Florence M. Chardon, Chase C. Suiter, Riza Daza, Nahum T. Smith, Phoebe Parrish, Troy A. McDiarmid, Jean-Benoît Lalanne, Amira Ellison, Beth Martin, Diego Calderon, Alice H. Berger, Jay Shendure, and Lea M. Starita

Abstract

CRISPR-based genome editing has revolutionized the field of functional genomics, allowing for the study of precise insertions, substitutions, and deletions in desired genes. However, precise genome edits have, until recently, required a homology-directed repair (HDR) template in addition to a locus-directing gRNA to install the desired edit. This decoupling of target and edit information has inhibited the study of precise edits on a genome-wide scale, requiring individual experiments for each locus of interest. CRISPR prime editing, which couples target and edit information in a single prime editing gRNA (pegRNA), has been used successfully to install precise edits at single loci, but its low efficiency has precluded its use in a genome-wide manner. Here, we introduce a multiplex prime editing framework to assay the effects of thousands of single nucleotide variants across many genes at once in a single, pooled experiment. We overcome the low efficiency of prime editing by devising a positive selection screen and use this method to assess thousands of single nucleotide variants across eight oncogenes for drug resistance against three different EGFR tyrosine kinase inhibitors. Our multiplex prime editing method is able to resolve both well-characterized drug resistance mutations and identify potential novel drug resistance mutations. This method, in theory, scales without limits and this approach can be extended to assay the drug resistance behavior of mutations in additional genes to other therapeutics.

Files

We have made all data files available for further analysis or visualization. The following are available:

1. epegRNA designs, sequences, and primer sequences for the EGFR C797S editing experiments, 121 epegRNA pooled screens, and 3,825 epegRNA pooled screen

  1. EGFR C797S arrayed editing experiments

  2. 121 epegRNA pooled screens

  3. 3,825 epegRNA pooled screen

2. PrimeDesign input and output files:

  1. 121 epegRNA pooled screens

  2. 3,825 epegRNA pooled screen

3. Sequencing fastq and count output files:

  1. EGFR C797S arrayed editing experiments

  2. 121 epegRNA pooled screens

  3. 3,825 epegRNA pooled screen

4. Sequence search file with detailed information about each epegRNA in the 3,825 epegRNA pooled screen:

  1. detailed csv file

5. Statistical testing results for pooled screens

  1. 121 epegRNA pooled screens

  2. 3,825 epegRNA pooled screens

KRAS G12C and EGFR C797S variant trends in drug treated cells versus DMSO treated cells.

KRAS G12C and EGFR C797S variant trends in drug treated cells versus DMSO treated cells.

Quantile-quantile plots showing the distribution of expected versus observed p-values for the three different drug screens. Each point represents a unique genetic variant encoded by 1-4 epegRNAs.

Quantile-quantile plots showing the distribution of expected versus observed p-values for the three different drug screens. Each point represents a unique genetic variant encoded by 1-4 epegRNAs.