Limits...
Distinct and predictive histone lysine acetylation patterns at promoters, enhancers, and gene bodies.

Rajagopal N, Ernst J, Ray P, Wu J, Zhang M, Kellis M, Ren B - G3 (Bethesda) (2014)

Bottom Line: Unexpectedly, we found that histone acetylation alone performs well in distinguishing these unique genomic regions.Further, we found the association of characteristic acetylation patterns with genic regions and association of chromatin state with splicing.Taken together, our work underscores the diverse functional roles of histone acetylation in gene regulation and provides several testable hypotheses to dissect these roles.

View Article: PubMed Central - PubMed

Affiliation: Ludwig Institute for Cancer Research, 9500 Gilman Drive, La Jolla, California 92093-0653 Bioinformatics and Systems Biology Program, University of California, San Diego, La Jolla, California 92037 Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139.

Show MeSH

Related in: MedlinePlus

Classification of distal enhancers and promoters. (A) Preference of various histone modifications for either enhancer or promoter using a Z-score normalized score of histone modification levels measured as input-subtracted RPKM (reads per kilobase per million) in 1H (blue bars) and IMR90 (red bars).Modifications with preference for promoters, measured as enrichment on the positive y-axis, in both 1H and IMR90, are shown indicated in red text color on the x-axis label while preference for enhancers or enrichment on the negative y-axis in both cell-types is indicated in blue text color. (B) Classification accuracy achieved using each of the 24 histone modifications individually to separate enhancers from promoters using RFECS in three distinct cell-lines: 1H (blue line), IMR90 (red line), and H9 (green line). H9 is another embryonic stem cell line that was used in this case to see if 1H-specific trends were in fact embryonic stem cell–specific. Modifications with the topmost classification accuracy in both 1H and IMR90 are shown in either red or blue text color, depending on whether they have preference for promoters or enhancers in both cell types. (C) Comparison of classification accuracy of acetylations with that of all 24 modifications. (D) Ordering of histone acetylations by their out-of-bag variable importance in classification of enhancers against promoters in 1H. Correlation clustering of histone acetylations at promoters and enhancers in (E) IMR90 and (F) 1H. Acetylations that show similar patterns of co-occurrence in both cell types are indicated in red text color along the axes.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4232531&req=5

fig1: Classification of distal enhancers and promoters. (A) Preference of various histone modifications for either enhancer or promoter using a Z-score normalized score of histone modification levels measured as input-subtracted RPKM (reads per kilobase per million) in 1H (blue bars) and IMR90 (red bars).Modifications with preference for promoters, measured as enrichment on the positive y-axis, in both 1H and IMR90, are shown indicated in red text color on the x-axis label while preference for enhancers or enrichment on the negative y-axis in both cell-types is indicated in blue text color. (B) Classification accuracy achieved using each of the 24 histone modifications individually to separate enhancers from promoters using RFECS in three distinct cell-lines: 1H (blue line), IMR90 (red line), and H9 (green line). H9 is another embryonic stem cell line that was used in this case to see if 1H-specific trends were in fact embryonic stem cell–specific. Modifications with the topmost classification accuracy in both 1H and IMR90 are shown in either red or blue text color, depending on whether they have preference for promoters or enhancers in both cell types. (C) Comparison of classification accuracy of acetylations with that of all 24 modifications. (D) Ordering of histone acetylations by their out-of-bag variable importance in classification of enhancers against promoters in 1H. Correlation clustering of histone acetylations at promoters and enhancers in (E) IMR90 and (F) 1H. Acetylations that show similar patterns of co-occurrence in both cell types are indicated in red text color along the axes.

Mentions: We previously observed that H3K4me1 and H3K4me3 are the most distinctive marks between promoters and enhancers among a limited set of five histone modifications (Heintzman et al. 2007). To further define the marks that distinguish these two regulatory elements in genome-wide maps of 24 histone modifications (Rajagopal et al. 2013), we compared active TSSs (TSSs overlapping DNase-I HS sites) with an equal number of enhancers defined by TSS-distal p300 binding. After normalization (Materials and Methods), we observe that the mean histone modification profile of either class separates clearly into TSS-preferred and enhancer-preferred groups (Figure 1A, positive vs. negative axes). We confirmed that the deviation of most of the histone modifications from a set of elements with randomly shuffled labels is statistically significant for total normalized read counts within −1 to +1 kb of the element (Figure 1A; p-value <10−5 using Wilcoxon test, except for bars marked by black dots). In both 1H and IMR90 cells, we consistently found that H3K4me1, H2BK20ac, and H2BK120ac are significantly enhancer-preferred, whereas H3K4me3, H3K4me2, H3K9ac, H3K56ac, H4K5ac, and H3K27ac are TSS-preferred (Figure 1B). The histone modification profiles in bin sizes of 100 bp between −1 and +1 kb along these elements are also observed to be different from the random set (Supporting Information, Figure S1, A and B, blue vs. red).


Distinct and predictive histone lysine acetylation patterns at promoters, enhancers, and gene bodies.

Rajagopal N, Ernst J, Ray P, Wu J, Zhang M, Kellis M, Ren B - G3 (Bethesda) (2014)

Classification of distal enhancers and promoters. (A) Preference of various histone modifications for either enhancer or promoter using a Z-score normalized score of histone modification levels measured as input-subtracted RPKM (reads per kilobase per million) in 1H (blue bars) and IMR90 (red bars).Modifications with preference for promoters, measured as enrichment on the positive y-axis, in both 1H and IMR90, are shown indicated in red text color on the x-axis label while preference for enhancers or enrichment on the negative y-axis in both cell-types is indicated in blue text color. (B) Classification accuracy achieved using each of the 24 histone modifications individually to separate enhancers from promoters using RFECS in three distinct cell-lines: 1H (blue line), IMR90 (red line), and H9 (green line). H9 is another embryonic stem cell line that was used in this case to see if 1H-specific trends were in fact embryonic stem cell–specific. Modifications with the topmost classification accuracy in both 1H and IMR90 are shown in either red or blue text color, depending on whether they have preference for promoters or enhancers in both cell types. (C) Comparison of classification accuracy of acetylations with that of all 24 modifications. (D) Ordering of histone acetylations by their out-of-bag variable importance in classification of enhancers against promoters in 1H. Correlation clustering of histone acetylations at promoters and enhancers in (E) IMR90 and (F) 1H. Acetylations that show similar patterns of co-occurrence in both cell types are indicated in red text color along the axes.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4232531&req=5

fig1: Classification of distal enhancers and promoters. (A) Preference of various histone modifications for either enhancer or promoter using a Z-score normalized score of histone modification levels measured as input-subtracted RPKM (reads per kilobase per million) in 1H (blue bars) and IMR90 (red bars).Modifications with preference for promoters, measured as enrichment on the positive y-axis, in both 1H and IMR90, are shown indicated in red text color on the x-axis label while preference for enhancers or enrichment on the negative y-axis in both cell-types is indicated in blue text color. (B) Classification accuracy achieved using each of the 24 histone modifications individually to separate enhancers from promoters using RFECS in three distinct cell-lines: 1H (blue line), IMR90 (red line), and H9 (green line). H9 is another embryonic stem cell line that was used in this case to see if 1H-specific trends were in fact embryonic stem cell–specific. Modifications with the topmost classification accuracy in both 1H and IMR90 are shown in either red or blue text color, depending on whether they have preference for promoters or enhancers in both cell types. (C) Comparison of classification accuracy of acetylations with that of all 24 modifications. (D) Ordering of histone acetylations by their out-of-bag variable importance in classification of enhancers against promoters in 1H. Correlation clustering of histone acetylations at promoters and enhancers in (E) IMR90 and (F) 1H. Acetylations that show similar patterns of co-occurrence in both cell types are indicated in red text color along the axes.
Mentions: We previously observed that H3K4me1 and H3K4me3 are the most distinctive marks between promoters and enhancers among a limited set of five histone modifications (Heintzman et al. 2007). To further define the marks that distinguish these two regulatory elements in genome-wide maps of 24 histone modifications (Rajagopal et al. 2013), we compared active TSSs (TSSs overlapping DNase-I HS sites) with an equal number of enhancers defined by TSS-distal p300 binding. After normalization (Materials and Methods), we observe that the mean histone modification profile of either class separates clearly into TSS-preferred and enhancer-preferred groups (Figure 1A, positive vs. negative axes). We confirmed that the deviation of most of the histone modifications from a set of elements with randomly shuffled labels is statistically significant for total normalized read counts within −1 to +1 kb of the element (Figure 1A; p-value <10−5 using Wilcoxon test, except for bars marked by black dots). In both 1H and IMR90 cells, we consistently found that H3K4me1, H2BK20ac, and H2BK120ac are significantly enhancer-preferred, whereas H3K4me3, H3K4me2, H3K9ac, H3K56ac, H4K5ac, and H3K27ac are TSS-preferred (Figure 1B). The histone modification profiles in bin sizes of 100 bp between −1 and +1 kb along these elements are also observed to be different from the random set (Supporting Information, Figure S1, A and B, blue vs. red).

Bottom Line: Unexpectedly, we found that histone acetylation alone performs well in distinguishing these unique genomic regions.Further, we found the association of characteristic acetylation patterns with genic regions and association of chromatin state with splicing.Taken together, our work underscores the diverse functional roles of histone acetylation in gene regulation and provides several testable hypotheses to dissect these roles.

View Article: PubMed Central - PubMed

Affiliation: Ludwig Institute for Cancer Research, 9500 Gilman Drive, La Jolla, California 92093-0653 Bioinformatics and Systems Biology Program, University of California, San Diego, La Jolla, California 92037 Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139.

Show MeSH
Related in: MedlinePlus