Clinical Significance of LNK (SH2B3) Expression in Pediatric B Cell Precursor Acute Lymphoblastic Leukemia

Background: In pediatric B cell precursor acute lymphoblastic leukemia (BCP-ALL), aberrant JAK-STAT signaling is one of the key leukemogenic mechanisms. Although activating mutations of JAK2 are found in high-risk BCP-ALL patients in western countries, such mutations are rare in Japanese cohorts, led to speculation of other factors contributing abnormal activation of JAK-STAT pathway. The adaptor protein LNK (SH2B3) is one of the negative regulator of JAK-STAT signaling, and its loss of function mutations have been identified in myeloproliferative neoplasms and high-risk BCP-ALL. In addition, the loss of function of LNK has been demonstrated to increase proliferation of B cells in vivo. Based on these findings, we conducted genetic analysis to determine the prognostic impact of LNKin BCP-ALL, and functional analysis to investigate its possible mechanisms.

Methods: For genetic analysis, we evaluated diagnostic bone marrow or peripheral blood samples of 164 pediatric BCP-ALL patients treated with the Japan Association of Childhood Leukemia Study (JACLS) ALL02 protocol along with peripheral blood samples from 9 healthy volunteers. The LNK expression level was determined by qRT-PCR. Deletion of IKZF1 was determined by MLPA, and direct sequencing was employed to detect LNK mutations in patients with IKZF1 deletion. For functional analysis, thrombopoietin (TPO)-dependent BaF3-MPL cells expressing LNK (BaF3-MPL-LNK) was established by retroviral transduction to investigate if LNKexpression levels impact on drug sensitivity for prednisolone (PSL), doxorubicin (DOX) or vincristine (VCR). For growth assay, BaF3-MPL cells with or without LNK were cultured for four days in the presence of TPO and the viable cell number was counted. For drug sensitivity test, BaF3-MPL cells with or without LNK were treated with each drug for 48 hours in the presence of TPO, then the IC50 was calculated. Phospho-specific flow cytometory (Phos-Flow) was performed to measure JAK-STAT activation.

Results: The LNK expression levels in pediatric BCP-ALL patients’ samples were significantly lower than in samples from healthy donors (P < 0.01). When analyzing all 164 cases, the expression level of LNK was decreased in relapsed patients but there was no statistical significance (P = 0.067). IKZF1 deletion was found in 25 (15% of all) patients, and in these patients, LNK expression level dose not relate to relapse (P = 0.39). Intriguingly, when patients with known high-risk factor (i.e., IKZF1 deletion or poor response to PSL) were excluded, the expression level of LNK was significantly higher in non-relapsed patients (P < 0.05). In functional assay, we observed inhibition of TPO dependent growth of BaF3-MPL cells by expression of LNK (P < 0.01), consistent with previous reports. Phos-Flow analysis revealed that LNK expression suppressed TPO–induced phosphorylation of STAT5 in BaF3-MPL cells. In drug sensitivity test, we found that IC50 of PSL and DOX were substantially lower in BaF3-MPL-LNK cells from in BaF3-MPL-mock cells (0.70 vs 3.93 nM, P < 0.01 and 0.61 vs 1.14 nM, P < 0.05, respectively). Decline in IC50 of VCR by LNK expression was not statistically significant (1.38 vs 2.45 nM, P = 0.056). We next compared the impact of LNK with Ruxolitinib (RUX), a potent synthesized JAK2 inhibitor. The diminution in IC50 of PSL in BaF3-MPL-mock cells treated with RUX (50 nM) was comparable of that in BaF3-MPL-LNK cells (0.31 nM, combination index (CI) = 0.39), consistent with our hypothesis that LNK is working as a JAK2 inhibitor. Since we identified two amino-acid substitutions in N-terminal proline-rich dimerization domain (R139H) and PH domain (P242S), we also examined their function. Transductions of these genes in BaF3-MPL cells, however, did not alter cell growth, suggesting they are single nucleotide variants.

Discussions: Our findings that high LNK expression is associated with low relapse rate in intermediate risk (IKZF1 intact, good PSL response) patients indicate potential of LNK to restrain relapse in such patients, presumably by suppressing JAK-STAT signaling. Since we proved the impact of LNK expression to improve sensitivity of PSL in vitro which was comparable to RUX, RUX could compensate lack of internal LNK expression to induce cell death of BCP-ALL cells. Collectively, targeting JAK-STAT pathway will be promising therapeutic option for intermediate risk BCP-ALL patients with low expression level of LNK.