Limits...
Novel Human Embryonic Stem Cell Regulators Identified by Conserved and Distinct CpG Island Methylation State.

Pells S, Koutsouraki E, Morfopoulou S, Valencia-Cadavid S, Tomlinson SR, Kalathur R, Futschik ME, De Sousa PA - PLoS ONE (2015)

Bottom Line: Transcriptional repressors and activators were over-represented amongst genes whose associated CGIs were methylated or unmethylated specifically in hESCs, respectively.Chromatin immunoprecipitation confirmed interaction between the candidates and the core pluripotency transcription factor network.We thus identify novel pluripotency genes on the basis of a conserved and distinct epigenetic configuration in human stem cells.

View Article: PubMed Central - PubMed

Affiliation: MRC Centre for Regenerative Medicine, School of Clinical Studies, University of Edinburgh, Edinburgh, EH16 4SB, United Kingdom; Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, EH16 4SB, United Kingdom.

ABSTRACT
Human embryonic stem cells (hESCs) undergo epigenetic changes in vitro which may compromise function, so an epigenetic pluripotency "signature" would be invaluable for line validation. We assessed Cytosine-phosphate-Guanine Island (CGI) methylation in hESCs by genomic DNA hybridisation to a CGI array, and saw substantial variation in CGI methylation between lines. Comparison of hESC CGI methylation profiles to corresponding somatic tissue data and hESC mRNA expression profiles identified a conserved hESC-specific methylation pattern associated with expressed genes. Transcriptional repressors and activators were over-represented amongst genes whose associated CGIs were methylated or unmethylated specifically in hESCs, respectively. Knockdown of candidate transcriptional regulators (HMGA1, GLIS2, PFDN5) induced differentiation in hESCs, whereas ectopic expression in fibroblasts modulated iPSC colony formation. Chromatin immunoprecipitation confirmed interaction between the candidates and the core pluripotency transcription factor network. We thus identify novel pluripotency genes on the basis of a conserved and distinct epigenetic configuration in human stem cells.

No MeSH data available.


Related in: MedlinePlus

Genome-wide CGI methylation analysis of hESC lines.(A) Table showing the number of methylated CGIs (Me-CGI) in each hESC line, and for adult somatic tissues (Illingworth et al., 2008), the percentage of methylated CGIs (% Me-CGI), number of methylated gene-associated CGIs (Me-GA-CGI) and the percentage of methylated CGIs that are gene-associated (% Me-GA-CGI). Me-CGIs are given in the supplementary file S1 Table. (B) Venn Diagram shows heterogeneity of hESC CGI methylation. 1111 CGIs are methylated in all 4 lines. (C-G) Genome maps depicting locations of various CGI groups: (C) Me-CGIs in hESC lines (red, gene-associated; green, not gene-associated). (D) 201 GA-CGIs methylated in hESCs but unmethylated in somatic tissues. (E) 98 GA-CGIs unmethylated in hESCs but methylated in somatic tissues. (F) hESC-expressed genes whose associated CGIs are hESC-methylated and unmethylated in somatic tissues. (G) hESC-expressed genes whose associated CGIs are hESC-unmethylated and methylated in somatic tissues. Autosomes ordered 1–22, 23 = X, 24 = Y.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4495055&req=5

pone.0131102.g001: Genome-wide CGI methylation analysis of hESC lines.(A) Table showing the number of methylated CGIs (Me-CGI) in each hESC line, and for adult somatic tissues (Illingworth et al., 2008), the percentage of methylated CGIs (% Me-CGI), number of methylated gene-associated CGIs (Me-GA-CGI) and the percentage of methylated CGIs that are gene-associated (% Me-GA-CGI). Me-CGIs are given in the supplementary file S1 Table. (B) Venn Diagram shows heterogeneity of hESC CGI methylation. 1111 CGIs are methylated in all 4 lines. (C-G) Genome maps depicting locations of various CGI groups: (C) Me-CGIs in hESC lines (red, gene-associated; green, not gene-associated). (D) 201 GA-CGIs methylated in hESCs but unmethylated in somatic tissues. (E) 98 GA-CGIs unmethylated in hESCs but methylated in somatic tissues. (F) hESC-expressed genes whose associated CGIs are hESC-methylated and unmethylated in somatic tissues. (G) hESC-expressed genes whose associated CGIs are hESC-unmethylated and methylated in somatic tissues. Autosomes ordered 1–22, 23 = X, 24 = Y.

Mentions: HESC MAP-gDNA hybridisation data showed that 12–16% of CGIs were methylated in hESCs, depending on cell line (2119 Me-CGIs in RH4 to 2717 in RH3). For consistency, somatic tissue data reported previously[26] were reanalysed in parallel. Similar proportions of CGIs were methylated in somatic tissues as in hESCs, varying from 10–14% (1785 Me-CGIs in male blood to 2546 Me-CGIs in muscle [Fig 1A; corresponding numbers and proportions of unmethylated CGIs are listed in table Q in S1 File]). Overall CGI methylation levels are thus similar in hESCs and somatic tissues. There was no significant difference in CGI methylation rates between hESC lines and somatic tissues for CGIs generally, or for gene-associated CGIs specifically (P = 0.142 in both cases; Kruskall-Wallis).


Novel Human Embryonic Stem Cell Regulators Identified by Conserved and Distinct CpG Island Methylation State.

Pells S, Koutsouraki E, Morfopoulou S, Valencia-Cadavid S, Tomlinson SR, Kalathur R, Futschik ME, De Sousa PA - PLoS ONE (2015)

Genome-wide CGI methylation analysis of hESC lines.(A) Table showing the number of methylated CGIs (Me-CGI) in each hESC line, and for adult somatic tissues (Illingworth et al., 2008), the percentage of methylated CGIs (% Me-CGI), number of methylated gene-associated CGIs (Me-GA-CGI) and the percentage of methylated CGIs that are gene-associated (% Me-GA-CGI). Me-CGIs are given in the supplementary file S1 Table. (B) Venn Diagram shows heterogeneity of hESC CGI methylation. 1111 CGIs are methylated in all 4 lines. (C-G) Genome maps depicting locations of various CGI groups: (C) Me-CGIs in hESC lines (red, gene-associated; green, not gene-associated). (D) 201 GA-CGIs methylated in hESCs but unmethylated in somatic tissues. (E) 98 GA-CGIs unmethylated in hESCs but methylated in somatic tissues. (F) hESC-expressed genes whose associated CGIs are hESC-methylated and unmethylated in somatic tissues. (G) hESC-expressed genes whose associated CGIs are hESC-unmethylated and methylated in somatic tissues. Autosomes ordered 1–22, 23 = X, 24 = Y.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4495055&req=5

pone.0131102.g001: Genome-wide CGI methylation analysis of hESC lines.(A) Table showing the number of methylated CGIs (Me-CGI) in each hESC line, and for adult somatic tissues (Illingworth et al., 2008), the percentage of methylated CGIs (% Me-CGI), number of methylated gene-associated CGIs (Me-GA-CGI) and the percentage of methylated CGIs that are gene-associated (% Me-GA-CGI). Me-CGIs are given in the supplementary file S1 Table. (B) Venn Diagram shows heterogeneity of hESC CGI methylation. 1111 CGIs are methylated in all 4 lines. (C-G) Genome maps depicting locations of various CGI groups: (C) Me-CGIs in hESC lines (red, gene-associated; green, not gene-associated). (D) 201 GA-CGIs methylated in hESCs but unmethylated in somatic tissues. (E) 98 GA-CGIs unmethylated in hESCs but methylated in somatic tissues. (F) hESC-expressed genes whose associated CGIs are hESC-methylated and unmethylated in somatic tissues. (G) hESC-expressed genes whose associated CGIs are hESC-unmethylated and methylated in somatic tissues. Autosomes ordered 1–22, 23 = X, 24 = Y.
Mentions: HESC MAP-gDNA hybridisation data showed that 12–16% of CGIs were methylated in hESCs, depending on cell line (2119 Me-CGIs in RH4 to 2717 in RH3). For consistency, somatic tissue data reported previously[26] were reanalysed in parallel. Similar proportions of CGIs were methylated in somatic tissues as in hESCs, varying from 10–14% (1785 Me-CGIs in male blood to 2546 Me-CGIs in muscle [Fig 1A; corresponding numbers and proportions of unmethylated CGIs are listed in table Q in S1 File]). Overall CGI methylation levels are thus similar in hESCs and somatic tissues. There was no significant difference in CGI methylation rates between hESC lines and somatic tissues for CGIs generally, or for gene-associated CGIs specifically (P = 0.142 in both cases; Kruskall-Wallis).

Bottom Line: Transcriptional repressors and activators were over-represented amongst genes whose associated CGIs were methylated or unmethylated specifically in hESCs, respectively.Chromatin immunoprecipitation confirmed interaction between the candidates and the core pluripotency transcription factor network.We thus identify novel pluripotency genes on the basis of a conserved and distinct epigenetic configuration in human stem cells.

View Article: PubMed Central - PubMed

Affiliation: MRC Centre for Regenerative Medicine, School of Clinical Studies, University of Edinburgh, Edinburgh, EH16 4SB, United Kingdom; Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, EH16 4SB, United Kingdom.

ABSTRACT
Human embryonic stem cells (hESCs) undergo epigenetic changes in vitro which may compromise function, so an epigenetic pluripotency "signature" would be invaluable for line validation. We assessed Cytosine-phosphate-Guanine Island (CGI) methylation in hESCs by genomic DNA hybridisation to a CGI array, and saw substantial variation in CGI methylation between lines. Comparison of hESC CGI methylation profiles to corresponding somatic tissue data and hESC mRNA expression profiles identified a conserved hESC-specific methylation pattern associated with expressed genes. Transcriptional repressors and activators were over-represented amongst genes whose associated CGIs were methylated or unmethylated specifically in hESCs, respectively. Knockdown of candidate transcriptional regulators (HMGA1, GLIS2, PFDN5) induced differentiation in hESCs, whereas ectopic expression in fibroblasts modulated iPSC colony formation. Chromatin immunoprecipitation confirmed interaction between the candidates and the core pluripotency transcription factor network. We thus identify novel pluripotency genes on the basis of a conserved and distinct epigenetic configuration in human stem cells.

No MeSH data available.


Related in: MedlinePlus