Deep mutational scanning of CHD2 for variant interpretation in neurodevelopmental disorders [RNA-seq]

De novo pathogenic variants in CHD2 are one of the most common causes of the neurodevelopmental disorders (NDDs) that include refractory epilepsy, intellectual disability and autism spectrum disorders. In addition, somatic CHD2 variants, acquired later in life, are highly recurrent in certain types of leukemias. For both types of conditions, loss-of-CHD2 function is the likely pathogenic mechanism, and most variants are truncations. However, there is also evidence that missense variants can disrupt or abrogate CHD2 function, leading to loss-of-function, though they are harder to interpret, and many are classified as variants of uncertain significance (VUS). VUS are one of the biggest challenges for human geneticists, and represent the biggest class of variants returned on clinical genetic test reports, in part because we have a poor understanding of how missense variants impact protein function. Our research addresses these challenges by developing a multiplex assay of variant effect (MAVE) to discriminate between pathogenic and benign missense variants. We leveraged the fact that CHD2 is a chromatin remodeler and thus affects gene expression, protein abundance and the DNA methylation landscape to attempt to develop high-throughput assays. We aimed to establish a CHD2 transcriptomic signature using mRNA sequencing from multiple pathogenic and benign variant HAP1 cell lines. While CHD2 is a well-known chromatin remodeler and is expressed at appreciable levels in these cells, we were unable to establish a CHD2-related transcriptomic signature in HAP1s, possibly due to the cell line choice as most manifestations in patients are neurological. Overall design: We aimed to establish a CHD2 transcriptomic signature using short read mRNAseq in clonal HAP1 cell lines, including five benign, 9 pathogenic CHD2 variants using prime-editing and assessed alongside two clonal wildtype lines. mRNA libraries were prepared and sequenced using a paired-end 150bp protocol on an Illumina instrument. Reads we mapped to the genome (hg38) using STAR and quantified using RSEM using default parameters using the nf-core/rnaseq v3.18.0 pipeline. Differentially expressed genes were called with DESeq2, with an adjusted p-value cutoff of 0.05 and an absolute log2FC cutoff of 1.