Dead and Deah-Box Helicases Other Than DDX41 Are Mutated in Myeloid Neoplasia

DEAD (DDX)/ DEAH (DHX)-box family of putative RNA helicases are enzymes belonging to the SF2 helicase superfamily. These enzymes are essential for RNA metabolism and integrity of RNA machinery by conserving the helicase cores while unwinding RNA helices. Mutations of several DDX/DHX-family members have been associated with a variety of conditions such as X-linked neurodevelopmental disorders (DDX3X and DDX54) and early-onset autosomal recessive retinitis pigmentosa (DHX38). To date, the only family member involved in the pathogenesis of myeloid neoplasia (MN) is DDX41, which was identified as a causative gene for late onset familial myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). In light of the correlation between gene paralog and human diseases, we explored whether variants in members of the DDX and DHX-box helicases other than DDX41 may also been observed in MN. To that end, we compiled the most comprehensive phylogenetic analysis of 52 RNA helicases genes in MN, describing variants ontogeny (somatic-SOM vs. germline-GL), molecular associations and exploiting informative cases.

A total of 1,843 patients (including public sources as TCGA and Beat AML) with MN were analyzed by WES/WGS^1,2 and targeted sequencing. We have identified 112 variants belonging to the DDX and 15 to the DHX family of genes in 131 patients. (Fig.1) As in our original report on DDX41 mutations³, the majority (68%) of the variants were GL with DDX41 being the most recurrently mutated gene accounting for up to 75% of cases, of which 15% displaying a biallelic GL plus SOM configuration. We then tested whether GL or SOM hits in other helicases may substitute for the DDX41 GL or SOM mutations carriers, respectively. We did not find any GL variants of other helicases to form compound heterozygous configuration with SOM DDX41 mutations. However, there were two patients who harbored DDX54^SOM/DDX41^SOM, one patient carried DHX29^SOM/DDX41^SOM/DDX54^SOM and one displayed somatic mosaicism with DDX6^SOM/DHX29^SOM configuration.

The second most commonly mutated DDXs gene in our cohort was DDX54 (n=8; GL/SOM=3/5) while the most common DHX mutated gene was DHX29 (n=10; GL/SOM=1/9). In analogy to DDX41, 3/5 with DDX54 mutations and 1/9 DHX20 showed biallelic GL/SOM configuration. Furthermore, we have also found 11 variants in DDX4/DHX29 located as DDX41 on 5q. Consistent with the topography of DDX41, which was recapitulated in our cohort (60% of the DDX41 variants mapped within the DEAD/Helicase ATP binding and Helicase C-terminal functional domains), half of the DDX54 variants also were located in the DEAD/Helicase ATP binding domains.

In terms of MN diagnosis, MDS phenotype was the most frequent (50%), followed by AML (41%) and MPN (9%). Overall, median age of patients was 69 (range, 28-90) and did not differ between carriers of GL or SOM variants. Male gender was predominant in both cases with GL or SOM hits (M/F^GL=2.1; M/F^SOM=5.5). Abnormal karyotype (mostly including +8, del5q and del20q) was found in 40% of our cohort without differences between patients with GL or SOM. Of note, no family history of hematologic malignancies was noted in cases with DDX54 or DHX29 variants. Phenotypically, one male patient with DDX54^GL presented with pancytopenia at 30 years of age, one had a 20-year long-standing history of anemia while the only patient with a DXH29^GL variant developed MDS at 48 years of age. A Medline search led us to associate two GL stopgains in DDX11 (R558* and R829*), a gene belonging to other human helicases (e.g.,XPD, FANCJ, RTEL1) implicated in MN, to Warsaw breakage syndrome. Noteworthy is that we detected a DDX11^R558* (previously reported in consanguineous with such disorder) in a patient with RTEL1 and RAD51 variants and a somatic NRAS^G12D. Finally, with regards to additional myeloid driver lesions, TET2 and DNMT3A were among the most frequently mutated genes in all 3 cohorts of DDX41, DDX54 and DHX29-mutants.

Our study represents the first in-depth analysis of DDX and DHX helicases in MN. One could speculate that paralogs DDX-DHX genes in MN exhibit a redundant function justifying why deleterious variants in more than one gene of the SF2 family do not propagate in the population. Beside the established role of DDX41 in inherited familiar cases, the rare occurrence of mutations in other helicases and their contribution to MN pathogenesis deserves attention and further studies.

Haferlach:Munich Leukemia Laboratory: Current Employment, Other: Part ownership. Maciejewski:Apellis Pharmaceuticals: Consultancy; Alexion: Consultancy.