If you’ve read our recent piece ‘Reprogramming Roadmap Reveals Fuzzy New Stem Cells‘ you’ll already know how “Project Grandiose” [1, 2], the brainchild of Andras Nagy, identified a new pluripotent state (the F-class stem cell). While the discovery of this new class is thrilling, the main purpose of this grand project was to uncover the molecular changes that underlie reprogramming of stem cells.
Therefore, accompanying the two ground breaking Nature studies previously discussed are three Nature Communications studies [3-5], which provide an in-depth analyses of the epigenetic (Lee et al.), protein expression (Benevento et al.), and small RNA changes (Clancy et al.) that occur during the reprogramming of mouse embryonic fibroblasts (MEFs) to two distinct pluripotent states; the embryonic stem cell (ESC) state and the Fuzzy, or F-class state. These analyses revealed diverse changes that occur at all levels during reprogramming.
DNA Methylation – The Main Barrier to Reprogramming?
In order to characterize the global dynamics of epigenetic modifications that occur during reprogramming, Lee et al. performed whole-genome bisulfite sequencing (MethylC-seq), chromatin immunoprecipitation sequencing (ChIP-seq) for H3K4me3/H3K27me3/H3K36me3 and RNA sequencing (RNA-Seq).
Using these techniques they uncovered some key findings:
- They identified 7,890 differentially methylated regions (DMRs) across all samples during the reprogramming process, and grouped these into three categories:
- DMR-1 – increased methylation levels after or during high-level reprogramming factor expression (includes genes related to development and cell differentiation).
- DMR-2 – represents differential methylation between the ESC and the F-class stem cell states.
- DMR-3 – low methylation levels in the ESC state with stable methylation persisting in the F-class state (includes multiple pluripotency genes).
- 98% of these DMRs contained H3K4me3 modified histones while 68% contained ESC-transcription factor binding sites (ESC-TFBSs).
- DMR low-level methylation, but not high-level methylation, allowed H3K4me3 and H3K27me3 gains.
- DMR-1 and -2 regions contained more Polycomb-repressive complex (PRC)-binding (SUZ12, EZH2 and RING1B) sites, while DMR-3 regions contained more ESC-TFBSs (Nanog, Oct4 and Klf4).
- Specific analysis of ESC-TFBSs found previously undetected focal demethylation and H3K4me3 gain linked to F-class cells.
- Broad demethylation in the ESC pluripotent state only.
From their findings the authors propose a model by which genes with CpG-poor promoters are controlled by DNA methylation (genes activated by demethylation and gain of H3K4me3), while for genes with CpG-rich promoters, low methylation levels allow histone modification mediated expression regulation. Overall, the authors strongly suggest that DNA methylation represents the major barrier to the reprogramming process.
Proteomics Plays its Part
While Lee et al. showed DNA methylation to be a major contributor to the reprogramming process, it isn’t the only change that was observed. Benevento et al. applied isobaric peptide labeling, strong cation exchange (SCX) chromatography and high-resolution tandem mass spectrometry (LC-MS/MS) analysis to probe dynamic proteomic changes during reprogramming.
Using this method the researchers demonstrated several interesting findings, including that 55% of the 4,454 quantified proteins changed more than two-fold in response to reprogramming factor expression, suggesting widespread proteome changes occur early in reprogramming.
These proteome changes were observed to occur in two waves, with the first happening 48 hours after activation of reprogramming factors and included activation of proteins involved in cell cycle control, cellular proliferation, metabolism and energy production as well as those involved in RNA processing, gene expression, and nucleolar proteins. The authors observed a strong enrichment for target genes of c-Myc, Kdm5b and Jarid2a, suggesting that these transcription factors are vitally important during this first wave.
As we’ve mentioned previously, the ‘Fuzzy’ or F-class of stem cells was named because of its fuzzy appearance under the microscope, which is caused by poor adhesion. The comparison between the proteome of F-class cells and ESCs was able to explain the cause of this appearance which was the due to the loss of epithelial and ESC-specific adhesion proteins in F-class cells.
In addition, comparison between F-class cells and ESCs highlighted the low abundance of several pluripotency markers and high levels of proteins associated with metabolism and cellular proliferation in F-class cells, regulated mainly by c-Myc. The observed loss of pluripotency-associated proteins was found to be mediated by DNA methylation and gain of H3K27me3.
Distinct miRNA Expression Distinguishes Stem Cell Types
In the last of these data rich studies, Clancy et al used next generation sequencing (NGS) to analyze small RNA expression at nucleotide resolution throughout the reprogramming process. This resulted in the discovery that while total miRNA levels are similar between cells, the expression of most individual miRNAs changes dramatically during reprogramming.
The researchers found that the miRNAs with the greatest change in expression during early reprogramming represent cell processes known to be vital for reprogramming, including cycle/proliferation, apoptosis, cell morphology, development/differentiation, signal transduction and gene regulation through chromatin modification, transcription and RNA processing.
This early miRNA expression pattern is controlled by rapid and dynamic changes to histone modifications, with H3K27me3 removal highly important, while DNA methylation was found to act later to set expression patterns present in the different pluripotent states. These later miRNA expression patterns are distinct and can be used to distinguish between ESCs and F-class cells.
The researchers uncovered 12 genomic loci that account for the expression of the ESC miRNAs (e.g. imprinted Dlk1-Dio3 locus, miR-302/367 locus, miR-290/295 locus) which are controlled by epigenetic/transcriptional regulation; DNA demethylation of these sites allows expression of these miRNAs late in reprogramming.
For F-class cells, 22 genomic loci were identified that account for expression of miRNAs (e.g. let-7b-5p, let-7-d-5p and let-7-c-5p/miR-99a-5p, miR-196a-5p and 181a-5p) and in contrast to ESC miRNAs, F-class miRNAs were found to be regulated at the post-transcriptional level.
Most excitingly, the group found that targets of F-class miRNAs could be grouped into an alternate pro-pluripotency regulatory network, which compensates for the lack of “classical” ESC-associated miRNAs.
Shedding light on the black box of reprogramming
Taken together these in-depth complimentary studies help to define the different stages of reprogramming and the different pluripotent states uncovered, and in particular they show that:
- DNA methylation is crucial to reprogramming and acts as an epigenetic switch between pluripotent states.
- Reorganization of protein expression occurs in two defined waves and protein expression patterns differ between pluripotent states.
- A distinct group of microRNAs support F-class pluripotency.
Further in-depth analysis of the huge amounts of freely available data (www.stemformatics.org) will hopefully shed yet more light on to what some call the “black-box” of reprogramming, allowing for greater control over this intricate process.
- Tonge PD, Corso AJ, Monetti C, et al. Divergent reprogramming routes lead to alternative stem-cell states. Nature 2014;516:192-197.
- Hussein SM, Puri MC, Tonge PD, et al. Genome-wide characterization of the routes to pluripotency. Nature 2014;516:198-206.
- Lee DS, Shin JY, Tonge PD, et al. An epigenomic roadmap to induced pluripotency reveals DNA methylation as a reprogramming modulator. Nat Commun 2014;5:5619.
- Clancy JL, Patel HR, Hussein SM, et al. Small RNA changes en route to distinct cellular states of induced pluripotency. Nat Commun 2014;5:5522.
- Benevento M, Tonge PD, Puri MC, et al. Proteome adaptation in cell reprogramming proceeds via distinct transcriptional networks. Nat Commun 2014;5:5613.