Novel germline JAK2 L808W and E305Vfs*4 mutations in familial essential thrombocythemia: Evidence from two pedigrees

The classical Philadelphia chromosome-negative myeloproliferative neoplasms (MPNs), including polycythemia vera, essential thrombocythemia (ET), and primary myelofibrosis, are clonal hematopoietic disorders driven by somatic mutations in JAK2 (most notably V617F), CALR, or MPL, all of which converge on constitutive activation of the JAK-STAT pathway (1). Triple-negative essential thrombocythemia (TN-ET) is a distinct subtype of ET, accounting for approximately 10%−20% of all ET cases. These cases lack MPN driver mutations but still exhibit histopathological and clinical features sufficient for the diagnosis of ET. A subset of TN-ET cases shows familial predisposition and is caused by germline mutations in JAK2 and MPL (2). Unlike the common somatic V617F mutation, these germline variants are often atypical, located outside classic mutational hotspots, and display variable penetrance (3).

Beyond canonical gain-of-function mutations, a recent paradigm-shifting study reported a germline JAK2 N533S mutation that alone failed to enhance proliferation or activate JAK-STAT signaling in vitro. Its pathogenicity emerged only upon acquiring a co-occurring somatic CALR mutation, where the double-mutant clone gained a significant competitive advantage in vivo (4). This finding introduced a novel “permissive” or “foundational” model of pathogenesis, wherein a germline variant creates a susceptible cellular background that facilitates clonal outgrowth following secondary hits. Here, we report two novel germline JAK2 mutations, L808W in the pseudokinase (JH2) domain and E305Vfs*4 in the band 4.1, ezrin, radixin, moesin (FERM) domain, identified in two independent families with ET. Through genetic and functional analyses, we investigated their distinct pathogenic mechanisms.

During routine clinical practice, we encountered two patients diagnosed with ET. After identifying that the mutations they carried were germline in origin, we systematically collected medical histories from their families. Peripheral blood and fingernail DNA from two ET pedigrees were subjected to next-generation sequencing and Sanger validation to confirm the germline origin of the JAK2 mutations. Pathogenicity was predicted using in silico tools (MutationTaster, PolyPhen2, and SIFT) and AlphaFold structural modeling. Full-length JAK2 L808W and E305Vfs*4 cDNAs were cloned and inserted into lentiviral vectors and transduced into HEL cells. Cell proliferation was assessed by a CCK-8 assay following IL-3 withdrawal. JAK-STAT pathway activation was evaluated by Western blotting for phospho-STAT5.

After excluding canonical JAK2 V617F, CALR, and MPL mutations by conventional testing, next-generation sequencing targeting 138 MPN-related genes was performed on bone marrow DNA from the probands. This analysis revealed two novel heterozygous JAK2 variants. In Family A, a c.2423T>G transversion in exon 19 resulted in a leucine to tryptophan substitution at codon 808 (p.L808W), with a variant allele frequency (VAF) of 48.03%, located within the pseudokinase (JH2) domain. In Family B, a c.914_917delAAAG deletion in exon 8 produced a frameshift and premature stop codon (p.E305Vfs*4; VAF 48.71%) within the FERM domain. Notably, the proband carrying JAK2 L808W also harbored concurrent somatic mutations in DNMT3A (VAF 1.03%) and TET2 (VAF 45.32%), whereas no additional driver mutations were identified in the E305Vfs*4 proband. To confirm the germline origin, we performed Sanger sequencing on DNA extracted from fingernail clippings of the probands and available family members. This analysis demonstrated perfect segregation of the respective JAK2 mutations with the thrombocytosis phenotype across both pedigrees, with all 10 affected individuals carrying the mutation and no unaffected family members testing positive (Figure 1). The clinical characteristics of affected individuals are summarized in Supplementary Table S1. In Family A, ages ranged from 18 to 52 years, with platelet counts between 388 and 511×10³/μL; in Family B, ages ranged from 20 to 80 years, with platelet counts between 389 and 650×10³/μL. No splenomegaly or hemorrhagic events were reported. Only individual B4 experienced a cerebrovascular event at 52 years of age, occurring in the context of ET with concomitant cardiovascular risk factors, including obesity and diabetes.

Figure 1.

Pedigree and genetic analysis. (A,B) The pedigree of the family with the JAK2 L808W and E305Vfs*4 mutations; (C,D) Sanger sequencing charts of heterozygous JAK2 L808W and E305Vfs*4 using fingernail samples at diagnosis.

To assess the potential functional impact of these mutations, we performed in silico pathogenicity prediction and structural modeling. For JAK2 L808W, prediction tools yielded discordant results: MutationTaster classified the variant as benign, whereas MetaDome, PolyPhen-2, and SIFT consistently predicted a damaging effect (Supplementary Table S2). Multiple sequence alignment demonstrated that the L808 residue is evolutionarily conserved across humans, pigs, mice, rats, and sheep (Supplementary Figure S1C), supporting its functional importance. Structural modeling based on the JAK2 pseudokinase-kinase domain template localized L808 to the JH2 domain, where its side chain was involved in hydrophobic interactions with leucine 780 (L780), potentially contributing to JH2 conformational stability. Substitution with tryptophan, a bulkier and more hydrophobic residue, may strengthen these interactions, possibly stabilizing the JH2 domain (Supplementary Figure S1D). In contrast, the JAK2 E305Vfs*4 frameshift deletion is predicted to introduce a premature termination codon after four altered amino acids, resulting in a severely truncated protein that retains only the N-terminal FERM domain (Supplementary Figure S1E). This truncation abolishes the entire JH2 and kinase domains, suggesting a potential dominant-negative or gain-of-function mechanism through aberrant FERM-mediated interactions.

To experimentally validate these predictions, we cloned full-length JAK2 L808W and E305Vfs*4 cDNAs into a lentiviral vector co-expressing GFP and a puromycin resistance gene. HEL cells, which endogenously express JAK2 V617F and served as a positive control, were transduced with lentiviruses carrying JAK2 L808W or JAK2 E305Vfs*4. Following puromycin selection, >90% GFP positivity was confirmed by flow cytometry (Supplementary Figure S2A), indicating efficient transduction. We then assessed cytokine-independent proliferation by withdrawing IL-3 and measured cell viability daily via a CCK-8 assay. Cells expressing JAK2 E305Vfs*4 exhibited robust IL-3-independent proliferation, comparable to JAK2 V617F-positive controls, whereas cells expressing JAK2 L808W showed no significant increase in proliferation and behaved similarly to those expressing wild-type JAK2 (Supplementary Figure S2B). To evaluate activation of the JAK-STAT pathway, we deprived cells of serum and cytokines for 48 h, and STAT5 phosphorylation was assessed by Western blotting. Consistent with the proliferation results, JAK2 E305Vfs*4 and V617F induced strong STAT5 phosphorylation, whereas only basal levels were observed in cells expressing JAK2 L808W or wild-type JAK2 (Supplementary Figure S2C). These functional results demonstrate that JAK2 E305Vfs*4 acts as a potent autonomous driver, whereas JAK2 L808W lacks intrinsic transforming activity and likely requires cooperative mutations for pathogenicity.

We identified two novel germline JAK2 mutations, L808W and E305Vfs*4, in independent essential thrombocythemia pedigrees that drive disease through fundamentally distinct mechanisms. The JAK2 L808W mutation, located in the pseudokinase (JH2) domain, did not exhibit intrinsic transforming capacity in vitro and failed to confer cytokine-independent proliferation or STAT5 phosphorylation. This functional profile closely resembles that of the recently reported germline JAK2 N533S mutation, which, similarly, lacks autonomous activity but enables clonal expansion upon acquisition of a secondary somatic mutation (4). Notably, the proband harboring L808W carried concurrent somatic mutations in TET2 (VAF 45.32%) and DNMT3A (VAF 1.03%), both of which are epigenetic regulators frequently co-mutated in MPNs (5). In a mouse model, TET2 deficiency confers a significant clonal advantage to JAK2-V617F-mutant hematopoietic stem cells (HSCs), ultimately promoting disease progression (6). Therefore, we propose that the JAK2 L808W mutation subtly alters the cellular signaling milieu, establishing a permissive state wherein the subsequent acquisition of a high-burden TET2 mutation synergistically drives overt ET pathogenesis. This mechanism may explain why a subset of familial MPNs lack canonical driver co-mutations.

In stark contrast to JAK2 L808W, JAK2 E305Vfs*4 exhibited potent gain-of-function activity, conferring IL-3-independent proliferation and constitutive STAT5 phosphorylation. This frameshift mutation produces a truncated protein retaining only the N-terminal FERM domain. The FERM domain is critical for JAK2 binding to cytokine receptors and for maintaining its autoinhibited conformation (3). The truncated FERM-only protein acts through a dominant-negative mechanism, in which it may aberrantly occupy receptor binding sites or engage in abnormal interactions with the FERM domain of wild-type JAK2. Such interactions could disrupt normal receptor-JAK2 complex assembly, leading to constitutive pathway activation (7). This mechanism positions E305Vfs*4 as a strong autonomous driver mutation.

Together with previously reported cases, our findings delineate a continuous functional spectrum of germline JAK2 mutations, ranging from strong autonomous drivers [e.g, E305Vfs*4, L604F (8), R715T (9)] to weak permissive variants requiring secondary hits [e.g, N533S (4), L808W], with intermediate-function mutations such as R564Q occupying the middle ground. This spectrum has clinical implications. For patients with permissive mutations such as L808W, disease manifestation may depend critically on the acquisition of secondary mutations, warranting regular molecular monitoring of clonal evolution. In our L808W proband, the high VAF of TET2 (45.32%) suggests that it may represent the key cooperative event; longitudinal tracking could inform risk stratification. Conversely, strong driver mutations such as E305Vfs*4 may confer enhanced sensitivity to JAK inhibitors, suggesting potential therapeutic stratification. These findings highlight the importance of comprehensive genetic characterization-including full JAK2 sequencing and assessment of epigenetic modifier genes-in MPN patients, particularly those with a young age at onset, a family history, or triple-negative status.

Finally, previous studies have shown that TN-ET follows an indolent clinical course, with longer overall survival and thrombosis-free survival. This favorable outcome can be attributed to clinical observations that patients with TN-ET are younger than those with JAK2 V617F mutations and exhibit significantly lower scores for platelet activation, apoptosis, and reactive oxygen species (ROS) pathways than patients with JAK2 V617F-mutated ET (10). This finding may explain why only one of our patients, who presented with multiple risk factors, experienced a thrombotic event.

These findings deepen our understanding of the genetic complexity of familial MPNs and underscore the core value of integrating precise genotyping with functional elucidation to advance personalized therapies.

SUPPLEMENTARY DATA

Supplementary data to this article can be found online.

Funding Statement

This work was supported by Peking University International Hospital Research Grant (No. YN2023QN01).

Author contributions

Conceptualization: JX Liu, WQ Hu; Methodology: JX Liu, CJ Wang; Investigation: YY Zhang, M Ouyang; Data curation: JX Liu; Writing – original draft: JX Liu, MX Zhang; Writing – review & editing: JX Liu, MX Zhang; Supervision: MX Zhang, HY Ren; Funding acquisition: JX Liu.