Functional Annotation of DDX41 Variants Using Base-Editing and cDNA Expression Library Screenings

Mutation of the DDX41 gene is one of the most common causes of germline predisposition to myeloid neoplasms (MNs). While multiple DDX41 variants potentially associated with the risk of developing MNs have been identified, pathogenicity of these variants is often ambiguous, making the management of patients and their relatives challenging. Due primarily to the rarity of the disease, information regarding the pathogenicity of each DDX41 variant is scarce in public databases such as ClinVar. Therefore, there is a critical need to validate the pathogenicity of each variant through functional assays. Here, we present novel tools to functionally assess loss-of-function (LOF) characteristics of DDX41 variants by employing comprehensive base-editing (BE) screening and a cDNA expression library screening of DDX41 variants.

We first established an engineered HAP1, a near-haploid leukemia line, in which a sequence encoding a FKBP12^F36V tag is knocked-in to the DDX41 locus (HAP1^FK), enabling us to degrade endogenous DDX41 protein following treatment with dTAG ligand. dTAG-mediated DDX41 depletion led to cell death, while exogenous expression of wild-type (WT) DDX41 prior to dTAG treatment rescued the phenotype. In contrast, exogenous expression of either common germline (A500fs and D140fs) or somatic (R525H and T227M) pathogenic variants did not rescue the phenotype, suggesting a loss of function (LOF) nature for these variants. To comprehensively identify LOF DDX41 variants in an unbiased manner, we conducted two independent BE screenings targeting DDX41 in HAP1 cells, using cell fitness as a readout. To do so, we established a lentiviral vector system that enables the expression of both a base editor and a sgRNA in a single vector format. We employed either an adenine base editor (ABE) or cytosine base editor (CBE), each fused with SpG, which recognizes an NGN PAM. A total of 3996 sg RNAs targeting all possible bases of DDX41, along with positive and negative control sgRNAs, were introduced into both ABE- and CBE-based vectors. As expected, either disruption of splice sites via ABE/CBE or introduction of nonsense mutations via CBE in DDX41 led to reduced cell fitness, demonstrating the validity of the screening methods. We identified 48 and 299 potential DDX41 LOF variants (Z-score < -2) via CBE and ABE screening, respectively. All these variants are classified as either of uncertain significance, conflicting classifications of pathogenicity, or unannotated in ClinVar (accessed on July 21, 2024).

To independently validate these findings, we next performed a cDNA expression library screening in HAP1^FK cells. We generated a lentiviral cDNA library consisting of all possible DDX41 point mutants (12,441) by oligo synthesis followed by Gibson cloning. Long-read HiFi sequencing revealed that 98.9% of the full-length DDX41 cDNAs were successfully cloned into the lentiviral vector. We introduced the library into the HAP1^FK cells, and cDNA abundance before and after culture in the presence of dTAG ligand was assessed by deep sequencing. We are currently analyzing the data, and the results will be discussed at the meeting.

In summary, we have established cellular systems that enable us to functionally assess the pathogenicity of DDX41 variants in a comprehensive manner. Our findings may help to interpret the pathogenicity of DDX41 variants, which are otherwise of unknown significance, and could guide clinical care for patients and their relatives harboring a germline DDX41 variant.

No relevant conflicts of interest to declare.