Dissecting direct reprogramming through integrative genomic analysis

Abstract

Somatic cells can be reprogrammed to a pluripotent state through the ectopic expression of defined transcription factors. Understanding the mechanism and kinetics of this transformation may shed light on the nature of developmental potency and suggest strategies with improved efficiency or safety. Here we report an integrative genomic analysis of reprogramming of mouse fibroblasts and B lymphocytes. Lineage-committed cells show a complex response to the ectopic expression involving induction of genes downstream of individual reprogramming factors. Fully reprogrammed cells show gene expression and epigenetic states that are highly similar to embryonic stem cells. In contrast, stable partially reprogrammed cell lines show reactivation of a distinctive subset of stem-cell-related genes, incomplete repression of lineage-specifying transcription factors, and DNA hypermethylation at pluripotency-related loci. These observations suggest that some cells may become trapped in partially reprogrammed states owing to incomplete repression of transcription factors, and that DNA de-methylation is an inefficient step in the transition to pluripotency. We demonstrate that RNA inhibition of transcription factors can facilitate reprogramming, and that treatment with DNA methyltransferase inhibitors can improve the overall efficiency of the reprogramming process.

Figure 1. Gene expression profiling

Relative expression levels across differentiated, partially reprogrammed and pluripotent cell populations. The dendrogram was generated by complete linkage hierarchical clustering using Pearson correlation on all measured genes. Only genes with at least twofold difference between any pair of samples from different classes are shown in the heat map. Red, white and blue indicate higher, identical and lower relative expression, respectively. ES cells, embryonic stem cells.

Figure 2. Chromatin state maps

a, Loss of H3K4me3 correlates with inactivation of MEF-specific low-CpG promoters (LCPs), such as that of Postn (periostin), during reprogramming. b, The transcription factor Zeb2 is marked by H3K4me3 and expressed in MEFs, but gains H3K27me3 and is silenced in partially and fully reprogrammed cells. c, The mesoderm/neural-crest transcription factor Sox9 is marked by H3K4me3 only and remains active in MCV6. d, The endodermal transcription factor Gata6 inappropriately lost H3K27me3 and is activated in MCV6 cells. e, The autocrine growth factor Fgf4 loses H3K27me3, gains H3K4me3 and becomes highly expressed in both partially and fully reprogrammed cells. f, The pluripotency gene Nanog gains H3K4me3 and is active only in iPS cells. g, The germline-specific gene Ddx4 gains H3K4me3 and H3K27me3 in iPS cells only, and remains poised for activation in germ cells. h, Chromatin states for high-CpG promoters (HCPs) in MEFs and reprogrammed cells, conditional on their state in embryonic stem cells. i, Fraction of genes with HCPs expressed in embryonic stem cells, but not wild-type MEFs, that have been re-activated in cells at various stages of reprogramming, conditional on their chromatin state inMEFs. Most HCPs markedbyH3K27me3 onlyorby neither mark are not re-actived in partially reprogrammed cells. d4, day 4.

Figure 3. DNA methylation analysis

Bisulphite sequencing of promoters or enhancers with Oct4/Sox2 binding sites near pluripotency-related and germ-cell-specific (Stella and Cyct) genes, as catalogued in ref. . Empty squares indicate unmethylated and filled squares methylated CpG dinucleotides. Most assayed sites are hypermethylated in differentiated and partially reprogrammed cells. Sox2 is enriched with H3K27me3 in non-pluripotent cells and accordingly hypomethylated in all cell types. Triangles show sites used for COBRA analysis (see text).

Figure 4. Inhibition of Dnmt1 accelerates the transition to pluripotency

a, MCV8 (sorted by FACS using a SSEA1-specific antibody) and BIV1 (−Dox) were either exposed to AZA for 48h (green) or kept in regular embryonic stem cell medium (grey). The number of Oct4–GFP-positive cells was analysed over multiple passages (P) by FACS. b, Untreated MCV8 control cells from passage 5 were subsequently subjected to AZA treatment for 48 h (+) or 120h (++), and resulting Oct4–GFP-positive cells were counted after one passage. P6/P1, total passage 6/passage 1 after AZA treatment. c, AZA treatment does not influence retroviral expression levels. d, AZA treatment has no influence on lentiviral expression in uninduced or induced BIV1 cells. e, Pluripotencyof all AZA-treated lines and MCV8.1 was demonstrated by teratoma formation. ECT, ectoderm; END, endoderm; MES, mesoderm. f, Overall efficiency of AZA treatment. Nanog–GFP MEFs were plated on 6-well plates (4 wells per time point with Dox, and 2 wells without). Cells were treated with AZA during one of the indicated intervals. On day 14, colony formation was analysed by fluorescence microscopy (representative panels are shown). g, Number of alkaline-phosphatase-positive, embryonic-stem-cell-like colonies obtained from each treatment. AZA treatment during days 8–10 resulted in a ~4-fold increase in efficiency over untreated controls. For a, c, d and g, error bars show standard deviations (n = 2, 2, 2 and 4, respectively).

Figure 5. Transcription factor knockdown facilitates reprogramming

MCV6 cells were plated onto 24-well dishes and transfected with siRNAs targeting expressed (Pax7, Pax3, Gata6, Sox9) or non-expressed (Zic1, Meox2) transcription factors. One plate was kept in embryonic stem cell medium and the second was exposed to AZA for 48h. Two independent siRNA sequences were used for duplicate experiments (red and green). FACS analysis was performed 48 h after AZA treatment (96 h after transfection) without passaging. The transfection efficiency was estimated as ~20% using Cy3-coupled GADPH control siRNA.