Evolution of transcription factor DNA binding sites

Area of Science:

• Bioinformatics
• Genomics
• Molecular Biology

Background:

• Transcription factor binding sites are typically modeled using sequence-specific binding energy, often employing Position Weight Matrix (PWM) or consensus-based models.
• Existing models assume sequence variations are neutral unless they significantly alter binding energy, potentially overlooking functional nuances.

Purpose of the Study:

• To evaluate the sufficiency of available data for constructing relevant PWMs.
• To explore the evolutionary stability and functional significance of non-consensus nucleotides within transcription factor binding sites.
• To propose a modified consensus-based model as a potentially more effective descriptor of binding properties.

Main Methods:

• Analysis of existing bioinformatics data on transcription factor binding sites.
• Comparison of data sufficiency for PWM construction versus modified consensus-based models.
• Examination of nucleotide conservation patterns in orthologous binding sites across different genomes.

Main Results:

• Available data are frequently insufficient for constructing accurate PWMs, suggesting modified consensus models may be superior.
• Non-consensus nucleotides in orthologous binding sites exhibit remarkable evolutionary stability, contrary to expectations of neutrality.
• Conservation of non-consensus nucleotides across distant genomes indicates significant selective pressure, implying specific functional roles in gene regulation.

Conclusions:

• The study challenges the prevailing view of neutral variation in transcription factor binding sites.
• Non-consensus nucleotides are not merely irrelevant but are evolutionarily conserved, suggesting specific functional importance.
• Modified consensus-based models may offer a more accurate representation of transcription factor binding site properties and evolutionary dynamics.