Predictive features of gene expression variation reveal mechanistic link with differential expression

Area of Science:

• Genomics
• Developmental Biology
• Systems Biology

Background:

• Organisms must balance responding to external signals with maintaining core gene expression.
• Controlling gene expression variation in multicellular organisms remains poorly understood, unlike in single cells.

Purpose of the Study:

• To identify genomic features predicting gene expression variation across individuals in multicellular organisms.
• To understand the mechanisms underlying robust versus variable gene expression.
• To investigate the evolutionary conservation and functional implications of expression variation predictors.

Main Methods:

• Utilized a machine-learning approach on bulk embryonic gene expression data from 75 Drosophila isogenic lines.
• Identified genomic features, particularly promoter-related ones, predictive of expression variation, controlling for expression level.
• Applied the developed framework to human data to assess evolutionary conservation.

Main Results:

• Promoter architecture significantly predicts gene expression variation across individuals in both Drosophila and humans.
• Identified two distinct classes of genes with low expression variation, suggesting different robustness mechanisms.
• Genes with high variation appear to lack these stabilizing mechanisms.
• Expression variation features partially predicted differential expression following perturbations, independent of treatment.

Conclusions:

• Promoter architecture is an ancient, conserved mechanism for controlling gene expression variation.
• Genetically encoded, gene-specific features can influence responses to perturbations, partially explaining differential expression signatures.
• Understanding expression variation provides insights into gene regulation and organismal robustness.