Phylogenetic Modeling of Regulatory Element Turnover Based on Epigenomic Data

Area of Science:

• Genomics
• Evolutionary Biology
• Bioinformatics

Background:

• Gene expression evolution is driven by changes in cis-regulatory elements (CREs).
• Previous analyses of CRE evolution using epigenomic data were largely descriptive and model-free.
• Understanding CRE dynamics requires integrating phylogenetic information with epigenomic data.

Purpose of the Study:

• To introduce a novel probabilistic modeling framework for analyzing CRE evolution.
• To directly utilize raw chromatin immunoprecipitation and sequencing (ChIP-seq) data, incorporating phylogenetic relationships.
• To investigate the evolutionary dynamics of CREs, including gains and losses, across mammalian species.

Main Methods:

• Developed a phylogenetic hidden Markov model (epiPhyloHMM) for identifying multiply aligned CREs.
• Created a combined phylogenetic and generalized linear model (phyloGLM) to analyze CRE evolutionary dynamics.
• Applied these models to H3K4me3 and H3K27ac ChIP-seq data from nine mammalian liver tissues.

Main Results:

• Enhancers are gained and lost at approximately twice the rate of promoters during mammalian evolution.
• CRE turnover rates show negative correlations with DNA sequence conservation, expression level, and tissue breadth.
• Turnover rates positively correlate with distance from the transcription start site.
• Dosage sensitivity of target genes correlates with CRE DNA sequence constraint, but not turnover rates.

Conclusions:

• The developed probabilistic framework provides a powerful new approach for analyzing CRE evolution.
• Findings reveal distinct evolutionary dynamics for enhancers and promoters.
• The study highlights the interplay between CRE evolution, sequence constraint, and gene properties like dosage sensitivity.