SH2B3 Loss of Function Variants Are Potential Drivers of Severe Thrombocytosis

Introduction: LNK/SH2B3 adaptor protein is an important negative regulator of cytokine signaling pathways in the hematopoietic and vascular system. The identification of acquired and germline SH2B3 mutations alone or in co-existence with known driver mutations in myeloid and lymphoid malignancies, autoimmune and cardiovascular disorders confirmed LNK implication in these neoplastic and chronic inflammatory diseases (Maslah, N., et al., Leukemia, 2017). Platelet dysfunction and megakaryocyte hyperplasia are two biological features of Sh2b3 KO mice; the latter phenotype being recently observed in two SH2B3 TN MPN cases (Zhang, X., et al., AJH, 2024). However, association of SH2B3 loss of function (LoF) variants with thrombocytosis and thrombotic clinical traits, in the absence of a driver mutation, has not yet been reported. Our aim was to identify LoF SH2B3 variants associated with TN thrombocytosis and to examine their mechanism of action by characterizing affected patients, murine and human cell line models.

Methods: A comprehensive NGS panel of 36 genes involved in myeloproliferative neoplasms (MPN) pathogenesis was performed to screen a cohort of 117 patients with TN thrombocytosis followed in our institution. Clinical and biological data, as well as blood samples were obtained from patients carrying an SH2B3 variant. Nail DNA was extracted to confirm or not germline origin. Platelets isolated from patients and human cell line transduced to express SH2B3 variants were examined by flow cytometry and western blots.

Results: We identified an unreported subset of five patients with isolated SH2B3 variants of germline and somatic origin in our cohort of TN thrombocytosis patients. All variants were within the pleckstrin (PH) domain, with the exception of one located at the beginning of the SH2 domain, and were characterized by frameshift (fs) mutations, all heterozygous. The analysis of nail DNA from these patients highlighted three germline and two somatic cases. Most important was the clinical features of patients with SH2B3 fs variants. Indeed, all five probands presented severe thrombotic events with portal vein thrombosis, myocardial infarction, several stroke episodes, and cerebral venous thrombosis. We identified two different mutations, the novel mutation c.1093_1094ins31 p.(F365Wfs*30) in one proband and the mutation c.685_691dup p.(D231Gfs*39) in 4 distinct patients. To confirm the latter result, we cross-examined an external independent cohort of TN MPN patients, out of which two probands from the same family (mother and daughter) were identified with that mutation. This result strongly suggests that SH2B3 D231Gfs*39 variant is involved in the thrombocytosis phenotype and severe thrombotic events affecting these patients.

To assess the mechanism of action of our recurrent SH2B3 fs mutation, we analyzed LNK expression level in platelets isolated from p.(D231Gfs*39) mutated patients by western blot. No truncated LNK form was detected; instead, a significant reduction of the WT LNK protein was observed compared to healthy donors. Human HEK cells expressing this SH2B3 variant confirmed this result, indicating that this fs mutation leads to a defect in total LNK protein. Indeed, in silico modeling of this SH2B3 variant showed a disruption of the LNK PH domain, causing an unstable WT LNK protein unable to effectively regulate signaling complexes at the membrane. Flow cytometry analysis of TPO signaling of these platelets showed a significant increase in STAT5 activation compared to healthy controls, confirming MPL/JAK2/STAT5 hyper-signaling as a functional consequence of these LoF SH2B3 mutations.

Conclusions: We identified frameshift heterozygous SH2B3 germline variants which provide for the first-time evidence of SH2B3 association with the thrombocytosis phenotype displayed in TN patients. A defect on LNK expression can cause abnormal function and JAK/STAT upregulation, possibly in a driver-like fashion. Notably, D231Gfs*39 heterozygous variant displayed different clinical traits from those reported at the homozygous state (Perez-Garcia, A., et al, Blood, 2013; Arfeuille, C., et al, Haematologica, 2023), suggesting that SH2B3 genetic status may determine the severity of the outcome phenotype observed in patients. Ongoing studies on KO and heterozygous Sh2b3 mice expressing this SH2B3 variant will shed light on the scope of the genetic condition.