,3,4 — naive cells, called embryonic stem cells, and those primed for differentiation, epiblast stem cells. Furthermore, a defined cocktail of transcription factors, called reprogramming factors, can reinstate pluripotency when introduced into mature cells, producing induced pluripotent stem cells (iPSCs)5,6,7. In addition to known pluripotent cell types5,8, iPSC generation yields a spectrum of distinct cell types, hinting at the existence of uncharacterized pluripotent states. A collection of five manuscripts (two in this issue9,10 and three in Nature Communications11,12,13), now uncover and characterize an alternative pluripotent outcome of iPSC reprogramming: F-class cells (Fig. 1).
These five manuscripts are part of an international collaboration called Project Grandiose, in which the researchers set out to reanalyse the process of iPSC reprogramming from an unbiased perspective. They reasoned that, by extensively documenting the molecular and cellular transitions occurring at each stage of the process, they could provide both the first thorough roadmap for iPSC reprogramming, and an explanation for the emergence during reprogramming of undefined pluripotent cell types, which have been mostly overlooked by previous studies.
In the first paper, Tonge et al.9 (page 192) identify F-class cells — named because of their unusual, fuzzy-looking colony morphology — as a pluripotent cell type distinct from embryonic stem cells (ES cells) and epiblast stem cells. Maintenance of F-class cells depends on continuing high expression of reprogramming factors. In conventional reprogramming methods, the expression of introduced genes (transgenes) is silenced by factors that are expressed in the host cells once pluripotency is achieved, and thus F-class cells could not have been identified in those assays. The researchers' use of a host-factor-independent reprogramming method bypasses transgene silencing and thereby allows sustained high-level expression of reprogramming factors14.
Tonge and colleagues report that the fuzzy morphology of F-class cells arises from their low adhesiveness, which, along with their fast proliferation, makes these cells more amenable to large-scale production than ES cells. This is a desirable feature for cell-based therapies, which demand large quantities of specific cell types. For example, pancreatic β-cells, which store and release insulin, can be derived from pluripotent cells and might be used to treat people with diabetes15. However, F-class cells' dependence on transgenes could be problematic for their safe clinical application, because mutations arising from either improper transgene insertion into the genome or incomplete inactivation of reprogramming factors when the cells begin differentiation might ultimately lead to tumour formation.
One solution might be to stabilize the F-state independent of transgenes, using small molecules. This strategy has been successful for stabilizing naive-like human pluripotent stem cells16,17. Tonge et al. show that ES-like cells convert to the F-state following forced expression of reprogramming factors. Conversely, F-class cells can be converted to an ES-cell-like state using small molecules that inhibit the activity of a class of enzymes called histone deacetylases, which modulate gene expression by removing acetyl molecules from the histone proteins around which DNA is packaged. Such interconvertibility may lead to insights into how pluripotency is stabilized in distinct cellular contexts.
In the second paper, Hussein et al.10 (page 198) define the different molecular routes to pluripotency by performing the most detailed analysis of reprogramming so far. Among other findings, the authors uncover key determinants for the emergence of ES-cell-like or F-class states. Emergence of the F-class state relies on repression of genes that are expressed in ES cells. This is achieved through a molecular modification associated with gene repression — the attachment of three methyl molecules to an amino-acid residue, lysine 27, of histone H3 proteins. By contrast, the loss of the DNA methylation marks inherited from mature cells is necessary for cells to take on an ES-cell-like state, but some of these marks are retained in F-class cells.
The remaining three studies complement Hussein and colleagues' work by providing descriptive, in-depth analyses of the changes in molecular pathways en route to pluripotency, generating large data sets that are freely available at www.stemformatics.org. Lee et al.11 interrogate the epigenetic changes (those modifications to the genome that affect gene expression without altering DNA sequence) that occur during the transition to pluripotency. They conclude that DNA methylation has a crucial role in iPSC reprogramming and acts as an epigenetic switch between F-class and ES-cell-like states. Clancy and colleagues12 delineate the dynamic changes in small RNAs — post-transcriptional regulators of gene expression — during iPSC reprogramming, and find that a distinct group of microRNAs supports the F-class pluripotency program. Finally, Benevento et al.13 show that reorganization of protein expression occurs in two defined waves during cellular reprogramming. The authors show that patterns of protein expression differ between ES-cell-like and F-class states.
These five manuscripts mark the first steps towards understanding F-class pluripotency and thus towards making the most of their clinical potential. The molecular mechanisms underpinning the F-state warrant further investigation, as do the metabolic cues that contribute to sustaining F-class cells, because different pluripotent stem cells probably have distinct metabolic requirements18. Remaining questions include whether human F-class cells can be generated through cellular reprogramming, and if functional differentiated cells can be obtained from F-class cells.
In embracing the inherent artificiality of iPSC reprogramming, Project Grandiose has opened up the field to fresh avenues of research. This work shows that a third pluripotent state can be engineered in vitro, and it may be that there are other pluripotent endpoints of reprogramming (Fig. 1). Moreover, there may be other pluripotent states in the developing embryo. If there are, it would be interesting to determine whether such states could be captured and cultured in vitro. To investigate these avenues, an unbiased approach, such as that taken by Tonge et al., will probably prevail.
Looking ahead, customized stem cells designed for specific applications — such as large-scale expansion, or fast, synchronized differentiation — may soon become a reality. The existence of alternative pluripotent states adds another dimension to the potential of pluripotent stem cells in regenerative medicine. The results of Project Grandiose call for future work that catalogues myriad molecularly and functionally distinct pluripotent stem cells to harness their full potential.