Identification of DHX40 as a candidate susceptibility gene for colorectal and hematological neoplasia

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Identification of DHX40 as a candidate susceptibility gene for colorectal and hematological neoplasia © 2023 the Authors Published version Olkinuora, Alisa; Nieminen, Taina T.; Douglas, Suvi; Kauppinen, Anni; Kontro, Mika; Väänänen, Juho; Kankainen, Matti; Ristimäki, Ari; Mäkinen, Markus; Lahermo, Päivi; Heckman, Caroline; Saarela, Janna; Salonen, Milla; Lepistö, Anna; Järvinen, Heikki; Mecklin, Jukka-Pekka; Kilpivaara, Outi; Wartiovaara-Kautto, Ulla; Porkka, Kimmo; Peltomäki, Päivi Olkinuora, A., Nieminen, T. T., Douglas, S., Kauppinen, A., Kontro, M., Väänänen, J., Kankainen, M., Ristimäki, A., Mäkinen, M., Lahermo, P., Heckman, C., Saarela, J., Salonen, M., Lepistö, A., Järvinen, H., Mecklin, J.-P., Kilpivaara, O., Wartiovaara-Kautto, U., Porkka, K., & Peltomäki, P. (2023). Identification of DHX40 as a candidate susceptibility gene for colorectal and hematological neoplasia. Leukemia, 37, 2301-2305. https://doi.org/10.1038/s41375-023-02021- 9 2023 LETTER OPEN CYTOGENETICS AND MOLECULAR GENETICS Identification of DHX40 as a candidate susceptibility gene for colorectal and hematological neoplasia Alisa Olkinuora 1 ✉, Taina T. Nieminen 1 ✉, Suvi Douglas 1,2 , Anni Kauppinen 1 , Mika Kontro 3,4,5 , Juho Väänänen 2 , Matti Kankainen 6,7,8,9 , Ari Ristimäki 2,10 , Markus Mäkinen 11 , Päivi Lahermo 4 , Caroline Heckman 4,8 , Janna Saarela 4,12 , Milla Salonen 1,13,14 , Anna Lepistö 2,15 , Heikki Järvinen 15 , Jukka-Pekka Mecklin 16,17 , Outi Kilpivaara 1,2,6 , Ulla Wartiovaara-Kautto 2,3 , Kimmo Porkka 3,8 ✉and Päivi Peltomäki 1,6 ✉ © The Author(s) 2023 Leukemia; https://doi.org/10.1038/s41375-023-02021-9 TO THE EDITOR: Accumulating evidence suggests shared susceptibility factors between colorectal and hematological malignancies, in line with observations of cross-cancer associations of multiple cancer predisposing genes [1]. Familial colorectal cancer type X (FCCTX) refers to colorectal cancer families that fulfill the diagnostic (Amsterdam) criteria for Lynch syndrome but lack DNA mismatch repair (MMR) defects. The underlying genes are mostly unknown. We recently linked inactivation of RPS20 (ribosomal protein S20) to FCCTX predisposition [2]. The phenotypic spectrum of RPS20 was later expanded to include Diamond Blackfan anemia [3] thus connecting colorectal and hematological carcinogenesis. Observations of inherited variants in DDX41, a DEAD box RNA helicase gene, causing susceptibility to myelodysplastic syndrome and myeloid leukemias, offer additional proof of the important regulatory roles of ribosome biosynthesis and RNA processing in cancer [4]. DDX41 variants are currently known to represent the most common germline alterations in adult myelodysplastic syndromes, accounting for 0.5–4% of all acute myeloid leukemia or myelodysplastic syndrome cases in adults [5]. We conducted exome sequencing (ES) on 28 unexplained FCCTX families to identify their predisposing genes (Supplementary Materials and Methods, Supplementary Table 1). The investigations described in this paper were approved by the Institutional Review Boards of the Helsinki University Hospital, Helsinki, Finland (approval nos. 466-46-2001, 206/13/03/03/2016 and 303/13/03/01/2011). Family F32 with myelodysplastic syndrome/acute leukemia coexisting with colorectal cancer caught our attention (Fig. 1A). All variants fulfilling our selection criteria (high quality, minor allele frequency <0.001, nonsynonymous, and predicted pathogenic with at least 5/6 in silico programs if missense) are listed for the index individual III.1 in Supplementary Table 2. His granddaughter V.1 presented with an unusual form of hematologic disorder at the age of 18 years. Initially, the investigations started due to pancytopenia which was treated with allo-hematopoietic stem cell transplantation (HSCT). The disease relapsed 2 years later with blood counts: hemoglobin 9.5, platelets 50, neutrophils 0.7. At that time, morphological examination of the bone marrow showed dysplasia in megakaryopoiesis and excess of blasts, and was subsequently diagnosed as myelodysplastic syndrome (MDS)/refractory anemia with excess blasts, type 2 (RAEB2) [6]. When variant sharing in the key affected members III.1, IV.3, and V.1 was set as a requirement, the only variant that remained was a heterozygous truncating variant affecting the DEAH-box RNA helicase gene DHX40 (NM_024612.4:c.710_713delTCAG, p.Val237GlufsX7) (Fig. 1A). The variant is predicted to lead to the deletion of the helicase domain and all downstream C-terminal portion of DHX40. Seven of eight cancer-affected members had this variant (the only one without was individual III.3 with adenocarcinoma of unknown origin). To functionally characterize the DHX40 variant, we evaluated its expressional consequences on RNA and protein level, undertook tumor studies to analyze somatic variants, changes in gene methylation and DHX40 protein expression, and conducted siRNA analyses of cell lines (see Supplementary Materials and Methods). By a primer extension assay, the dosage ratios of mutant to wild- Received: 21 June 2023 Revised: 15 August 2023 Accepted: 1 September 2023 1 Department of Medical and Clinical Genetics, University of Helsinki, 00014 Helsinki, Finland. 2 Applied Tumor Genomics Research Program, Faculty of Medicine, University of Helsinki, 00014 Helsinki, Finland. 3 Department of Hematology, Helsinki University Hospital, Comprehensive Cancer Center and University of Helsinki, 00014 Helsinki, Finland. 4 HiLIFE Institute for Molecular Medicine Finland (FIMM), University of Helsinki, 00014 Helsinki, Finland. 5 Foundation for the Finnish Cancer Institute, 00014 Helsinki, Finland. 6 HUSLAB Laboratory of Genetics, HUS Diagnostic Center, HUS, Helsinki University Hospital, 00029 Helsinki, Finland. 7 Hematology Research Unit Helsinki, University of Helsinki, 00014 Helsinki, Finland. 8 iCAN Digital Precision Cancer Medicine Flagship, Helsinki, Finland. 9 Translational Immunology Research Program and Department of Clinical Chemistry and Hematology, University of Helsinki, 00014 Helsinki, Finland. 10 Department of Pathology, HUSLAB, HUS Diagnostic Center, Helsinki University Hospital and University of Helsinki, 00014 Helsinki, Finland. 11 Research Unit of Cancer and Translational Medicine, Department of Pathology, 90014, University of Oulu, and Department of Pathology, Oulu University Hospital, OYS, 90029 Oulu, Finland. 12 Centre for Molecular Medicine Norway, NCMM, University of Oslo, 0318 Oslo, Norway. 13 Department of Veterinary Biosciences, University of Helsinki, 00014 Helsinki, Finland. 14 Folkhälsan Research Center, 00290 Helsinki, Finland. 15 Department of Abdominal Surgery, Helsinki University Hospital and University of Helsinki, 00014 Helsinki, Finland. 16 Department of Education & Research and Surgery, Jyväskylä Central Hospital, 40620 Jyväskylä, Finland. 17 Department of Sports & Health Sciences, Jyväskylä University, 40014 Jyväskylä, Finland. ✉email: alisa.olkinuora@helsinki.fi; taina.nieminen@helsinki.fi; kimmo.porkka@helsinki.fi; paivi.peltomaki@helsinki.fi www.nature.com/leu Leukemia 1234567890();,: type alleles were reduced up to fivefold (Supplementary Fig 1A), likely reflecting nonsense-mediated RNA decay. In agreement, Western blot analysis with N-terminal DHX40 antibody showed no visible truncated protein (Supplementary Fig. 2A). However, a truncated product of expected size could be clearly visualized by transfecting HEK293 cells with eGFP-tagged DHX40 expression constructs (Supplementary Fig. 2B). To evaluate if Knudson’s twohit mechanism applied to DHX40, available neoplastic tissues from individuals heterozygous for the DHX40 c.710_713delTCAG variant were examined for somatic variants, loss of the wild-type allele, and promoter methylation of DHX40. There was no unequivocal evidence of “second hits”(Supplementary Fig. 1B, Supplementary -3 -2.5 -2 -1.5 -1 -0.5 0 unique to DHX40 siRNA treated cells (absent in GAPDH or Non-Target siRNA treated cells) log2FC change versus Untreated HEK293 CCD841 K562 DHX40 SAMM50 MBNL1 MEX3A CREBL2 EMP2 RBL2 LZIC PURA MAEA SLC30A7 GCLM RAB29 MTR PTGES3 KRBOX5 MSANTD4 CDR2L PRKRA CLTC 0 200 400 600 800 1000 1200 1400 1600 * * * * * * siRNA •DHX40 •GAPDH •Non-Target Cell lines •CCD841CoN •HEK293 •K562 RNA-seq DTE ASGAL A B DEADDEAD Helicase_CHelicase_ C HA2 H A 2 OB_NTP_bind O B_NTP_bin d 0 779aa200 400 600 p.(Val237Glufs*7) DHX40 GAPDH Non-Target A3 A5 ES IR A. Olkinuora et al. 2 Leukemia Table 3, Supplementary Table 4), supporting the idea that loss of function of the constitutionally mutant allele alone was sufficient for tumorigenesis, without a second hit (haploinsufficiency). As no suitable RNA samples from the patients were available for investigations of the consequences of DHX40 inactivation on global transcriptomes, we undertook siRNA experiments on cell lines representing normal (CCD841) or cancer tissues (HEK293 and K562) of different cellular origins (Fig. 1B, Supplementary Materials and Methods), followed by RNA-sequencing. We observed 71 differentially expressed transcripts (q-value < 0.01) unique to DHX40-siRNA-treated cells and shared by all three cell lines (Fig. 1B, Supplementary Fig. 3). Genes functioning in RNA metabolism (RNA helicase, RNA binding, or transcription-related function) were enriched, comprising 31% of the unique transcripts. ASGAL (Alternative Splicing Graph ALigner) directly aligns RNA-seq data to a splicing graph, which results in a list of novel splice events in respect to gene annotation. By ASGAL analysis, DHX40-siRNA-treated cells exhibited a 13% increase in novel splice events, compared to GAPDH-siRNA-treated or non-target siRNA- treated cells (Fig. 1B). DHX40-siRNA treatment associated with elevation of all types of splice events, especially exon skipping. No germline variants suspected pathogenic were detectable in DHX40 or other DEAD/H box genes (from here on, referred to as DDX/DHX genes) in the remaining 27 FCCTX families. However, one family displayed a heterozygous truncating variant in a related RNA helicase gene, TDRD9 (NM_153046.2:c.2261delC, p.Thr754IlefsX11) that co-segregated with colorectal cancer in two siblings. This finding further strengthens the importance of impaired RNA metabolism behind hereditary cancer susceptibility. Next, we ascertained close to 400 patients with acute leukemia, myelodysplastic syndrome, or myeloma with germline and/or somatic exomic data available from a hospital-based repository and analyzed their samples for possibly pathogenic variants in DDX/DHX genes. Sixty-six skin fibroblast samples out of 367 (18%) revealed a possibly pathogenic germline variant in at least one DDX/DHX gene (Fig. 2A). Most variants were of the missense type (Fig. 2B). According to the FinnGen database (finngen.fi), of single nucleotide variants included in their genome-wide association study and located nearby our most frequently affected genes, DDX58,DDX54,DHX38, and DDX10 showed significant (p<10 −4 ) association to several hematological malignancies. Additionally, DDX10 showed significant (p<10 −4 ) association to colon adenocarcinomas, and DDX54 to benign colon tumors. While DHX40 was not affected with any such germline alterations that would fulfill our stringent selection criteria, the common p.M1I pathogenic loss of-function variant of DDX41 [5] was detected twice. One hundred and twenty-two neoplastic bone marrow samples out of 432 (28%) revealed at least one possibly pathogenic somatic DHX/DDX single-nucleotide variant or small indel variant (mostly of the missense type), including DHX40 variants in four (two with AML and two with myeloma) (Fig. 2C, D). The nonsense variant NM_024612.4:c.361G > T, p.E121* was particularly noteworthy being present in two consecutive specimens of CD138+cells taken with a 15-month interval from a patient with multiple myeloma, and having a high variant allele frequency (37%) in both samples. This variant (together with p.S210* identified in an AML patient) affects the DEAD/H box helicase domain and is predicted to give rise to a severely truncated protein (Fig. 2E). In summary, we describe a rare truncating germline variant of DHX40 in a multi-generation family with hematological and solid malignancies. Loss-of-function nature of the variant, cosegregation with neoplasia phenotypes, and functional evidence suggest a role in cancer predisposition. Our findings add DHX40 as a new candidate to the growing list of RNA metabolism-related genes that may underlie predisposition to FCCTX [2,7] and myeloid disorders [4]. Moreover, our findings from a large hospital-based patient series suggest the involvement of the broader DDX/DHX gene family in hematological malignancies. Our DHX40 findings resemble those earlier described for DDX41 in several respects. First, while DDX41 primarily associates with hematological neoplasia, solid malignancies, especially prostate cancer, colorectal cancer, and melanoma, have been reported in individuals with pathogenic germline variants of DDX41 [8]. Second, our patient V.1 from F32 responded well to combination therapy (clofarabin, plerixafor and lenalidomide) and had no need for additional therapy regimens after a second allo-HSCT following the combination therapy; the patient has remained in remission for over 10 years. Interestingly, sensitivity of DDX41-mutated patients to lenalidomide treatment has been described in the literature [4]. Third, although the frameshift nature of the DHX40 germline variant implied loss of function, no apparent “second hits”were detectable in neoplastic tissues from individuals from F32 (Supplementary Table 3), a situation compatible with dominant-negative mechanism or alternatively, haploinsufficiency. Experience from DDX41 suggests haploinsufficiency: although individuals with pathogenic germline variants often acquire somatic DDX41 variants as second hits, the latter occur at low (0–20%) frequencies in bone marrow cells because biallelic alterations are not compatible with proliferating hematopoietic cells [9,10]. It is likewise possible that biallelic DHX40 defects are not tolerated. Interestingly, a duodenum adenoma from F32 showed a mosaic loss of DHX40 protein by immunohistochemical analysis (Supplementary Fig. 1C), suggesting that biallelic defects could be present, but in only a small proportion of tumor cells. Fourth, DDX41-deficient blast cells were shown to exhibit aberrant exon skipping and retention [4] resembling our findings from siRNA-mediated knockdown of DHX40 in cell lines (Fig. 1B). Fig. 1 Genetic and functional characterization of the DHX40 c.710_713delTCAG variant. A Pedigree of F32. Numbers below the symbols are patient identifiers. Arrow denotes the index person. Tumor manifestations and age at diagnosis (years) are given below patient symbols. Nonessential pedigree features have been excluded or modified to protect confidentiality. A plus sign (+) denotes the presence of the variant in a heterozygous state and a minus sign (-) the absence of the variant. Sequence chromatogram on the right displays the normal (top) and altered sequence (bottom), where arrowhead denotes the site of the germline change. Lollipop diagram (bottom) indicates location of the predisposing variant against the main functional domains of the encoded protein. The functional domains of DHX40 are: DEAD, DEAD/H box helicase domain; Helicase_C, Helicase conserved C-terminal domain; HA2, Helicase associated domain; OB_NTP bind, Oligonucleotide/ oligosaccharide-binding (OB)-fold. BConsequences of siRNA-mediated knockdown of DHX40. Workflow of siRNA experiments on CCD841CoN, HEK293, and K562 cell lines is shown on the left. Results from analyses of RNA-sequencing data for differential transcript expression (DTE) and novel splicing events (by the ASGAL tool) (see Supplementary Materials and Methods) are depicted on the right. The bar graph of DTE analysis displays top 20 differentially expressed genes among 71 unique transcripts shared between the three cell lines. Genes whose products participate in RNA metabolism are in bold (see Supplementary Fig. 3 for all 71 genes). The ASGAL analysis shows the number of novel splicing events detected after treatment with DHX40-siRNA, GAPDH-siRNA, or non-target siRNA, vs. untreated cells, and stratified by the type of splicing alteration (A3, alternative 3’site; A5, alternative 5’site; ES, exon skipping; IR, intron retention). Splicing events for all three cell lines (HEK293, K562, and CCD841CoN) were combined. Asterisk denotes statistically significant differences (p< 0.0001 for ES and p< 0.05 for all other events, by pairwise chi-square test with FDR correction) in the number of specific types of splice events after DHX40-siRNA treatment vs. GAPDH-siRNA or non-target siRNA treatment. A. Olkinuora et al. 3 Leukemia At least 59 established DDX/DHX helicases are known [11]. In our hospital-based hematological series, DDX/DHX genes were fairly frequently affected by missense variants and occasionally (7–9%) by truncating variants (Fig. 2A–D) Although several DDX/ DHX genes pinpointed in this series showed significant association to hematological and colorectal neoplasia according to public databases, the true significance of the finding remains to be determined by additional studies. Apart from ATP-dependent RNA-duplex unwinding, DDX/DHX proteins participate in the regulation of long and short noncoding RNAs, pre-mRNA splicing, ribosome biogenesis, and many other cellular functions [11–14]. Besides shedding light to the basic biological mechanisms of neoplasia, our results are clinically relevant since DDX/DHX variant status may guide therapy options ([4] and this study); moreover, participation in translational control makes DDX/DHX helicases attractive targets for novel anti-cancer therapies [15]. Our findings encourage additional investigations on DHX40 as well as studies into other DHX/DDX genes as possible predisposing factors for colorectal and hematological neoplasia. DATA AVAILABILITY Raw RNA-seq data and DTE analysis results from siRNA treated cell lines can be obtained from GSE228991. Our IRB approvals do not allow sharing of raw sequencing data from patients. All variants fulfilling our filtering criteria can be found in the Supplementary Files. Requests to access additional datasets should be directed to the corresponding authors. REFERENCES 1. Huang KL, Mashl RJ, Wu Y, Ritter DI, Wang J, Oh C, et al. Pathogenic germline variants in 10,389 adult cancers. Cell. 2018;173:355–70.e314. E121* R135C DEADDEAD Helicase_CHelicase _ CHA2HA 2 OB_NTP_bindOB _ NTP _ bind 0 5 splice R135C A175T V237fs L292P P317L W572L C573R R688H R712*/Q 0779aa200 400 600 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 1111 12 12 131 3 14 1 415 15 161 6 1717 181 8 E121* R135C S210* G281D Y468F D474YV237fs 5 0 Colorectal cancer Myeloid and lymphoid neoplasia Missense Splice Exon Number of germline variants per gene (total n = 69) Number of somac variants per gene (total n = 320) 0 2 4 6 8 10 12 14 16 18 20 DHX36 DDX58 DDX10 DHX57 DHX37 DHX9 DDX60 DDX26B DHX29 DDX41 DDX20 DHX33 DDX3Y DHX34 DDX4 DDX3X DDX59 INTS6 DDX52 DDX25 DDX5 DHX35 DDX50 DDX27 DDX53 DHX40 DHX15 DDX17 DDX56 DDX1 DDX42 DDX28 DDX54 EIF4A1 EIF4A3 DDX24 DHX16 DDX55 DHX38 DDX31 DDX6 DDX21 DDX49 DDX39A DDX23 DDX18 DDX46 DDX19A DDX43 DHX32 DDX12P EIF4A2 DDX51 DDX39B DDX19B DHX30 DDX47 Missense Truncang Somac exomes Invesgated With DDX/DHX variants With DHX40 variants AML 159 44 (28%) 2 (1%) ALL 46 13 (28%) 0 MDS 17 4 (24%) 0 Myeloma 121 44 (36%) 2 (2%) MPN 2 0 0 other/combined 87 13 (15%) 0 Total 432 122 (28%) 4 (1%) Number of paents Germline exomes Invesgated With DDX/DHX variants With DHX40 variants AML 166 31 (19%) 0 ALL 52 10 (19%) 0 MDS 19 4 (21%) 0 Myeloma 107 19 (18%) 0 MPN 23 2 (9%) 0 Total 367 66 (18%) 0 (0%) Number of paents AB CD D 0 1 2 3 4 5 6 7 DDX58 DDX54 DHX38 DDX10 DDX59 DDX52 DHX16 DDX20 DDX49 DDX3Y DHX36 DDX50 DHX37 DDX11 DDX28 DHX34 DHX57 DDX41 INTS6 DDX1 DDX51 DHX9 DDX23 DDX19B PMFBP1 DDX42 DHX35 DHX29 DDX56 DDX25 DDX47 DDX55 DHX8 EIF4A1 Missense Truncang E DHX40 variantsDHX40 variants Fig. 2 Prevalence of DDX/DHX variants in hematological neoplasia and summary of DHX40 variants in colorectal and hematological neoplasia. A,BGermline exomes (fibroblasts) from our hospital-based hematological series. Variant selection criteria were the same as for FCCTX, except that in silico evaluations were based on five programs instead of six. Bar chart in 3B includes DDX/DHX genes affected by at least one variant. See Supplementary Table 5 for variant details. C, D Somatic exomes (leukemic bone marrow) from our hospital-based hematological series. Nonsynonymous variants with allele frequency (VAF) 1% or higher and somatic pvalue < 0.01 by VarScan2 were selected. The “other/combined”group in 3C consists of mainly combined diagnoses of different hematological lineages. The four cases of somatic DHX40 variants listed in 3C include p.S210*, VAF 7% (AML_3 with recurrent AML); p.Y468F, VAF 7% (AML_43 with polycythemia vera and subsequent acute promyelocytic leukemia); p.D474Y, VAF 7% (MM_14 with multiple myeloma); and p.E121*, VAF 37% (MM_20 with recurrent multiple myeloma). Bar chart in 3D includes DDX/DHX genes affected by at least one variant. See Supplementary Table 6 for variant details. ELollipop diagram of germline (triangle) and somatic (circle) variants in DHX40 present in colorectal, myeloid, and lymphoid neoplasia. Variants shown in red font are reported in this article, and variants in black are from publicly available datasets in cBioPortal (see Supplementary Materials and Methods). The germline variant p.V237Efs comprised both colorectal and myeloid/lymphoid phenotypes (Fig. 1A). See legend for Fig. 1A for functional domains of DHX40 (the corresponding exons of DHX40 are underneath the diagram). A. Olkinuora et al. 4 Leukemia 2. Nieminen TT, O’Donohue M-F, Wu Y, Lohi H, Scherer SW, Paterson AD, et al. Germline mutation of rps20, encoding a ribosomal protein, causes predisposition to hereditary nonpolyposis colorectal carcinoma without dna mismatch repair deficiency. Gastroenterology. 2014;147:595–8.e595. 3. Bhar S, Zhou F, Reineke LC, Morris DK, Khincha PP, Giri N, et al. Expansion of germline RPS20 mutation phenotype to include Diamond-Blackfan anemia. Hum Mutat. 2020;41:1918–30. 4. Polprasert C, Schulze I, Sekeres MA, Makishima H, Przychodzen B, Hosono N, et al. Inherited and somatic defects in DDX41 in myeloid neoplasms. Cancer Cell. 2015;27:658–70. 5. Quesada AE, Routbort MJ, DiNardo CD, Bueso-Ramos CE, Kanagal-Shamanna R, Khoury JD, et al. DDX41 mutations in myeloid neoplasms are associated with male gender, TP53 mutations and high-risk disease. Am J Hematol. 2019;94:757–66. 6. Hong M, He G. The 2016 revision to the world health organization classification of myelodysplastic syndromes. J Transl Int Med. 2017;5:139–43. 7. Bellido F, Sowada N, Mur P, Lázaro C, Pons T, Valdés-Mas R, et al. Association between germline mutations in BRF1, a subunit of the RNA polymerase III transcription complex, and hereditary colorectal cancer. Gastroenterology. 2018;154:181–94.e120. 8. Churpek JE, Smith-Simmer K. DDX41-associated familial myelodysplastic syndrome and acute myeloid leukemia. In: Margaret P Adam, Ghayda M Mirzaa, Roberta A Pagon, Stephanie E Wallace, Lora JH Bean, Karen W Gripp, Anne Amemiya, editors. GeneReviews.University of Washington: Seattle; 2021. 9. Chlon TM, Stepanchick E, Hershberger CE, Daniels NJ, Hueneman KM, Kuenzi Davis A, et al. Germline DDX41 mutations cause ineffective hematopoiesis and myelodysplasia. Cell Stem Cell. 2021;28:1966–81.e1966. 10. Makishima H, Saiki R, Nannya Y, Korotev S, Gurnari C, Takeda J, et al. Germ line DDX41 mutations define a unique subtype of myeloid neoplasms. Blood. 2023;141:534–49. 11. Cai W, Xiong Chen Z, Rane G, Satendra Singh S, Choo Z, Wang C, et al. Wanted DEAD/H or alive: helicases winding up in cancers. J Natl Cancer Inst. 2017;109:djw278. 12. Rocak S, Linder P. DEAD-box proteins: the driving forces behind RNA metabolism. Nat Rev Mol Cell Biol. 2004;5:232–41. 13. Liu YC, Cheng SC. Functional roles of DExD/H-box RNA helicases in Pre-mRNA splicing. J Biomed Sci. 2015;22:54. 14. Russon MP, Westerhouse KM, Tran EJ. Transcription, translation, and DNA repair: new insights from emerging noncanonical substrates of RNA helicases. Biol Chem. 2021;402:637–44. 15. Heerma van Voss MR, van Diest PJ, Raman V. Targeting RNA helicases in cancer: the translation trap. Biochim Biophys Acta Rev Cancer. 2017;1868:510–20. ACKNOWLEDGEMENTS We thank Saila Saarinen for expert laboratory assistance and Henrikki Almusa and Samuli Eldfors for contributing to VarScan2 analyses. We are grateful to Kirsi Pylvänäinen, Tuula Lehtinen, Beatriz Alcala-Repo, and Maija Röntynen for collecting clinical data. AUTHOR CONTRIBUTIONS Conceptualization: AO, TTN, SD, OK, UW-K, KP, and PP. Data curation: JV and MKa. Formal analysis: AO, TTN, SD, AK, JV, and MKa. Funding acquisition: MKo, AR, MM, CH, J-PM, OK, UW-K, KP, and PP. Investigation: AO, TTN, SD, AK, MKa, JV, MKo, CAH, JS, MS, OK, UW-K, AR, AL, KP, and PP. Methodology: AO, TTN, SD, AK, and PL. Project administration: HJ, AR, MM, J-PM, OK, UW-K, KP, and PP. Resources: AR, MM, CAH, HJ, J-PM, AL, OK, UW-K, KP, and PP. Software: AO, JV, and MKa. Supervision: AO, TTN, AR, CAH, OK, UW-K, KP, and PP. Validation: AO, TTN, and SD. Visualization: AO and PP. Writing—original draft: AO, TTN, and PP. Writing—review & editing: AO, TTN, SD, AR, OK, UW-K, KP, PP. All authors approved the final version of the manuscript. FUNDING This work was supported by grants from Biomedicum Helsinki Foundation (to AO); the Jalmari and Rauha Ahokas Foundation (to TTN); the Maud Kuistila Memorial Foundation (to TTN); the Finnish Medical Foundation (to MKo); the Finska Läkaresällskapet (to AR); Medicinska Understödsföreningen Liv & Hälsa (to AR); the Jane and Aatos Erkko Foundation (to PP); the Academy of Finland (grant number 330606 to PP); Cancer Foundation Finland sr (to AO, MKo, AR, OK, and PP); and Helsinki University Central Hospital Research Funds (to MKo and AR); and the Sigrid Juselius Foundation (to AR, OK, and PP). The Doctoral Program in Biomedicine of Helsinki University offered a paid doctoral student position to AO. Open Access funding provided by University of Helsinki including Helsinki University Central Hospital. COMPETING INTERESTS The authors declare no competing interests. ADDITIONAL INFORMATION Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41375-023-02021-9. Correspondence and requests for materials should be addressed to Alisa Olkinuora, Taina T. Nieminen, Kimmo Porkka or Päivi Peltomäki. Reprints and permission information is available at http://www.nature.com/ reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by/4.0/. © The Author(s) 2023 A. Olkinuora et al. 5 Leukemia