Limits...
Signalling couples hair follicle stem cell quiescence with reduced histone H3 K4/K9/K27me3 for proper tissue homeostasis.

Lee J, Kang S, Lilja KC, Colletier KJ, Scheitz CJ, Zhang YV, Tumbar T - Nat Commun (2016)

Bottom Line: The level of marks over specific gene promoters did not correlate to mRNA level changes in quiescent HFSCs.Furthermore, removal of proliferation factors and addition of BMP4 reduced histone methylases and increased demethylases mRNAs in cultured skin epithelial cells.We conclude that signalling couples hair follicle stem cell quiescence with reduced H3 K4/K9/K27me3 levels for proper tissue homeostasis.

View Article: PubMed Central - PubMed

Affiliation: Department of Molecular Biology and Genetics, Cornell University, Ithaca, New York 14853, USA.

ABSTRACT
Mechanisms of plasticity to acquire different cell fates are critical for adult stem cell (SC) potential, yet are poorly understood. Reduced global histone methylation is an epigenetic state known to mediate plasticity in cultured embryonic SCs and T-cell progenitors. Here we find histone H3 K4/K9/K27me3 levels actively reduced in adult mouse skin and hair follicle stem cells (HFSCs) during G0 quiescence. The level of marks over specific gene promoters did not correlate to mRNA level changes in quiescent HFSCs. Skin hypomethylation during quiescence was necessary for subsequent progression of hair homeostasis (cycle). Inhibiting BMP signal, a known HFSC anti-proliferative factor, elevated HFSC methylation in vivo during quiescence prior to proliferation onset. Furthermore, removal of proliferation factors and addition of BMP4 reduced histone methylases and increased demethylases mRNAs in cultured skin epithelial cells. We conclude that signalling couples hair follicle stem cell quiescence with reduced H3 K4/K9/K27me3 levels for proper tissue homeostasis.

No MeSH data available.


Related in: MedlinePlus

Correlation of level changes in H3 K4/K9/K27me3 changes and mRNA.(a) Density plot of all genes in the mouse genome that show signal by either microarray or ChIP-seq (N=24,778 genes and N=10,345 genes plotted for H3K4me3/H3K9me3 and H3K27me3, respectively, out of total N=34,497 genes. Genes that showed the absence of microarray or ChIP signals in comparing populations were excluded). Genes changed ≥1.5-fold are marked outside the black lines flanking the origin; genes with <1.5-fold variations are considered ‘unchanged'. (b) % of genes that are decreased (blue), unchanged (green) and increased (red) in their histone marks from EA-HFSCs to LC-HFSCs. Each functional category was defined based on changes in mRNA levels in HFSCs at distinct stages of differentiation and self-renewal (GDB: genes downregulated in the bulge). See the Methods for description of arrays used and Supplementary Data 2 for the lists of genes. Note atypical behaviour of genes upregulated in the bulge (GUB) and tumour suppressors. (c) Dot plot of cyclin-dependent kinase inhibitors (CDKis) and known tumour suppressors that play a role in skin and hair follicle biology (Supplementary Data 2). Expression of CDKis was adopted from our previous qRT–PCR data46, whereas mRNA values for the other genes were from microarrays (Supplementary Data 1).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4835553&req=5

f4: Correlation of level changes in H3 K4/K9/K27me3 changes and mRNA.(a) Density plot of all genes in the mouse genome that show signal by either microarray or ChIP-seq (N=24,778 genes and N=10,345 genes plotted for H3K4me3/H3K9me3 and H3K27me3, respectively, out of total N=34,497 genes. Genes that showed the absence of microarray or ChIP signals in comparing populations were excluded). Genes changed ≥1.5-fold are marked outside the black lines flanking the origin; genes with <1.5-fold variations are considered ‘unchanged'. (b) % of genes that are decreased (blue), unchanged (green) and increased (red) in their histone marks from EA-HFSCs to LC-HFSCs. Each functional category was defined based on changes in mRNA levels in HFSCs at distinct stages of differentiation and self-renewal (GDB: genes downregulated in the bulge). See the Methods for description of arrays used and Supplementary Data 2 for the lists of genes. Note atypical behaviour of genes upregulated in the bulge (GUB) and tumour suppressors. (c) Dot plot of cyclin-dependent kinase inhibitors (CDKis) and known tumour suppressors that play a role in skin and hair follicle biology (Supplementary Data 2). Expression of CDKis was adopted from our previous qRT–PCR data46, whereas mRNA values for the other genes were from microarrays (Supplementary Data 1).

Mentions: The global hypomethylation observed in LC-HFSCs prompted us to question the correlation between changes in levels of mRNAs and H3 K4/K9/K27me3 levels at specific TSS gene loci. We used our Affymetrix microarray data from sorted HFSCs (Supplementary Data 1, see the Methods for details) and selected 24,778 genes (for H3 K4/K9me3) and 10,345 genes (for H3K27me3) that showed detectable signal in either microarray or ChIP-seq analysis. We found that in LC-HFSCs, irrespective of changes in mRNA levels, ∼90% genes with H3K4me3 or H3K9me3 on TSS and 64% with H3K27me3 on TSS downregulated the mark (Fig. 4a). As with the analysis of the whole genome, this was again most pronounced for H3K9me3 and least pronounced for H3K4me3. Our analysis of data from the study by Lien et al.23 showed similar reduction for H3K27me3 but not for H3K4me3 in LT-HFSCs (Supplementary Fig. 3); it is likely that H3K4me3 was already regained by the late telogen stage analysed in that study, as also suggested by our western blot analysis (Fig. 1j). Also note that normalization to IgG (not available for Lien et al.23) accounts for accessibility of antibodies to the chromatin and noticeably modify quantitative ChIP-seq data distribution (Supplementary Fig. 3). Comparison of EA-HFSCs and non-EA-HFSCs showed better correlation of mRNA and histone mark levels (Supplementary Fig. 3b), suggesting that the LC-HFSCs might represent a special case (see the Discussion for details).


Signalling couples hair follicle stem cell quiescence with reduced histone H3 K4/K9/K27me3 for proper tissue homeostasis.

Lee J, Kang S, Lilja KC, Colletier KJ, Scheitz CJ, Zhang YV, Tumbar T - Nat Commun (2016)

Correlation of level changes in H3 K4/K9/K27me3 changes and mRNA.(a) Density plot of all genes in the mouse genome that show signal by either microarray or ChIP-seq (N=24,778 genes and N=10,345 genes plotted for H3K4me3/H3K9me3 and H3K27me3, respectively, out of total N=34,497 genes. Genes that showed the absence of microarray or ChIP signals in comparing populations were excluded). Genes changed ≥1.5-fold are marked outside the black lines flanking the origin; genes with <1.5-fold variations are considered ‘unchanged'. (b) % of genes that are decreased (blue), unchanged (green) and increased (red) in their histone marks from EA-HFSCs to LC-HFSCs. Each functional category was defined based on changes in mRNA levels in HFSCs at distinct stages of differentiation and self-renewal (GDB: genes downregulated in the bulge). See the Methods for description of arrays used and Supplementary Data 2 for the lists of genes. Note atypical behaviour of genes upregulated in the bulge (GUB) and tumour suppressors. (c) Dot plot of cyclin-dependent kinase inhibitors (CDKis) and known tumour suppressors that play a role in skin and hair follicle biology (Supplementary Data 2). Expression of CDKis was adopted from our previous qRT–PCR data46, whereas mRNA values for the other genes were from microarrays (Supplementary Data 1).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4835553&req=5

f4: Correlation of level changes in H3 K4/K9/K27me3 changes and mRNA.(a) Density plot of all genes in the mouse genome that show signal by either microarray or ChIP-seq (N=24,778 genes and N=10,345 genes plotted for H3K4me3/H3K9me3 and H3K27me3, respectively, out of total N=34,497 genes. Genes that showed the absence of microarray or ChIP signals in comparing populations were excluded). Genes changed ≥1.5-fold are marked outside the black lines flanking the origin; genes with <1.5-fold variations are considered ‘unchanged'. (b) % of genes that are decreased (blue), unchanged (green) and increased (red) in their histone marks from EA-HFSCs to LC-HFSCs. Each functional category was defined based on changes in mRNA levels in HFSCs at distinct stages of differentiation and self-renewal (GDB: genes downregulated in the bulge). See the Methods for description of arrays used and Supplementary Data 2 for the lists of genes. Note atypical behaviour of genes upregulated in the bulge (GUB) and tumour suppressors. (c) Dot plot of cyclin-dependent kinase inhibitors (CDKis) and known tumour suppressors that play a role in skin and hair follicle biology (Supplementary Data 2). Expression of CDKis was adopted from our previous qRT–PCR data46, whereas mRNA values for the other genes were from microarrays (Supplementary Data 1).
Mentions: The global hypomethylation observed in LC-HFSCs prompted us to question the correlation between changes in levels of mRNAs and H3 K4/K9/K27me3 levels at specific TSS gene loci. We used our Affymetrix microarray data from sorted HFSCs (Supplementary Data 1, see the Methods for details) and selected 24,778 genes (for H3 K4/K9me3) and 10,345 genes (for H3K27me3) that showed detectable signal in either microarray or ChIP-seq analysis. We found that in LC-HFSCs, irrespective of changes in mRNA levels, ∼90% genes with H3K4me3 or H3K9me3 on TSS and 64% with H3K27me3 on TSS downregulated the mark (Fig. 4a). As with the analysis of the whole genome, this was again most pronounced for H3K9me3 and least pronounced for H3K4me3. Our analysis of data from the study by Lien et al.23 showed similar reduction for H3K27me3 but not for H3K4me3 in LT-HFSCs (Supplementary Fig. 3); it is likely that H3K4me3 was already regained by the late telogen stage analysed in that study, as also suggested by our western blot analysis (Fig. 1j). Also note that normalization to IgG (not available for Lien et al.23) accounts for accessibility of antibodies to the chromatin and noticeably modify quantitative ChIP-seq data distribution (Supplementary Fig. 3). Comparison of EA-HFSCs and non-EA-HFSCs showed better correlation of mRNA and histone mark levels (Supplementary Fig. 3b), suggesting that the LC-HFSCs might represent a special case (see the Discussion for details).

Bottom Line: The level of marks over specific gene promoters did not correlate to mRNA level changes in quiescent HFSCs.Furthermore, removal of proliferation factors and addition of BMP4 reduced histone methylases and increased demethylases mRNAs in cultured skin epithelial cells.We conclude that signalling couples hair follicle stem cell quiescence with reduced H3 K4/K9/K27me3 levels for proper tissue homeostasis.

View Article: PubMed Central - PubMed

Affiliation: Department of Molecular Biology and Genetics, Cornell University, Ithaca, New York 14853, USA.

ABSTRACT
Mechanisms of plasticity to acquire different cell fates are critical for adult stem cell (SC) potential, yet are poorly understood. Reduced global histone methylation is an epigenetic state known to mediate plasticity in cultured embryonic SCs and T-cell progenitors. Here we find histone H3 K4/K9/K27me3 levels actively reduced in adult mouse skin and hair follicle stem cells (HFSCs) during G0 quiescence. The level of marks over specific gene promoters did not correlate to mRNA level changes in quiescent HFSCs. Skin hypomethylation during quiescence was necessary for subsequent progression of hair homeostasis (cycle). Inhibiting BMP signal, a known HFSC anti-proliferative factor, elevated HFSC methylation in vivo during quiescence prior to proliferation onset. Furthermore, removal of proliferation factors and addition of BMP4 reduced histone methylases and increased demethylases mRNAs in cultured skin epithelial cells. We conclude that signalling couples hair follicle stem cell quiescence with reduced H3 K4/K9/K27me3 levels for proper tissue homeostasis.

No MeSH data available.


Related in: MedlinePlus