Limits...
Distinct polymer physics principles govern chromatin dynamics in mouse and Drosophila topological domains.

Ea V, Sexton T, Gostan T, Herviou L, Baudement MO, Zhang Y, Berlivet S, Le Lay-Taha MN, Cathala G, Lesne A, Victor JM, Fan Y, Cavalli G, Forné T - BMC Genomics (2015)

Bottom Line: Using simple polymer models, we previously showed that, in mouse liver cells, gene-rich domains tend to adopt a statistical helix shape when no significant locus-specific interaction takes place.Interestingly, this statistical helix organization is considerably relaxed in mESC compared to liver cells, indicating that the impact of the constraints responsible for this organization is weaker in pluripotent cells.Finally, depletion of histone H1 in mESC alters local chromatin flexibility but not the statistical helix organization.

View Article: PubMed Central - PubMed

Affiliation: Institut de Génétique Moléculaire de Montpellier, UMR5535, CNRS, Université de Montpellier, 1919 Route de Mende, 34293, Montpellier, Cedex 5, France. vuthy.ea@igmm.cnrs.fr.

ABSTRACT

Background: In higher eukaryotes, the genome is partitioned into large "Topologically Associating Domains" (TADs) in which the chromatin displays favoured long-range contacts. While a crumpled/fractal globule organization has received experimental supports at higher-order levels, the organization principles that govern chromatin dynamics within these TADs remain unclear. Using simple polymer models, we previously showed that, in mouse liver cells, gene-rich domains tend to adopt a statistical helix shape when no significant locus-specific interaction takes place.

Results: Here, we use data from diverse 3C-derived methods to explore chromatin dynamics within mouse and Drosophila TADs. In mouse Embryonic Stem Cells (mESC), that possess large TADs (median size of 840 kb), we show that the statistical helix model, but not globule models, is relevant not only in gene-rich TADs, but also in gene-poor and gene-desert TADs. Interestingly, this statistical helix organization is considerably relaxed in mESC compared to liver cells, indicating that the impact of the constraints responsible for this organization is weaker in pluripotent cells. Finally, depletion of histone H1 in mESC alters local chromatin flexibility but not the statistical helix organization. In Drosophila, which possesses TADs of smaller sizes (median size of 70 kb), we show that, while chromatin compaction and flexibility are finely tuned according to the epigenetic landscape, chromatin dynamics within TADs is generally compatible with an unconstrained polymer configuration.

Conclusions: Models issued from polymer physics can accurately describe the organization principles governing chromatin dynamics in both mouse and Drosophila TADs. However, constraints applied on this dynamics within mammalian TADs have a peculiar impact resulting in a statistical helix organization.

No MeSH data available.


Epigenetic landscapes and chromatin dynamics of the Drosophila chromosome 2 L. “Virtual 3C”, obtained from Hi-C experiments in the Drosophila, were classified according to the four previously defined epigenetic domains (D1 to D4) [6]: D1 (“red chromatin”) corresponds to domains with “active” epigenetic marks, D2 (“black chromatin”) displays no specific epigenetic modifications, D3 (“blue chromatin”) is PcG associated chromatin and D4 (“green chromatin”) is HP1/heterochromatin. The unconstrained chromatin model [eqs.1 and 2] was then fitted and the three best-fit parameters (K = crosslinking efficiency; L = compaction; S = flexibility) were recovered from each “virtual 3C”. Statistical analyses of best-fit parameters were performed separately according to the epigenetic domains. Box-plots show the results obtained for each type of domains on the chromosome 2 L. Stars indicate statistically significant differences: single asterisk indicates a p-value < 0.05 and > 0.01, a double asterisk a p-value < 0.01 and > 0.001 and a triple asterisk a p-value < 0.001 (all p-values are given in Additional file 5). The number of best-fits (n) performed in each domain is as follows: D1: n = 990; D2: n = 2481; D3: n = 624; D4: n = 239). The results obtained from the other Drosophila chromosomes are given in Additional file 3
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4536789&req=5

Fig4: Epigenetic landscapes and chromatin dynamics of the Drosophila chromosome 2 L. “Virtual 3C”, obtained from Hi-C experiments in the Drosophila, were classified according to the four previously defined epigenetic domains (D1 to D4) [6]: D1 (“red chromatin”) corresponds to domains with “active” epigenetic marks, D2 (“black chromatin”) displays no specific epigenetic modifications, D3 (“blue chromatin”) is PcG associated chromatin and D4 (“green chromatin”) is HP1/heterochromatin. The unconstrained chromatin model [eqs.1 and 2] was then fitted and the three best-fit parameters (K = crosslinking efficiency; L = compaction; S = flexibility) were recovered from each “virtual 3C”. Statistical analyses of best-fit parameters were performed separately according to the epigenetic domains. Box-plots show the results obtained for each type of domains on the chromosome 2 L. Stars indicate statistically significant differences: single asterisk indicates a p-value < 0.05 and > 0.01, a double asterisk a p-value < 0.01 and > 0.001 and a triple asterisk a p-value < 0.001 (all p-values are given in Additional file 5). The number of best-fits (n) performed in each domain is as follows: D1: n = 990; D2: n = 2481; D3: n = 624; D4: n = 239). The results obtained from the other Drosophila chromosomes are given in Additional file 3

Mentions: Box-plots in Fig. 4 show the results of statistical analyses of best-fit parameters obtained for chromosome 2 L. We found that “active” domains (D1, “red chromatin”) are less compact (median value of L parameter = 10.81 nm/kb), more efficiently cross-linked (median value of K parameter = 0.85) and more flexible (median value of S parameter = 4.15 kb) than the other domains (L = 10.56/10.66/10.32 nm/kb for D2/D3/D4 respectively while K = 1.49/1.34/2.40 and S = 4.92/4.84/5.30 kb for D2/D3/D4 respectively) (Fig. 4). As expected, we found that HP1/heterochromatin (D4) is much less flexible and more compact than any other type of chromatin. However, “black” (D2) and PcG (D3) chromatins have very similar flexibility and compaction, suggesting that PcG proteins do not significantly impact on local chromatin dynamics (Fig. 4). Identical results were found for all the other Drosophila chromosomes, except for the tiny chromosome 4, which displayed quite flexible and poorly compacted chromatin despite being entirely heterochromatic (Table 2) (full data are in Additional file 3. Additional file 5 gives Wilcox p-values of differences observed between the different epigenetic domains for parameters shown in Table 2). This finding is consistent with a recent work demonstrating that chromosome 4 displays distinct epigenetic profiles compared to both pericentric heterochromatin and euchromatic regions and that enrichment of HP1a on chromosome 4 genes creates an alternate chromatin structure which is critical for their regulation [29]. Globally, these experiments confirm that the epigenetic contexts influence significantly the local chromatin dynamics in vivo. However, quantitatively, their effects on chromatin compaction and flexibility appear as being quite limited. Indeed, the largest variations observed (between the “active” and HP1/heterochromatin domains) for chromatin compaction and flexibility are 10.76 to 9.99 nm/kb, i.e. about 7 %, on chromosome 2R, and 4.090 to 5.382 kb, i.e. about 24 %, on chromosome 3 L, respectively (Table 2). Therefore, the epigenetic landscape in the fly appears to be involved in fine-tuning the local chromatin dynamics.Fig. 4


Distinct polymer physics principles govern chromatin dynamics in mouse and Drosophila topological domains.

Ea V, Sexton T, Gostan T, Herviou L, Baudement MO, Zhang Y, Berlivet S, Le Lay-Taha MN, Cathala G, Lesne A, Victor JM, Fan Y, Cavalli G, Forné T - BMC Genomics (2015)

Epigenetic landscapes and chromatin dynamics of the Drosophila chromosome 2 L. “Virtual 3C”, obtained from Hi-C experiments in the Drosophila, were classified according to the four previously defined epigenetic domains (D1 to D4) [6]: D1 (“red chromatin”) corresponds to domains with “active” epigenetic marks, D2 (“black chromatin”) displays no specific epigenetic modifications, D3 (“blue chromatin”) is PcG associated chromatin and D4 (“green chromatin”) is HP1/heterochromatin. The unconstrained chromatin model [eqs.1 and 2] was then fitted and the three best-fit parameters (K = crosslinking efficiency; L = compaction; S = flexibility) were recovered from each “virtual 3C”. Statistical analyses of best-fit parameters were performed separately according to the epigenetic domains. Box-plots show the results obtained for each type of domains on the chromosome 2 L. Stars indicate statistically significant differences: single asterisk indicates a p-value < 0.05 and > 0.01, a double asterisk a p-value < 0.01 and > 0.001 and a triple asterisk a p-value < 0.001 (all p-values are given in Additional file 5). The number of best-fits (n) performed in each domain is as follows: D1: n = 990; D2: n = 2481; D3: n = 624; D4: n = 239). The results obtained from the other Drosophila chromosomes are given in Additional file 3
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4536789&req=5

Fig4: Epigenetic landscapes and chromatin dynamics of the Drosophila chromosome 2 L. “Virtual 3C”, obtained from Hi-C experiments in the Drosophila, were classified according to the four previously defined epigenetic domains (D1 to D4) [6]: D1 (“red chromatin”) corresponds to domains with “active” epigenetic marks, D2 (“black chromatin”) displays no specific epigenetic modifications, D3 (“blue chromatin”) is PcG associated chromatin and D4 (“green chromatin”) is HP1/heterochromatin. The unconstrained chromatin model [eqs.1 and 2] was then fitted and the three best-fit parameters (K = crosslinking efficiency; L = compaction; S = flexibility) were recovered from each “virtual 3C”. Statistical analyses of best-fit parameters were performed separately according to the epigenetic domains. Box-plots show the results obtained for each type of domains on the chromosome 2 L. Stars indicate statistically significant differences: single asterisk indicates a p-value < 0.05 and > 0.01, a double asterisk a p-value < 0.01 and > 0.001 and a triple asterisk a p-value < 0.001 (all p-values are given in Additional file 5). The number of best-fits (n) performed in each domain is as follows: D1: n = 990; D2: n = 2481; D3: n = 624; D4: n = 239). The results obtained from the other Drosophila chromosomes are given in Additional file 3
Mentions: Box-plots in Fig. 4 show the results of statistical analyses of best-fit parameters obtained for chromosome 2 L. We found that “active” domains (D1, “red chromatin”) are less compact (median value of L parameter = 10.81 nm/kb), more efficiently cross-linked (median value of K parameter = 0.85) and more flexible (median value of S parameter = 4.15 kb) than the other domains (L = 10.56/10.66/10.32 nm/kb for D2/D3/D4 respectively while K = 1.49/1.34/2.40 and S = 4.92/4.84/5.30 kb for D2/D3/D4 respectively) (Fig. 4). As expected, we found that HP1/heterochromatin (D4) is much less flexible and more compact than any other type of chromatin. However, “black” (D2) and PcG (D3) chromatins have very similar flexibility and compaction, suggesting that PcG proteins do not significantly impact on local chromatin dynamics (Fig. 4). Identical results were found for all the other Drosophila chromosomes, except for the tiny chromosome 4, which displayed quite flexible and poorly compacted chromatin despite being entirely heterochromatic (Table 2) (full data are in Additional file 3. Additional file 5 gives Wilcox p-values of differences observed between the different epigenetic domains for parameters shown in Table 2). This finding is consistent with a recent work demonstrating that chromosome 4 displays distinct epigenetic profiles compared to both pericentric heterochromatin and euchromatic regions and that enrichment of HP1a on chromosome 4 genes creates an alternate chromatin structure which is critical for their regulation [29]. Globally, these experiments confirm that the epigenetic contexts influence significantly the local chromatin dynamics in vivo. However, quantitatively, their effects on chromatin compaction and flexibility appear as being quite limited. Indeed, the largest variations observed (between the “active” and HP1/heterochromatin domains) for chromatin compaction and flexibility are 10.76 to 9.99 nm/kb, i.e. about 7 %, on chromosome 2R, and 4.090 to 5.382 kb, i.e. about 24 %, on chromosome 3 L, respectively (Table 2). Therefore, the epigenetic landscape in the fly appears to be involved in fine-tuning the local chromatin dynamics.Fig. 4

Bottom Line: Using simple polymer models, we previously showed that, in mouse liver cells, gene-rich domains tend to adopt a statistical helix shape when no significant locus-specific interaction takes place.Interestingly, this statistical helix organization is considerably relaxed in mESC compared to liver cells, indicating that the impact of the constraints responsible for this organization is weaker in pluripotent cells.Finally, depletion of histone H1 in mESC alters local chromatin flexibility but not the statistical helix organization.

View Article: PubMed Central - PubMed

Affiliation: Institut de Génétique Moléculaire de Montpellier, UMR5535, CNRS, Université de Montpellier, 1919 Route de Mende, 34293, Montpellier, Cedex 5, France. vuthy.ea@igmm.cnrs.fr.

ABSTRACT

Background: In higher eukaryotes, the genome is partitioned into large "Topologically Associating Domains" (TADs) in which the chromatin displays favoured long-range contacts. While a crumpled/fractal globule organization has received experimental supports at higher-order levels, the organization principles that govern chromatin dynamics within these TADs remain unclear. Using simple polymer models, we previously showed that, in mouse liver cells, gene-rich domains tend to adopt a statistical helix shape when no significant locus-specific interaction takes place.

Results: Here, we use data from diverse 3C-derived methods to explore chromatin dynamics within mouse and Drosophila TADs. In mouse Embryonic Stem Cells (mESC), that possess large TADs (median size of 840 kb), we show that the statistical helix model, but not globule models, is relevant not only in gene-rich TADs, but also in gene-poor and gene-desert TADs. Interestingly, this statistical helix organization is considerably relaxed in mESC compared to liver cells, indicating that the impact of the constraints responsible for this organization is weaker in pluripotent cells. Finally, depletion of histone H1 in mESC alters local chromatin flexibility but not the statistical helix organization. In Drosophila, which possesses TADs of smaller sizes (median size of 70 kb), we show that, while chromatin compaction and flexibility are finely tuned according to the epigenetic landscape, chromatin dynamics within TADs is generally compatible with an unconstrained polymer configuration.

Conclusions: Models issued from polymer physics can accurately describe the organization principles governing chromatin dynamics in both mouse and Drosophila TADs. However, constraints applied on this dynamics within mammalian TADs have a peculiar impact resulting in a statistical helix organization.

No MeSH data available.