Limits...
Aberrant DNA methylation reprogramming during induced pluripotent stem cell generation is dependent on the choice of reprogramming factors.

Planello AC, Ji J, Sharma V, Singhania R, Mbabaali F, Müller F, Alfaro JA, Bock C, De Carvalho DD, Batada NN - Cell Regen (Lond) (2014)

Bottom Line: Strikingly, not only were the genomic locations of the aberrations different but also their types: reprogramming with Yamanaka factors mainly resulted in failure to demethylate CpGs, whereas reprogramming with Thomson factors mainly resulted in failure to methylate CpGs.Our study thus reveals that the choice of reprogramming factors influences the amount, location, and class of DNA methylation aberrations in iPSCs.These findings may provide clues into how to produce human iPSCs with fewer DNA methylation abnormalities.

View Article: PubMed Central - PubMed

Affiliation: Campbell Family Cancer Research Institute, Ontario Cancer Institute, Princess Margaret Cancer Centre, University Health Network, Toronto, ON M5G 2 M9 Canada ; Department of Morphology, Piracicaba Dental School, University of Campinas, Piracicaba, SP Brazil.

ABSTRACT
The conversion of somatic cells into pluripotent stem cells via overexpression of reprogramming factors involves epigenetic remodeling. DNA methylation at a significant proportion of CpG sites in induced pluripotent stem cells (iPSCs) differs from that of embryonic stem cells (ESCs). Whether different sets of reprogramming factors influence the type and extent of aberrant DNA methylation in iPSCs differently remains unknown. In order to help resolve this critical question, we generated human iPSCs from a common fibroblast cell source using either the Yamanaka factors (OCT4, SOX2, KLF4 and cMYC) or the Thomson factors (OCT4, SOX2, NANOG and LIN28), and determined their genome-wide DNA methylation profiles. In addition to shared DNA methylation aberrations present in all our iPSCs, we identified Yamanaka-iPSC (Y-iPSC)-specific and Thomson-iPSC (T-iPSC)-specific recurrent aberrations. Strikingly, not only were the genomic locations of the aberrations different but also their types: reprogramming with Yamanaka factors mainly resulted in failure to demethylate CpGs, whereas reprogramming with Thomson factors mainly resulted in failure to methylate CpGs. Differences in the level of transcripts encoding DNMT3b and TET3 between Y-iPSCs and T-iPSCs may contribute partially to the distinct types of aberrations. Finally, de novo aberrantly methylated genes in Y-iPSCs were enriched for NANOG targets that are also aberrantly methylated in some cancers. Our study thus reveals that the choice of reprogramming factors influences the amount, location, and class of DNA methylation aberrations in iPSCs. These findings may provide clues into how to produce human iPSCs with fewer DNA methylation abnormalities.

No MeSH data available.


Related in: MedlinePlus

Experimental design and comparison of DNA methylation profiles of iPSCs and ESCs. A) Scheme showing the experimental design. Cells were treated identically and 25 days post-infection, colonies were picked randomly and maintained in culture for 6 passages. We profiled DNA methylation during early passage iPSCs to reduce the confounding contribution of negative selection during culture conditions [32], which may mask the differences between these 2 reprogramming factor combinations. B) Pair-wise plot shows that all iPSC methylomes are similar to ESC methylomes and significantly different from fibroblast methylomes. The raw methylation data was normalized and background-subtracted using the Illumina Genome Studio software. Intensity values were converted to beta-values where the value of 0 represents unmethylated and the value of 1 represents fully methylated. Fib stands for fibroblasts. C) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, ESCs, Y-iPSCs and T-iPSCs show that their overall levels in iPSCs are largely similar to those in ESCs. Hypermethylated CpGs are defined as those with values > 0.7 and hypomethylated CpGs are defined as those with values < 0.3. Hyper and Hypo stand for hypermethylated and hypomethylated, respectively. D) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, Y-iPSCs and T-iPSCs that are significantly different from ESCs (defined as those with difference in methylation > =0.2) show that only a small proportion of the CpGs in iPSCs are different from those in ESCs. E) Boxplot made as in Figure 1D but restricted to only those CpGs that change their methylation state from fibroblasts to ESCs. CpGs hypomethylated in fibroblasts but hypermethylated in ESCs (green) are more likely to be significantly different in iPSCs than CpGs that are hypermethylated in fibroblasts and hypomethylated in ESCs (yellow).
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4230737&req=5

Fig1: Experimental design and comparison of DNA methylation profiles of iPSCs and ESCs. A) Scheme showing the experimental design. Cells were treated identically and 25 days post-infection, colonies were picked randomly and maintained in culture for 6 passages. We profiled DNA methylation during early passage iPSCs to reduce the confounding contribution of negative selection during culture conditions [32], which may mask the differences between these 2 reprogramming factor combinations. B) Pair-wise plot shows that all iPSC methylomes are similar to ESC methylomes and significantly different from fibroblast methylomes. The raw methylation data was normalized and background-subtracted using the Illumina Genome Studio software. Intensity values were converted to beta-values where the value of 0 represents unmethylated and the value of 1 represents fully methylated. Fib stands for fibroblasts. C) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, ESCs, Y-iPSCs and T-iPSCs show that their overall levels in iPSCs are largely similar to those in ESCs. Hypermethylated CpGs are defined as those with values > 0.7 and hypomethylated CpGs are defined as those with values < 0.3. Hyper and Hypo stand for hypermethylated and hypomethylated, respectively. D) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, Y-iPSCs and T-iPSCs that are significantly different from ESCs (defined as those with difference in methylation > =0.2) show that only a small proportion of the CpGs in iPSCs are different from those in ESCs. E) Boxplot made as in Figure 1D but restricted to only those CpGs that change their methylation state from fibroblasts to ESCs. CpGs hypomethylated in fibroblasts but hypermethylated in ESCs (green) are more likely to be significantly different in iPSCs than CpGs that are hypermethylated in fibroblasts and hypomethylated in ESCs (yellow).

Mentions: iPSCs were derived from human neonatal foreskin fibroblasts via retroviral infection of reprogramming factors [16, 17]. A common batch of human neonatal foreskin fibroblasts was split into 3 groups (Figure 1A). The first set was directly processed for DNA methylation analysis. The second set was infected with retroviruses encoding the Yamanaka factors, i.e. OCT4, SOX2, KLF4 and cMYC reprogramming factors. The third set was infected with retroviruses encoding the Thomson factors, i.e. OCT4, SOX2, NANOG and LIN28 reprogramming factors. We selected 9 Y-iPSCs and 6 T-iPSCs that had characteristic ESC-like morphology, express several pluripotency genes and can be stably maintained in culture (Figure 1A; Additional file 1: Figure S1). Genomic DNA from each of the iPSCs was subjected to DNA methylation profiling using the Illumina HumanMethylation450 platform. We augmented our data set with published Illumina HumanMethylation450 methylation data for several male ESCs (SIVF002, SIVF025, SIVF043, SIVF044, SIVF050) [14]. To validate the DNA methylation data obtained by the Illumina HumanMethylation450 platform we performed high throughput sequencing based DNA methylation assessment called reduced representation bisulfite sequencing (RRBS) in one of the Y-iPSCs and one of the T-iPSCs. The overall methylation data from Illumina HumanMethylation450 and RRBS data were concordant (with an average correlation coefficient of 0.8) (Additional file 1: Figure S2). These results are very similar to the Pearson correlation between RRBS and Illumina’s DNA methylation arrays obtained in a different study using the same cutoffs [19] and argue in favor of the reliability of the data obtained in our study.Figure 1


Aberrant DNA methylation reprogramming during induced pluripotent stem cell generation is dependent on the choice of reprogramming factors.

Planello AC, Ji J, Sharma V, Singhania R, Mbabaali F, Müller F, Alfaro JA, Bock C, De Carvalho DD, Batada NN - Cell Regen (Lond) (2014)

Experimental design and comparison of DNA methylation profiles of iPSCs and ESCs. A) Scheme showing the experimental design. Cells were treated identically and 25 days post-infection, colonies were picked randomly and maintained in culture for 6 passages. We profiled DNA methylation during early passage iPSCs to reduce the confounding contribution of negative selection during culture conditions [32], which may mask the differences between these 2 reprogramming factor combinations. B) Pair-wise plot shows that all iPSC methylomes are similar to ESC methylomes and significantly different from fibroblast methylomes. The raw methylation data was normalized and background-subtracted using the Illumina Genome Studio software. Intensity values were converted to beta-values where the value of 0 represents unmethylated and the value of 1 represents fully methylated. Fib stands for fibroblasts. C) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, ESCs, Y-iPSCs and T-iPSCs show that their overall levels in iPSCs are largely similar to those in ESCs. Hypermethylated CpGs are defined as those with values > 0.7 and hypomethylated CpGs are defined as those with values < 0.3. Hyper and Hypo stand for hypermethylated and hypomethylated, respectively. D) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, Y-iPSCs and T-iPSCs that are significantly different from ESCs (defined as those with difference in methylation > =0.2) show that only a small proportion of the CpGs in iPSCs are different from those in ESCs. E) Boxplot made as in Figure 1D but restricted to only those CpGs that change their methylation state from fibroblasts to ESCs. CpGs hypomethylated in fibroblasts but hypermethylated in ESCs (green) are more likely to be significantly different in iPSCs than CpGs that are hypermethylated in fibroblasts and hypomethylated in ESCs (yellow).
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4230737&req=5

Fig1: Experimental design and comparison of DNA methylation profiles of iPSCs and ESCs. A) Scheme showing the experimental design. Cells were treated identically and 25 days post-infection, colonies were picked randomly and maintained in culture for 6 passages. We profiled DNA methylation during early passage iPSCs to reduce the confounding contribution of negative selection during culture conditions [32], which may mask the differences between these 2 reprogramming factor combinations. B) Pair-wise plot shows that all iPSC methylomes are similar to ESC methylomes and significantly different from fibroblast methylomes. The raw methylation data was normalized and background-subtracted using the Illumina Genome Studio software. Intensity values were converted to beta-values where the value of 0 represents unmethylated and the value of 1 represents fully methylated. Fib stands for fibroblasts. C) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, ESCs, Y-iPSCs and T-iPSCs show that their overall levels in iPSCs are largely similar to those in ESCs. Hypermethylated CpGs are defined as those with values > 0.7 and hypomethylated CpGs are defined as those with values < 0.3. Hyper and Hypo stand for hypermethylated and hypomethylated, respectively. D) Boxplot of the percentages of highly methylated (red) and lowly methylated (blue) CpGs in fibroblasts, Y-iPSCs and T-iPSCs that are significantly different from ESCs (defined as those with difference in methylation > =0.2) show that only a small proportion of the CpGs in iPSCs are different from those in ESCs. E) Boxplot made as in Figure 1D but restricted to only those CpGs that change their methylation state from fibroblasts to ESCs. CpGs hypomethylated in fibroblasts but hypermethylated in ESCs (green) are more likely to be significantly different in iPSCs than CpGs that are hypermethylated in fibroblasts and hypomethylated in ESCs (yellow).
Mentions: iPSCs were derived from human neonatal foreskin fibroblasts via retroviral infection of reprogramming factors [16, 17]. A common batch of human neonatal foreskin fibroblasts was split into 3 groups (Figure 1A). The first set was directly processed for DNA methylation analysis. The second set was infected with retroviruses encoding the Yamanaka factors, i.e. OCT4, SOX2, KLF4 and cMYC reprogramming factors. The third set was infected with retroviruses encoding the Thomson factors, i.e. OCT4, SOX2, NANOG and LIN28 reprogramming factors. We selected 9 Y-iPSCs and 6 T-iPSCs that had characteristic ESC-like morphology, express several pluripotency genes and can be stably maintained in culture (Figure 1A; Additional file 1: Figure S1). Genomic DNA from each of the iPSCs was subjected to DNA methylation profiling using the Illumina HumanMethylation450 platform. We augmented our data set with published Illumina HumanMethylation450 methylation data for several male ESCs (SIVF002, SIVF025, SIVF043, SIVF044, SIVF050) [14]. To validate the DNA methylation data obtained by the Illumina HumanMethylation450 platform we performed high throughput sequencing based DNA methylation assessment called reduced representation bisulfite sequencing (RRBS) in one of the Y-iPSCs and one of the T-iPSCs. The overall methylation data from Illumina HumanMethylation450 and RRBS data were concordant (with an average correlation coefficient of 0.8) (Additional file 1: Figure S2). These results are very similar to the Pearson correlation between RRBS and Illumina’s DNA methylation arrays obtained in a different study using the same cutoffs [19] and argue in favor of the reliability of the data obtained in our study.Figure 1

Bottom Line: Strikingly, not only were the genomic locations of the aberrations different but also their types: reprogramming with Yamanaka factors mainly resulted in failure to demethylate CpGs, whereas reprogramming with Thomson factors mainly resulted in failure to methylate CpGs.Our study thus reveals that the choice of reprogramming factors influences the amount, location, and class of DNA methylation aberrations in iPSCs.These findings may provide clues into how to produce human iPSCs with fewer DNA methylation abnormalities.

View Article: PubMed Central - PubMed

Affiliation: Campbell Family Cancer Research Institute, Ontario Cancer Institute, Princess Margaret Cancer Centre, University Health Network, Toronto, ON M5G 2 M9 Canada ; Department of Morphology, Piracicaba Dental School, University of Campinas, Piracicaba, SP Brazil.

ABSTRACT
The conversion of somatic cells into pluripotent stem cells via overexpression of reprogramming factors involves epigenetic remodeling. DNA methylation at a significant proportion of CpG sites in induced pluripotent stem cells (iPSCs) differs from that of embryonic stem cells (ESCs). Whether different sets of reprogramming factors influence the type and extent of aberrant DNA methylation in iPSCs differently remains unknown. In order to help resolve this critical question, we generated human iPSCs from a common fibroblast cell source using either the Yamanaka factors (OCT4, SOX2, KLF4 and cMYC) or the Thomson factors (OCT4, SOX2, NANOG and LIN28), and determined their genome-wide DNA methylation profiles. In addition to shared DNA methylation aberrations present in all our iPSCs, we identified Yamanaka-iPSC (Y-iPSC)-specific and Thomson-iPSC (T-iPSC)-specific recurrent aberrations. Strikingly, not only were the genomic locations of the aberrations different but also their types: reprogramming with Yamanaka factors mainly resulted in failure to demethylate CpGs, whereas reprogramming with Thomson factors mainly resulted in failure to methylate CpGs. Differences in the level of transcripts encoding DNMT3b and TET3 between Y-iPSCs and T-iPSCs may contribute partially to the distinct types of aberrations. Finally, de novo aberrantly methylated genes in Y-iPSCs were enriched for NANOG targets that are also aberrantly methylated in some cancers. Our study thus reveals that the choice of reprogramming factors influences the amount, location, and class of DNA methylation aberrations in iPSCs. These findings may provide clues into how to produce human iPSCs with fewer DNA methylation abnormalities.

No MeSH data available.


Related in: MedlinePlus