Association of a single-nucleotide polymorphism in the SH2B3 gene with JAK2V617F-positive myeloproliferative neoplasms

To the editor:

In a recent paper, Perez-Garcia et al described an inherited mutation in the SH2B3 gene associated with the development of acute lymphoblastic leukemia.¹ SH2B3 encodes the lymphocyte adaptor protein (LNK) that negatively modulates the signaling of several cytokine receptors, including the thrombopoietin receptor (myeloproliferative leukemia virus oncogene [MPL]) and the erythropoietin receptor, by attenuating JAK2 kinase activation. LNK was also shown to bind and regulate mutant signaling molecules found in myeloproliferative neoplasms (MPNs) like MPL-W515L and JAK2-V617F^2-4  through its SH2 domain and a novel site at its amino-terminal region. Lastly, several acquired SH2B3 mutations in the pleckstrin homology domain and NH2-terminal region of the protein were reported in MPNs.^4-6  Moreover a nonsynonymous single-nucleotide polymorphism (SNP), rs3184504 (p;R262W, c.784T>C), in exon 2 of the SH2B3 gene was reported to be associated with type 1 diabetes,⁷  risk of coronary artery disease, thrombocytosis,⁸  and celiac disease.

To explore a potential predisposing role of this SH2B3 SNP sequence in MPNs, we analyzed it in whole-blood DNA in a large cohort of MPN patients with and without the JAK2V617F mutation. In a control population of 2851 individuals the reported T- and C-allele frequencies were respectively 43.9% and 56.1%.⁹  Of note, similar figures were found in another control group of 1005 Russian subjects.¹⁰ 

We therefore analyzed the frequency of each allele of the rs3184504 in 317 MPN patients, including 169 patients with essential thrombocythemia (103 with the JAK2V617F mutation and 66 without) and 148 with myelofibrosis (81 with the JAK2V617F mutation and 67 without). We found that the T-allele frequency was significantly higher than reported in the control populations (Table 1) (odds ratio [OR] = 1.32; P = .0008). We then studied the T- and C-allele frequencies between JAK2V617F-positive and JAK2V617F-negative patients. There was no association between the T/T profile and the percentage of JAK2V617F. We found that the T allele was significantly more frequent in JAK2V617F-positive patients as compared with controls (OR = 1.68; P = .0001) and JAK2V617F-negative patients (OR 1.78; P = .0004), whereas no difference was found between JAK2V617F-negative MPN patients and the control group (OR = 0.95; P = .663). Interestingly, the higher frequency of the T allele was associated with the JAK2V617F myelofibrosis group (OR = 2.23; P = .0001).

In conclusion, these results reveal a significant association between the c.784T>C nonsynonymous polymorphism of LNK and JAK2V617F-positive MPNs, further supporting a role of the SH2B3 gene in the predisposition to hematologic malignancies. Functional studies are required to understand how the T risk allele may contribute to the development of MPNs. In particular, the charge reversal due to this nonsynonymous SNP on a surface-exposed residue in the pleckstrin homology domain of the molecule may affect its interactions with downstream signaling molecules.