Massively parallel phenotyping of coding variants in cancer with Perturb-seq

Area of Science:

• Genomics
• Cancer Biology
• Single-cell analysis

Background:

• Millions of cancer somatic variants are identified, but predicting their functional impact remains difficult.
• Current experimental methods assess variant impact in bulk cell populations, limiting resolution.
• Distinguishing gain-of-function, loss-of-function, and dominant-negative variants is crucial for understanding cancer progression.

Purpose of the Study:

• To develop a scalable, gene-agnostic method for functional variant impact phenotyping in single cancer cells.
• To assess the phenotypic impact of TP53 and KRAS variants using pooled Perturb-seq.
• To categorize variants based on their RNA profiles and validate findings with orthogonal assays.

Main Methods:

• Pooled Perturb-seq was employed to measure the RNA profiles of over 300,000 single lung cancer cells.
• The impact of 200 TP53 and KRAS variants on cellular RNA profiles was analyzed.
• Variants were categorized into functional subsets, and findings were validated using orthogonal assays.

Main Results:

• A novel single-cell approach successfully phenotyped the impact of TP53 and KRAS variants.
• KRAS variants exhibited a continuum of gain-of-function phenotypes, defying discrete categorization.
• Variant functional impact could not be solely predicted by their frequency in patient cohorts.

Conclusions:

• The developed pooled Perturb-seq method offers a scalable and gene-agnostic approach for variant impact phenotyping.
• This method provides deeper insights into variant function beyond traditional classifications.
• The approach has broad potential applications in various disease settings, particularly in cancer research.