Titrating gene expression using libraries of systematically attenuated CRISPR guide RNAs.

Area of Science:

• Molecular Biology
• Genetics
• Bioinformatics

Background:

• Precise control over gene expression is crucial for understanding gene function and its link to cellular phenotypes.
• Existing tools lack the fine-tuning capability to systematically modulate gene expression levels.
• Evaluating the quantitative relationship between gene expression and phenotype remains challenging.

Purpose of the Study:

• To develop a novel CRISPR interference (CRISPRi) based system for titrating human gene expression.
• To create a library of single-guide RNAs (sgRNAs) with systematically modulated activities.
• To establish rules for sgRNA activity prediction using deep learning and apply them for large-scale gene expression control.

Main Methods:

• Utilized CRISPR interference (CRISPRi) with a series of single-guide RNAs (sgRNAs) possessing modulated activities.
• Performed large-scale measurements across multiple cell models to characterize sgRNA activities, including those with mismatches.
• Employed deep learning models to derive rules governing mismatched sgRNA activity and synthesized a compact sgRNA library.

Main Results:

• Characterized sgRNA activities and developed predictive rules using deep learning, enabling precise gene expression titration.
• Synthesized a library to titrate expression of ~2,400 essential human genes and created an in silico library for the entire human genome.
• Observed sharp transitions in cellular behaviors at gene-specific expression thresholds when cells were staged along a gene expression continuum.

Conclusions:

• The developed CRISPRi system offers a generalizable tool for precise gene expression control.
• This approach facilitates the study of gene expression thresholds and their impact on cellular phenotypes.
• Applications include tuning biochemical pathways and identifying therapeutic targets for diseases characterized by gene dysregulation.