Mbd3 and deterministic reprogramming

Rais et al. (Nature 502: 65-70) reported that suppressing formation of the Nucleosome Remodelling and Deacetylase (NuRD) complex by depleting its structural component Mbd3 promotes near 100% induction of somatic cells reprogrammed to pluripotency. These results have not been independently reproduced and conflicting data were obtained in a recent study by our groups (Cell Stem Cell 15: 102-10). Here we investigate the source of these disparities through analysis of the genomic data accompanying Rais et al. In contrast to the conditions described in the study, Mbd3 remained expressed at high levels in both heterozygous and incompletely ablated knockout cell lines used to assess pluripotency induction in an Mbd3-depleted state, calling into question the involvement of Mbd3 in the reported improvements to reprogramming efficiency. Mismatched reporter constructs were found to be applied to Oct4-GFP experiment and control cells, where the primary enhancer element regulating Oct4 expression is present only in the test group where apparent enhancements in reprogramming kinetics were observed. This setup led to inappropriate comparison of test cells harboring a promiscuous Oct4-GFP reporter and control cells transformed with a much more stringent version. Such conditions are expected to incur substantial differences in GFP output and misleading interpretation of fluorescence data. We relay our findings to inform the community of these issues and to express our concerns regarding the conclusions presented in Rais et al.