Phosphoproteomic analysis identifies the M0-91 cell line as a cellular model for the study of TEL-TRKC fusion-associated leukemia

The TEL-TRKC gene fusion associated with the t(12;15)(p13;q25) translocation has been implicated in both hematological (acute myeloid leukemia (AML))1 and non-hematological malignancies (congenital fibrosarcoma, congenital mesoblastic nephroma and secretory breast carcinoma). In AML, the TEL-TRKC (TEL-TRKC(L)) derives from the in-frame fusion of exons 1–4 of TEL to exons 13–18 of TRKC. In contrast, the TEL-TRKC variant associated with solid tumors (TEL-TRKC(F)) contains exons 1–5 of the TEL gene. Activation of both the RAS-MAPK and PI3K-AKT pathways by TEL-TRKC contributes to oncogenic signaling in transfected NIH3T3 cells.2 However, no human cell lines are available to study the TEL-TRKC fusion. In this study, we screened over 40 AML cell lines for constitutive phosphorylation of STAT5 by Western blot. The M0-91, an AML-M0-derived cell line3, showed constitutive tyrosine phosphorylation of STAT5 (Figure 1a). To identify protein tyrosine kinases responsible for the constitutive phosphorylation of STAT5 in M0-91, cell lysates were trypsin-digested, and phosphopeptides were immunoprecipitated with phosphotyrosine antibody (pY-100), and analyzed by LC-MS/MS mass spectrometry.4, 5 LC-MS/MS mass spectrometry identified 393 phosphotyrosine sites in 265 proteins (Supplementary Table 1). Among these proteins, over 15 tyrosine kinases were tyrosine-phosphorylated (Figure 1b). Multiple tyrosine-phosphorylated peptides corresponded to either TRKB or TRKC, including three tyrosines in the activation loop. Thus, either TRKB or TRKC could be aberrantly activated in M0-91 cells. While full-length TRKB/C have a molecular weight of 140–145 kDa, we observed a 50 kDa by Western blot with a pan-TRK antibody in M0-91 cells (Figure 1c). In addition, we observed tyrosine-phosphorylated peptides deriving from the TEL protein (Supplementary Table 1). TEL is a member of the ETS family transcription factor and is essential for hematopoiesis. It is a frequent target of chromosomal translocations in human cancers, and one of its fusion partners is TRKC.1 To determine whether a chimeric TEL-TRKC transcript was present, we performed 3' rapid amplification of complementary DNA (cDNA) ends on the sequence encoding the HLH domain of TEL. Sequence analysis of the resultant product revealed that the kinase domain of TRKC was fused to TEL gene (Figure 1d). In agreement with a previous observation in a patient diagnosed with TEL-TRKC AML,1 the M0-91 cell line expressed TEL-TRKC fusion transcript (Figure 1e), with exon 4 of TEL fused in-frame to exon 14 of TRKC. It lacks a 42 bp alternative spliced exon in the TRKC moiety (data not shown). The lack of this exon enhances tyrosine kinase activity.6, 7 Thus, this result confirmed the presence of the 'leukemic' form of the TEL-TRKC fusion. No reciprocal TRKC-TEL fusion gene was detected. The M0-91 cell line did express wild type TEL, but not wild type TRKC (Figure 1e). To investigate whether the TEL-TRKC fusion protein contributes to the growth and viability of the M0-91 cells, the expression of the TEL-TRKC fusion was down regulated with an siRNA designed against the TRKC portion of the fusion protein. Western-blot analysis revealed that the expression of the TEL-TRKC fusion was specifically and significantly reduced at 24 and 48 h following transfection of the TRKC siRNA into M0-91 cells. This is accompanied by a decrease in the phosphorylation of previously known TRKC downstream effectors involving RAS-MAPK, PI3K-AKT and PLC pathways (AKT, ERK, PLC1 and PLC2) (Figure 2a). For the first time, we show that both STAT3 and STAT5 are tyrosine phosphorylated downstream to TEL-TRKC (Figure 2a), in contrasts to a previous report, where STATs were not activated in TEL-TRKC transformed BaF3 cells.8 Thus, activation of STATs by TEL-TRKC maybe cell-type specific. Down regulation of TRKC inhibits cell growth (Figure 2b), and treatment with TRKC siRNA also increases apoptosis of the M0-91 cell line as measured by a cleaved Caspase-3 flow cytometric assay (Figure 2c). Similar results were observed when M0-91 cells were treated with a relatively selective TRKC inhibitor AG879 (data not shown). These results suggest that the TEL-TRKC fusion is essential for the growth and survival of M0-91 cells. The siRNA knockdown experiments also revealed that phosphorylation of Shc, SHP-2 and SHIP1, as well as expression of both c-Myc and cyclin D1, is regulated by TEL-TRKC fusion (Figure 2d). In conclusion, we identified M0-91 as the first cell line expressing the TEL-TRKC fusion gene from its endogenous promoter, and therefore the first valuable cell culture model for the screening for TRKC inhibitors and for understanding both hematological and non-hematological diseases associated with TEL-TRKC. In addition, our phosphoproteomic analysis provided the most comprehensive tyrosine phosphorylation signaling profile reported for TEL-TRKC fusion to date. This information will be useful to develop novel therapeutic approaches for the effective treatment of human cancers associated with deregulated TRKC activity. Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)