Possible new <i>LNK</i> mutations in myeloproliferative neoplasms

Abstract

Recently, lymphocyte adaptor protein (LNK or SH2B3) has been reported as a new genetic abnormality in BCR-ABL negative myeloproliferative neoplasms (MPN). We performed LNK sequence analysis in 42 chronic phase MPN patients (14 polycythemia vera, 15 essential thrombocythemia, 9 primary myelofibrosis, and 4 myeloproliferative neoplasm, unclassified) and detected total three types of genetic mutations (7.1%) including p.Q423X, p.R551W and p.I568T. These mutations were accompanied by a JAK2 V617F mutation. The p.Q423X is a nonsence mutation located on exon 7 encoding the Src homology 2 domain. The p.R551W and p.I568T are missense mutations located on exon 8 encoding the C-terminal region. Our study suggests that LNK mutations occur in low frequency in human MPN, and can occur in several regions of the LNK gene not only on a pleckstrin homology domain which have been regarded as a ‘hot spot’. Recently, the mutation of lymphocyte adaptor protein (LNK or SH2B3) was detected in MPN. LNK is a negative regulator of thrombopoietin (TPO)-MPL and erythropoietin (EPO)-EPO receptor-mediated JAK2 activation [10, 11]. LNK belongs to a family of adaptor proteins possessing several domains: a proline-rich N-terminal dimerization domain, a pleckstrin homology (PH) domain, a Src homology 2 (SH2) domain, and a conserved C-terminal tyrosine residue [12]. Via its SH2 domain, LNK binds to JAK2 and inhibits its downstream signal transduction. LNK is also capable of regulating the aberrant signaling of MPL-W515L and JAK2 V617F in vitro [13, 14]. LNK-deficient mice show a phenotype consistent with MPN, involving an expanded hematopoietic stem cell compartment, megakaryocytic hyperplasia, splenomegaly, leukocytosis, and thrombocytosis [15]. In human MPN, three studies about LNK mutations have been reported, in which 12 of 102 patients were shown to harbor 14 different LNK mutations [16-18]. In this study, we performed LNK mutation analysis in 42 chronic-phase MPN patients to study additional mutations in human MPN. We also performed JAK2 and MPL W515L/K mutation analyses to identify the relationship with the LNK mutation. The patients comprised 27 males and 15 females with average ages of 63 and 64 years, respectively. All patients were on initial diagnosis with chronic phase, and the subtypes were 14 PV, 15 ET, 9 PMF, and 4 MPN, unclassified (MPN-U) (See Supporting Information I). Six genetic variations were detected including p.P242S, p.W262R, p.Q423X, p.A536T, p.R551W, and p.I568T (Fig. 1). The p.P242S (c.724C>T) were detected in six patients and was an already known polymorphism (rs78894077) listed in the National Center for Biotechnology Information Single Nucleotide Polymorphism (SNP) Database (dbSNP; http://ncbi.nlm.nih.gov/projects/SNP/). The p.W262R (c.784T>C) was detected in all 42 patients as a homozygote, was also known polymorphism (rs3184504). The p.A536T (c.1606G>A) variation was detected in a PV patient. Although it was not listed in the dbSNP, we regarded it as one of polymorphisms, because it was detected in one of 50 normal controls. In contrast, the other three variations, p.Q423X (c.1267C>T), p.R551W (c.1651C>T), and p.I568T (c.1703T>C), which were detected in each ET, PMF, and PV patient, were not found in 50 control samples. Moreover, those had not been reported in previous literatures or listed in the dbSNP, either. The p.Q423X is located on exon 7 of LNK gene and results in a premature stop codon in the SH2 domain. The p.R551W and p.I568T are missense variations located on exon 8, which encodes the C-terminal region of the LNK protein. Because we could not perform germline study with the paired normal cell in each patient, we could not definitely rule out the possibility that these variations were not true somatic mutations, especially for p.R551W and p.I568T. However, when we regarded these three variations as new mutations based on the study using 50 normal controls, the prevalence of LNK mutation in our study could be estimated as 7.1% in chronic-phase MPN. The evaluation of the subtype-specific prevalence of LNK mutation was not performed due to small numbers of cases harboring the mutations. Previous estimates of LNK mutation prevalence have ranged from 6.1 to 25% [16-18]. Oh et al. reported two types of LNK mutations in 33 (6.1%) JAK2 V617F-negative MPN patients [17], and Lasho et al. identified two mutations (25.0%) in eight JAK2 V617F-negative patients with unexplained erythrocytosis [19]. Pardanani et al. studied a relatively large cohort of 61 patients with blast-phase MPN, and postulated that LNK mutations are more prevalent in the blast-phase (13.1%) than in the chronic phase (4.9%). Nucleotide variations found in our study. Arrows in each DNA sequence trace are indicating the substituted position. Three variations are new mutations composed of two missense mutations (p.R551W and p.I568T) and one nonsense mutation (p.Q423X). The others (p.P242S, p.W262R and p.A536T) are known SNPs. The previously reported mutations were variable and differed across patients, except for one mutation (c.644C>T) that was detected in three cases (Table I). Interestingly, most of the mutations were located on the PH domain, which have led investigators to regard the PH domain as a mutational “hot spot” [18]. The PH domain of LNK is involved in the colocalization of LNK to the plasma membrane. These mutations are expected to disrupt the PH domain and consequently mislocalize LNK in the cytoplasm abrogating its function to JAK2. However, the mutations we detected were on the SH2 domain or C-terminus. The SH2 domain of LNK is known to bind JAK2 and do a critical role in negative regulation of the downstream signal transduction. So, we expected the p.Q423X mutation might result in the disruption of LNK function and have a role in MPN pathogenesis. If we perform the expression study for the production of truncated protein or no protein synthesis by this mutation, it would be more helpful to understand its effect to LNK function. Regarding the p.R551W and p.I568T in C-terminal region, actually we cannot expect the functional consequences because those are not located in the main domain. When we performed the conservation study, we could find these location were well-conserved interspecies, so we just guessed these amino acid changes might have effect on the LNK structure or its function (See Supporting Information II). According to previous investigations using transgenic mice, the mutational effect of LNK could be different according to the mutation type. The SH2-mutated type showed most severe disruption of LNK function compared to the mutated form in PH domain or C-terminal tyrosine residue [10, 19]. If large number of cases with LNK mutations would be collected in the future, the phenotypic difference according to mutation type could be evaluated. The JAK2 V617F (exon 14) was detected in 26 of 42 patients (61.9%), but the exon 12 mutation was not detected in this study. All of three patients with LNK mutation also harbored JAK2 V617F mutation. The MPL W515L was identified in two ET patients, and neither of them harbored LNK or JAK2 V617F mutations. The frequent cooccurrence of LNK mutation and JAK2 V617F was also observed in previous study, which reported LNK mutations were found at similar frequencies between JAK2 V617F-negative and positive patients [17]. Now we do not know whether the LNK mutation and JAK2 V617F occur in the same clone or not, and the effect of those cooccurrence on MPN pathogenesis or phenotype. We expected more severe phenotype will result in cases with two mutations than those with only one mutation, based on a previous study that reported the augmented ability of oncogenic JAK2 to expand myeloid progenitors in LNK-deficient mice [20]. Although only a few studies, including ours, have been published regarding LNK mutations in human MPN, the collective data does favor some suggestions. The prevalence of LNK mutations is not high in human MPN, especially among patients with chronic-phase MPN (<10% mutated). Second, LNK mutations can target several regions of the LNK gene, not only in the PH domain. Whether mutations in different domains contribute differently to phenotypic expression needs to be addressed in larger studies. Finally, the LNK mutation could be accompanied by JAK2 V617F. We cannot be sure if LNK mutations and consequent functional disruption are solely responsible or only provide a supportive role for MPN pathogenesis by other genetic changes like JAK2 V617F. Larger-scale studies of LNK mutation in human MPN are warranted. This study was approved by the Dongsan Medical Center's Institutional Review Board. Total 42 DNA samples that had been extracted and stored from peripheral blood or bone marrow of MPN patients between 2007 and 2010 were used. All samples were obtained on initial diagnosis and chronic phase of MPN. For control study, total 50 DNA samples from healthy controls were acquired. The primers for LNK (exon 2-8) and JAK2 (exon 12 and 14) were designed using Primer3 (http://frodo.wi.mit.edu/primer3/), and the sequencing was conducted using a BigDye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA) on an ABI 3730XL DNA analyzer (Applied Biosystems) (See Supporting Information III). The detection of MPL W515L or W515K mutation was performed using a Real-QTM MPL W515L/K screening kit (BioSewoom, Seoul, Korea) according to the manufacturer's instructions. The authors thank KS Park, Ph.D. and WA Bae, Ph.D. for good assistance with sample collection, DNA extraction, and sequencing analysis. Additional Supporting Information may be found in the online version of this article. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. Jung Sook Ha*, Dong-Seok Jeon*, * Department of Laboratory Medicine, Keimyung University School of Medicine, 194 Jung-gu Dongsan-dong, Daegu 700-712, South Korea.