To the editor:
Myeloproliferative neoplasms (MPNs), including polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis, have in most instances a sporadic occurrence, but familial clustering of MPNs has been reported and familial cases are about 7% to 8% of all MPN patients.1,2 Driver mutations in JAK2, CALR, or MPL are somatically acquired in familial cases as they are in sporadic patients.1,3,4 Common single-nucleotide polymorphisms in the JAK2 and TERT genes confer susceptibility to MPNs and contribute to the familial clustering of MPNs.5-7 Recently, germ line RBBP6 mutations have been identified in about 5% of familial MPN cases8 and germ line duplication of ATG2P and GSKIP genes has been reported in 4 families from the French West Indies.9
The SH2B adaptor protein 3 (SH2B3) gene, also known as the LNK gene, encodes a negative regulator of cytokine signaling. In mouse models, Lnk negatively regulates erythropoietin receptor signaling and thrombopoietin receptor signaling by attenuating Jak2 activation, and thus negatively modulating erythropoiesis and megakaryopoiesis, respectively.10,11
LNK mutations have been described in some patients with sporadic MPNs12,13 and in a small number of cases with idiopathic erythrocytosis and subnormal Epo levels.14 LNK mutations mainly affect exon 2 and may occur concurrently with the JAK2 (V617F) mutation.12,14
In an attempt to identify the germ line genetic factors that underlie familial clustering of MPNs, we applied next-generation sequencing to our MPN families. All samples were collected after subjects gave their written informed consent and the study was approved by the local ethics committee.
Our cohort of 94 MPN families was analyzed with 2 strategies. First, we applied whole-exome sequencing (HiSeq2000 system; Illumina) in a subgroup of 16 families with MPNs. This approach resulted in the identification of the LNK (E208Q) mutation in a patient with familial PV belonging to family 36; the variant was then validated by Sanger sequencing.
We next screened for exon 2 LNK mutations by Sanger sequencing in the remaining 93 families. All affected and healthy members for whom DNA was available were studied (149 patients and 89 healthy relatives). Of 149 patients affected with familial MPNs, 2 patients (1.4%) carried the LNK (E208Q) mutation (including the initial case, identified through exome sequencing). None of the 89 healthy relatives carried mutations in exon 2 of the LNK gene.
The pedigrees of the 2 mutated cases (MeF and MPC12_294, belonging to family 36 and family 38, respectively) are reported in Figure 1. Both patients were affected with JAK2 (V617F)-mutant PV, diagnosed at 42 years (MeF) and 66 years (MPC12_294). The 2 patients carried LNK (E208Q) both in granulocyte and T-lymphocyte DNA. To confirm the germ line nature of the mutation, we analyzed DNA extracted from hair roots of patient MeF, detecting LNK (E208Q) also in this nonhematopoietic tissue. In both families, the other family member affected with MPNs (MeR and MPC07_24) did not carry any mutation in the LNK gene, thus excluding segregation of the LNK (E208Q) mutation with the disease phenotype.
In conclusion, germ line LNK mutations rarely occur in familial MPNs and do not segregate with the disease phenotype. Our findings suggest that mutations in LNK, either germ line or acquired, may cooperate with acquired driver mutations in JAK2, CALR, or MPL to determine disease phenotype in MPNs. In the study by Oh et al, the patient with the missense mutation (E208Q) in the pleckstrin homology domain of LNK had an ET phenotype.13 This patient was negative for JAK2 (V617F) and MPL (W515) mutations; however, CALR mutations had not yet been described at the time of this report, and we cannot exclude that a CALR mutation was responsible for this ET. Two additional patients reported for their LNK mutations had idiopathic erythrocytosis and not a myeloproliferative disorder.14 Overall, it appears unlikely that LNK alterations may act as driver mutations in MPNs.
Authorship
Acknowledgments: The authors thank Giorgio Lo Russo and Laura Tassi for their technical support.
This work was supported by a grant from Associazione Italiana per la Ricerca sul Cancro (AIRC; Milan, Italy), Special Program Molecular Clinical Oncology 5x1000 to AIRC–Gruppo Italiano Malattie Mieloproliferative (AGIMM) project no. 1005, and by a grant from AIRC (my first AIRC grant MFAG-2014-15672; E. Rumi). The Sonderforschungsbereich (SFB) grant from the Austrian Science Fund (FWF; F4702-B20) is acknowledged for its generous support (J.D.M.F. and R.K.).
Contribution: E. Rumi and A.S.H. designed research and wrote the paper; A.S.H., D.P., J.D.M.F., C.M., and M.C.R. performed molecular investigations; C.C., E. Roncoroni, I.C., M.B., M.G., and C.A. collected clinical data; and M.C. and R.K. finalized the manuscript.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
Correspondence: Elisa Rumi, Department of Hematology Oncology, Fondazione IRCCS Policlinico San Matteo Pavia, Department of Molecular Medicine, University of Pavia, Viale Golgi 19, 27100 Pavia, Italy; e-mail: elisarumi@hotmail.com and elisa.rumi@unipv.it.