MSI2 and HOXA9 Expression Is Significantly Increased in Patients with Primary Myelofibrosis

Musashi 2 (also known as MSI2), a mRNA binding protein, is reported to control critical stem cell fate decisions by binding to the 3’untranslated region of target mRNAs, thereby inhibiting translation. MSI2 is preferentially expressed in hematopoietic tissue and in particular in early myeloid progenitors. Furthermore, investigators have shown MSI2 expression is upregulated by HOXA9 resulting in dysfunction of the regulatory pathway, leading to hematopoietic stem cell proliferation, impaired myeloid differentiation and is associated with worse prognosis in chronic myeloid leukemia (CML) and acute myeloid leukemia (Kharas, et al. Nat Med 2010; 16:903; Ito, et al. Nature 2010; 466:765). We verified MSI2 levels are significantly increased in CML patients in blast crisis compared with those in chronic phase, irrespective of lymphoid or myeloid transformation (Kaeda, et al. Blood. 2011;118;supplement). Given these observations we sought to assess MSI2 and HOXA9 expression in BCR-ABL1negative myeloproliferative neoplasm (MPN) patients.

We retrospectively quantified MSI2 and HOXA9 expression in 38 samples of which 22 were from highly heterogeneous MPN patients, all of whom are alive. These 22 MPN patients were classified as polycythemia vera (PV) n= 5; essential thrombocythemia (ET) n=3; primary myelofibrosis (PMF) n=11 and unclassified MPN (U-MPN) n=3, based on clinical hematologic and molecular parameters in accordance with the World Health Organization 2008 classification. The MPN patients demographic features were (M: 13; F:9) with median age of 69.5 years (range 53-88). Of the 22 patients 17 had the V617F allele, in one it was undetectable and 4 were not tested. The remaining cDNA samples were unselected normal controls (NC) from healthy blood donors (M: 6; F:10) with median age of 41 years (range 34-61). MSI2 and HOXA9 mRNA levels were quantified by quantitative real time PCR (Q-PCR) and normalized to GUSβendogenous control gene, expression levels. Patient and control data were subjected to Mann-Whitney unpaired test two-tailed anaylsis using Graphpad Prism version 6.04 software.

MSI2 and HOXA9 expression was detectable in all the samples by Q-PCR. However, HOXA9 transcript numbers, median 0.115% (range 0.020-2.360) were significantly increased (p=0.0123) in MPN patients when compared with the NC Group, median 0.060% (range 0.030-0.120). Conversely, median MSI2 expression level 0.715% (range 0.370-2.030) in MPN patients, was significantly decreased, p<0.0001 compared to that observed in the NC Group, median 2.817% (range 1.445-7.533). But both MSI2 and HOXA9 were significantly increased, p<0.0001, when comparing NC Group (n=16) with the PMF patients (n=11). Of these 11 PMF patients 6 had the V617F allele, in one it was undetectable and 4 were not screened. Interestingly, we noted a difference between PV and PMF data with respect to HOXA9 and MSI2 expression. Specifically, the PMF patients’ HOXA9 median [n=11; median age 62 years (range 53-81)], 0.790% (range 0.03-2.36); was significantly higher, (p=0.0021), than the PV patients’ [n=5; median age 77 years (range 69-88)], 0.030% (range 0.10-0.40). By contrast no significant difference, p=0.0504, was detected when comparing MSI2expression between patients with PV and PMF. The sample size was too small to reliably evaluate the other MPN groups.

Our data are consistent with the notion that MSI2 plays a significant role in the biology of myeloid malignancies. While PV and ET patient numbers in this study were too small for evaluation, we did note finding increased HOXA9 expression among PMF subjects is consistent with reports that up to 30% of PMF patients harbor ASXL1 mutations, which are reported to lead to increased HOXA9 expression. Furthermore, the ASXL1 mutations are associated with worse prognosis, such that stem cell transplant (SCT) is recommended for patients harboring these mutations. In this study 2 of the 11 PMF patients had SCT and were among three patients with the highest HOXA9 levels determined. Given the findings it would be interesting to screen the PMF patients for ASXL1 mutations. Equally, finding MSI2 and HOXA9 increased expression in PMF patients is supportive of the later regulating the former. However, we observed no correlation between MSI2 and HOXA9 levels. An expanded study is required to assess these preliminary data implying HOXA9 expression is a prognostic marker in PMF patients.