Frequent and clinically relevant germline DNA repair gene variants in young and familial myeloproliferative neoplasms

Myeloproliferative neoplasms (MPNs) comprise a group of hematological disorders characterized by overproduction of blood cells caused by somatic driver mutations. While most MPN cases arise sporadically, familial clustering is increasingly recognized, occurring in around 7% of cases. Within affected families, different driver mutations (JAK2, CALR, MPL) and disease subtypes—including polycythemia vera, essential thrombocythemia, and myelofibrosis —can be observed [1–3]. Several germline variants increasing susceptibility to MPNs have been identified, including variants in JAK2 (‘46/1’ haplotype; 4-fold increased MPN risk) [4], TERT (2-fold), SH2B3, TET2, CHEK2, ATM, PINT, and GFI1B [3, 5–7]. However, the underlying mechanisms by which these germline variants contribute to the acquisition of the somatic drivers remain unclear.

In a previous study, we performed whole-exome sequencing (WES) in familial cases of MPN and identified germline variants in DNA repair genes in 80% of cases [8]. Due to the limited sample size, however, it remained uncertain whether these germline variants were causally linked to the development of familial MPN.

Here, we investigated whether young age (<30 years) at driver-mutation-positive MPN onset and/or occurrence of familial MPN (at least one first-degree relative also affected) may increase the likelihood of detecting clinically relevant germline variants in DNA repair genes.

Whole exome (WES) or whole genome sequencing (WGS) was performed in 48 individuals who met at least one criterion to screen for variants in DNA repair genes leading to genetic tumor risk syndromes (TRS), with hair roots used as germline control (Supplementary Table 1). Three patients fulfilled both inclusion criteria.

In the cohort of patients diagnosed with MPN before the age of 30 years, 21 patients were recruited. The average age of onset was 22.9 years (range: 12–28, median: 24). A (likely) pathogenic germline variant was detected in five of these patients (5/21, 24%): three variants in CHEK2 and one variant each in BRIP1 and ATM (Table 1).

Table 1.

Pathogenic/likely pathogenic germline variants identified in the study cohort and control cohort.

Family Nr.	Pat. Nr.	Gene	Variant	Zygosity	Classification	Young	Familial	MPN subtype	MPN driver mutation carrying gene
Study cohort
F2	P3	CHEK2	NM_007194.4:c.1100del, p.(Thr367Metfs*15)	Het.	Pathogenic		X	Post-ET-MF	JAK2
F2	P4	CHEK2	NM_007194.4:c.1100del, p.(Thr367Metfs*15)	Het.	Pathogenic	X	X	PV	JAK2
F3	P5	BRCA1	NM_007294.3:c.2722G > T, p.(Glu908*)	Het.	Pathogenic		X	PV	JAK2
F3	P6	BRCA1	NM_007294.3:c.2722G > T, p.(Glu908*)	Het.	Pathogenic		X	PMF	JAK2
F9	P15	CHEK2	NM_007194.4:c.349A > G, p.(Arg117Gly)	Het.	Pathogenic		X	PV	JAK2
F18	P30	CHEK2	NM_007194.4:c.1100del, p.(Thr367Metfs*15)	Het.	Pathogenic		X	PMF	JAK2/MPL
		MSH6	NM_000179:c.467C > G, p.(Ser156*)	Het.	Pathogenic
F28	P40	BRIP1	NM_032043.3: c.2990_2993del, p.(Thr997Argfs*61)	Het.	Likely pathogenic	X		ET	JAK2
F31	P43	ATM	NM_000051.2:c.2921+1G > A, p.(Tyr947Glnfs*9)	Het.	Pathogenic	X		MPN-U (ET/PMF)	JAK2
F32	P44	CHEK2	NM_007194.4:c.1283C > T, p.(Ser428Phe)	Het.	Likely pathogenic	X		Post-ET-MF	JAK2
F34	P46	CHEK2	NM_007194.4:c.1100del, p.(Thr367Metfs*15)	Het.	Pathogenic	X		ET	JAK2
Control cohort
	K43	CHEK2	NM_007194.4:c.1100del, p.(Thr367Metfs*15)	Het.	Pathogenic			ET	CALR

Het heterozygous, ET essential thrombocythemia, MF myelofibrosis, MPN-U unclassifiable MPN, PMF primary myelofibrosis, PV polycythemia vera.

In the familial MPN cohort, we included 30 patients from 18 families (12 pairs of relatives affected by MPN and 6 single cases). A pathogenic germline variant was detected in six patients from four families (4/18 families, 22%): three variants in CHEK2 and one variant in BRCA1, while one of the CHEK2 variant carriers also harbored a pathogenic MSH6 variant (Table 1).

Altogether, across the entire cohort of young and familial MPN patients, we identified pathogenic or likely pathogenic variants in ten patients (10/48, 21%) from eight families (Table 1).

For comparison, we analyzed 52 unselected, driver-mutation-positive MPN patients (>30 years at onset; mean 55.0, median: 55, range: 34–81; no familial MPN) by WES (Supplementary Table 1). In this group, only one pathogenic germline DNA-repair variant (CHEK2) was detected (1/52, 2%) (Table 1) in a patient with a family history of solid tumors. However, the CHEK2 variant did not co-segregate with tumor cases in the family.

All variants identified in our study are considered actionable, with direct implications for patients and their families [9]. For many TRS, established prevention and surveillance programs exist. Clinical guidelines recommend genetic testing when the pre-test probability of detecting a pathogenic TRS-causing variant exceeds 10%, based on defined criteria that currently do not include MPNs [10, 11]. In our study, the detection rate in both young-onset and familial MPN clearly exceeded this threshold, supporting the rationale for offering germline testing to these subgroups. Recently, an analysis of the UK Biobank reported similar germline DNA-repair variants (e.g., CHEK2, ATM) in patients with clonal hematopoiesis of undetermined potential (CHIP), who showed an increased risk of myeloid malignancies compared with CHIP patients without such mutations [12].

In contrast, pathogenic variants in tumor-predisposition genes were considerably less frequent in the control cohort. Although sample size was limited and no causal link to MPN development can be established, our data support the use of simple clinical criteria to identify MPN patients who may benefit from genetic counseling and germline testing.

Interestingly, 10 of 11 germline variant carriers harbored the JAK2V617F driver mutation (and only one a CALR mutation), raising the possibility of a mechanism that preferentially supports this mutation or clone. Future studies are needed to clarify the underlying cause.

Further, we sought to elucidate whether the observed heterozygous germline variants have functional relevance in MPN. BRCA1, BRIP1, ATM, and CHEK2 encode proteins involved in the repair of DNA double-strand breaks (DSB) via homologous recombination (HR). In various solid tumors, the use of poly ADP-ribose polymerase (PARP) inhibitors, such as olaparib, has been approved in cases of identification of HR deficiency (HRD). Despite the absence of additional somatic mutations (i.e., second hits) in DNA repair genes that would result in complete HRD in our patient cohort, evidence has emerged that even heterozygous variants (haploinsufficiency) manifest altered physiology under stress conditions [13]. In previous work, we tested the impact of a Brca1 deletion in a murine MPN cell line model [14] and found that heterozygous loss of Brca1 in Jak2V617F-overexpressing 32D cells rendered these cells more sensitive to DSB and olaparib-induced apoptosis, along with impaired Rad51 foci formation, demonstrating defective HR.

Functional analyses of ATM and CHEK2 germline variants identified in families with hematologic malignancies have also been reported [3, 7, 15]. In particular, Bao et al. demonstrated that CHEK2 loss-of-function or inhibition promotes self-renewal of human hematopoietic stem and progenitor cells [15], which was supported by a knock-in mouse model generated by Stubbins et al., showing hematologic malignancy as well as increased cKit+ stem and progenitor cells in aged mice [7]. During the course of our analysis, a germline ATM L2307F variant found in familial MPN was described to impair ATM-mediated phosphorylation of CHEK2 in CRISPR-engineered myeloid cells, indicating partial loss of ATM function [3]. However, these analyses were performed in transformed cells and did not address the role of ATM variants in hematopoietic stem or progenitor cell biology.

To further investigate the clinical relevance of germline haploinsufficiency of DNA repair genes in JAK2V617F-driven MPN, we analyzed the impact of a heterozygous germline ATM variant (c.2921+1G > A; ATM^WT/mut, hereafter ATM^mut) identified in one patient with unclassifiable MPN (MPN-U) with features of essential thrombocythemia, prefibrotic myelofibrosis, and hypersplenism (Supplementary Fig. 1). We reprogrammed patient-derived hematopoietic cells into induced pluripotent stem cells (iPSCs) and established two isogenic lines carrying either JAK2^WT/V617F (JAK2^VF) or JAK2^WT/WT (JAK2^WT). As controls, we used previously established JAK2^WT and JAK2^VF iPSCs from another MPN patient with germline ATM^WT. This design enabled us to compare four genetically defined conditions combining germline (ATM^mut vs. ATM^WT) and somatic (JAK2^VF vs. JAK2^WT) variants (Fig. 1A).

Fig. 1. Characterization of ATM^WT and ATM^mut induced pluripotent stem cells (iPSCs).

A Schematic overview of PBMC reprogramming and iPSC clone isolation. PBMCs from a JAK2V617F-positive MPN patient carrying a heterozygous germline ATM mutation (c.2921+1G > A) were reprogrammed into iPSCs. Single-cell cloning produced iPSC lines with either JAK2V617F or JAK2WT. As ATM wild-type controls, we used previously established iPSC lines from another JAK2V617F-driven MPN patient (one JAK2V617F and one JAK2WT). B Western blot analysis of ATM protein levels in ATM^WT and ATM^mut iPSCs. Quantification is shown as mean ± SD, n = 4. Statistical analysis was performed using one-way ANOVA followed by Tukey’s test (*p < 0.05, **p < 0.01). C Percentage of γH2AX+ cells in hematopoietic cells derived from indicated iPSCs. γH2AX+ cells were quantified by FACS after 48 h of treatment with or without 5 µM olaparib. Data are presented as mean ± SD, n = 4. Statistical analysis was performed using two-way ANOVA followed by Tukey’s test (**p < 0.01, ***p < 0.001, ****p < 0.0001). D Hematopoietic cell production from indicated iPSCs. Cells obtained after EB-based differentiation were counted and enriched for CD34 using MACS. Data are shown as mean ± SD, n = 4–5. Statistical analysis was performed using one-way ANOVA followed by Tukey’s test (*p < 0.05, **p < 0.01, ****p < 0.0001).

Western blot analysis confirmed reduced ATM protein levels in ATM^mut iPSC lines (Fig. 1B). Olaparib-mediated PARP inhibition induced DNA double-strand breaks (DSBs) across all four genotypes, as evidenced by increased γH2AX-positive cells (Fig. 1C). Notably, the ATM^mut JAK2^VF combination displayed the highest levels of DSBs compared with the other three genotypes. In particular, ATM^mut JAK2^VF cells showed significantly more DSBs than ATM^WT JAK2^VF cells, indicating that acquisition of the JAK2V617F oncogene in the context of germline ATM haploinsufficiency amplifies DNA damage and reveals a synthetic vulnerability to PARP inhibition.

Next, we investigated the effects of the ATM variant, alone or in combination with JAK2V617F, on hematopoiesis. The double-mutant iPSCs (ATM^mut JAK2^VF) produced higher numbers of hematopoietic cells than either unmutated (ATM^WT JAK2^WT) or single-mutant (ATM^mut JAK2^WT, ATM^WT JAK2^VF) iPSCs (Fig. 1D). Consistently, the double-mutant iPSCs displayed a marked reduction in CD34+ progenitors compared with the three controls (Fig. 1D and Supplementary Fig. 1G). Together, these results suggest that ATM haploinsufficiency enhances the oncogenic effects of JAK2V617F by promoting proliferation, accelerating differentiation, and reducing stemness of hematopoietic stem cells [16]. Given the results in the present models, future work will aim to expand these models to cover additional DNA repair genes, including CHEK2 and BRIP1.

In conclusion, germline analyses in young and familial MPN patients represent a valuable strategy to identify functionally relevant and clinically actionable genetic alterations, with a detection rate that clearly surpasses the commonly recommended 10% threshold for germline testing. This approach not only provides important insights into disease biology but also has the potential to improve diagnostic evaluation, genetic counseling, and therapeutic decision-making. Beyond the diagnostic value, our iPSC-based analyses suggest that germline haploinsufficiency of DNA repair genes may amplify the oncogenic effects of JAK2V617F and create synthetic vulnerabilities to PARP inhibition. As PARP inhibitors are already approved in HR-deficient solid tumors, our findings raise the possibility that selected MPN patients—particularly those with germline DNA repair variants—might also benefit from such targeted therapies. Extending this rationale, a combination of approved MPN-directed drugs such as JAK inhibitors or peginterferon-α with PARP inhibitors may be particularly well-suited for these patients. Notably, given that HRD may also be induced by JAK inhibitor therapy itself [17], a first clinical trial (NCT06218628) is currently testing the combination of the JAK inhibitor pacritinib with the PARP inhibitor talazoparib in patients with sporadic MPNs. Together, these findings underscore the importance of integrating germline genetics into the routine evaluation of young and familial MPN patients to refine stratification, guide follow-up, and pave the way toward more personalized treatment options.

Author contributions

RM, MJR, ME, LLT, and SK designed the research, RM, MJR, MC KK, MB, FK, IS, DD, NG, KB, JB, MAST, SI, UG, SA, THB, IK, ME, LLT, and SK contributed and analyzed the data, RM, MJR, MB, FK, JB, and MAST performed research, RM, MJR, THB, IK, ME, LLT, and SK wrote the paper, RM, MJR, MC, KK, MB, FK, IS, DD, NG, KB, JB, MAST, SI, UG, SA, THB, IK, ME, LLT, and SK approved and revised the paper.

Funding

“This work was supported in part by funds from the German Cancer Aid (Deutsche Krebshilfe; DKH) to S.K. (70114726) and the Center of Integrated Oncology Aachen Bonn Cologne Düsseldorf (CIO ABCD), Germany”.

Data availability

The authors declare that all data supporting the findings of this study are available within the paper and its supplementary information files. In accordance with data protection requirements, we do not have the option of making the genetic data publicly available. However, the data can be shared upon request and with appropriate legal safeguards.

Competing interests

SK has received research grant/funding from Geron, Janssen, AOP Pharma, and Novartis; received consulting fees from AbbVie, Pfizer, Incyte, Ariad, Novartis, AOP Pharma, Bristol Myers Squibb, Celgene, Geron, Janssen, CTI BioPharma, Roche, Bayer, Protagonist, and PharmaEssentia; received payment or honoraria from AbbVie, Pfizer, Incyte, Ariad, Novartis, AOP Pharma, Bristol Myers Squibb, Celgene, Geron, Janssen, CTI BioPharma, Roche, Bayer, Protagonist, PharmaEssentia, Astra Zeneca, iOMEDICO, GSK; received travel/accommodation support from Alexion, Novartis, Bristol Myers Squibb, Incyte, AOP Pharma, CTI BioPharma, Pfizer, Celgene, Janssen, Geron, Roche, AbbVie, GSK, Sierra Oncology, Protagonist, Karthos, iOMEDICO, and Astra Zeneca; had a patent issued for a BET inhibitor at RWTH Aachen University; participated on advisory boards for Pfizer, Incyte, Ariad, Novartis, AOP Pharma, BMS, Celgene, Geron, Janssen, CTI BioPharma, Roche, Bayer, GSK, Sierra Oncology, AbbVie, Protagonist, and PharmaEssentia. The other authors declare that they do not have a conflict of interest.

Ethics approval and consent to participate

All methods were performed in accordance with the relevant guidelines and regulations. This study was approved by the ethics committee of the Medical Faculties of RWTH Aachen University (EK302/16, EK127/12) and the University of Bonn (EK 212/16) and was conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants.

Supplementary information

The online version contains supplementary material available at 10.1038/s41408-025-01433-3.