WTAP is a novel oncogenic protein in acute myeloid leukemia

The biology of acute myelogenous leukemia (AML) is characterized by a block in differentiation, increase in proliferation and inhibition of apoptosis, all of which when combined lead to an expansion of leukemic blasts.1 AML therapy cures ~20% of those affected,2 highlighting the need for a better understanding of leukemia biology in order to identify novel therapeutic targets. The Wilms’ tumor 1 (WT1) gene has an oncogenic role in leukemogenesis and its overexpression correlates with a poor prognosis.3 WTAP is a highly conserved protein that partners with WT14,5 to function as a switch gene, regulating the balance between quiescence and proliferation.6 WTAP-null mice exhibit embryonic lethality.7 WTAP has been recently described as an oncogenic factor in gliomas.8 We hypothesized that alterations in WTAP, whose role in leukemogenesis is unknown, might provide an alternate means of modulating the WT1 pathway in AML. We performed a series of experiments to test this possibility. First, we examined WTAP expression in several primary AML samples observing increased levels of WTAP in AML compared with normal peripheral blood mononuclear cells (Figure 1a) as well as in AML cell lines (data not shown). Next, the WTAP expression levels were determined in 511 newly diagnosed AML patients using the reverse-phase protein array (RPPA) technique. In comparison with normal bone marrow CD34+ cells, WTAP expression in bulk AML cells was above normal levels in 32% of patients (Figure 1b). Although WTAP levels were not associated with individual cytogenetic abnormalities, some specific molecular mutations such as NPM1 and FLT3-ITD were found to have significant correlation (P ≤ 0.05) with WTAP expression (Supplementary Figure 1A). In addition, RPPA analysis showed that WTAP levels were positively correlated (R > |0.2|) with levels of various cell proliferation-related proteins (cyclins and Hsp90), antiapoptotic proteins (Bcl-2 and Bax), oncoproteins (FLI1) and proteins important for stem cell functions such as Myc and Ash2L (Supplementary Figure 1B). To assess the functional significance of increased WTAP expression, its expression was silenced in K562 and HL-60 cells, leading to a significant reduction (P ≤ 0.05) in proliferation (Figure 1c), and clonogenic survival (P ≤ 0.01) (Figure 1d). Similar effects on proliferation were observed in the AML cell line OCI-AML3 and in primary AML cells (Supplementary Figure 1C), suggesting a pro-proliferative role for WTAP in AML. WTAP knockdown alone did not induce apoptosis but markedly increased (P ≤ 0.01) the extent of apoptosis following etoposide treatment (Figure 1e). These results provide evidence for an association between the increased expression of WTAP and chemoresistance in AML. Figure 1 Expression of WTAP in AML and effect of WTAP silencing on AML cell behavior. (a) Peripheral blood mononuclear cells from normal donors (NL) and AML patients (AML) were obtained by Ficoll–Paque density centrifugation, and protein extracts were ... To examine the role of WTAP in AML progression in vivo, we performed tumor xenograft experiments in nude mice. As shown in Figure 1f, the growth rates and masses of tumors derived from WTAP-knockdown cells were significantly reduced (P ≤ 0.01) compared with control. To complement this analysis, the in vitro transforming activity of WTAP was examined by investigating its effects on growth of the Ba/F3 cell line. This line depends on interleukin 3 (IL-3) for survival and proliferation, but this dependence can be released by the transgenic expression of suitable oncogenes.9 Whereas control Ba/F3 cells were not viable in the absence of IL-3 at 72 h, WTAP-expressing Ba/F3 cells were able to maintain growth factor-independent proliferation, as demonstrated by significantly higher (P ≤ 0.01) number of viable cells (Figure 1g), suggesting that WTAP harbors oncogenic activity. The aberrant cellular proliferation and terminal differentiation block of myeloid cells are two hallmarks of AML.10 Having shown that WTAP regulates growth and survival, we investigated whether WTAP has a role in myeloid cell differentiation. As shown in Figure 1h, knockdown of WTAP promoted phorbol 12-myristate 13-acetate (PMA)-induced myeloid differentiation, as revealed by an increase in the expression of myeloid differentiation markers CD11b and CD14 compared with control cells. These results suggest that increased expression of WTAP in AML not only supports cell proliferation but also induces the differentiation block. Our RPPA analysis suggested a link between WTAP and mammalian target of rapamycin (mTOR) expression; and given that the mTOR pathway is deregulated in a number of cancers including AML,11 we hypothesized a putative regulatory role of WTAP on mTOR activity in AML. As shown in Figure 1i, WTAP knockdown induced a decrease in the phosphorylation levels of mTOR and its downstream effector p70 ribosomal subunit 6 kinase (pS6K) compared with control shRNA. To further understand the participation of WTAP in leukemogenesis, we performed transcriptomic analysis with RNA-Seq on WTAP knockdown in K562 cells. Gene ontology analysis indicated that cell adhesion and regulation of cell proliferation are the most enriched functionalities (Supplementary Figure 1D and Supplementary Table 2). Among the most relevant genes affected by WTAP with recognized roles in leukemia are CD4, CD44, CEBPA, CSF1R, MPO, ABCG2, TCL1A, CYP1A1, CYP3A4, FGFR1, PTPRC (CD45), CD83, CD86, CD9 and CCR4. Consistent with its described role in RNA processing, we determined that WTAP knockdown affected exon usage of 93 genes, including important factors such as MLL and MSI2. Interestingly, we also observed that WTAP affects the processing of its own transcript, very likely via polyA site selection. Complete analysis can be accessed in GEO ({"type":"entrez-geo","attrs":{"text":"GSE46718","term_id":"46718"}}GSE46718—http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=jbkjjswaqscegjsa Fisher’s exact test) in their mRNA levels as determined by RNA-Seq. Mutations of WTAP were not observed in the TCGA analysis of AML.12 Therefore, the etiology of increased WTAP expression in AML remains unexplained. We next sought to determine the potential mechanism that may contribute to an increase in WTAP expression in AML. The molecular chaperone Hsp90 maintains the stability of many tumor-promoting oncoproteins,13 including WT1.14 Keeping in mind the connection between WT1 and WTAP, we investigated the potential interaction between Hsp90 and WTAP. First, we determined that WTAP co-immunoprecipitates with Hsp90 (Figure 2a), whereas treatment with the Hsp90 inhibitor ganetespib significantly reduced the binding of Hsp90 to WTAP. Therefore, formation of the WTAP–Hsp90 complex is dependent on the chaperoning activity of Hsp90. Studies have shown that Hsp90 client proteins shift the primary chaperone association from Hsp90 to Hsp70 following inhibition of Hsp90 activity.15 Accordingly, our results showed that ganetespib treatment increased WTAP association with Hsp70 (Figure 2a). Furthermore, the GST pull-down assay showed a direct interaction between WTAP and Hsp90 (Figure 2b). To understand the functional significance of the association of Hsp90 with WTAP, we investigated the effects of Hsp90 inhibitors on WTAP protein stability. Ganetespib treatment promoted the degradation of WTAP in K562 (CML), MV4-11 (AML) and Kasumi-1 (AML) cell lines, previously shown to be highly sensitive to ganetespib (Figure 2c).16 Marked reduction in WTAP expression was similarly observed in blasts exposed to ganetespib (Figure 2d). Furthermore, ganetespib inhibited AML tumor growth in nude mice, as we reported previously,14 and WTAP expression in xenograft tumors (Figure 2e). Inhibition of Hsp90 induces polyubiquitination and proteasomal degradation of Hsp90 client proteins.17 We confirmed that pretreatment with proteasomal inhibitor bortezomib largely prevented ganetespib-induced degradation of WTAP (Figure 2f), and higher levels of ubiquitinated WTAP were observed in the presence of the combination of bortezomib and ganetespib compared with either agent alone (Figure 2g). Together, these data show that ganetespib-mediated degradation of WTAP is dependent on the ubiquitin-proteasome pathway, very similar to other bonafide Hsp90 client proteins. Figure 2 Hsp90 associates with WTAP and is necessary for its stability. (a) The K562 cells were treated with vehicle (−) or Hsp90 inhibitor, ganetespib (1 µm for 6 h) followed by immunoprecipitation (IP) with IgG and WTAP antibody. The immunoprecipitates ... In summary, our study shows that WTAP has an important role in abnormal proliferation and arrested differentiation of leukemia cells, and WTAP is a novel client protein of Hsp90. Therefore, WTAP may be a promising novel therapeutic target in AML. Further investigation aimed to elucidate the mechanism of action of WTAP is warranted.