FLT3 Splice Variant (FLT3Va) As a Potential Immunotherapeutic Target in Patients with Acute Myeloid Leukemia (AML)

Alternative pre-mRNA splicing (AS) is a normal epigenetic phenomenon, a key regulator of gene expression, yields multiple transcripts and thus a variety of proteins from a single gene. Mutations in the spliceosome components resulting in aberrant splicing isoforms are common in AML, and other myeloid neoplasms, and may generate leukemia-specific neoantigens targetable with an antibody-drug conjugates (ADCs) or blocking antibodies.

Our previous studies revealed that the FLT3 cell surface receptor is one of the most commonly misspliced genes in AML (54-63% of ~400 AML patients). We conducted cloning and sequencing analyses in AML cells and identified multiple aberrant splice-variants of FLT3 that resulted from either skipping of one or more exons or activation of cryptic splicing sites. Transfection of cDNA with three of these variants in TF-1 (AML cell line) cells resulted in expression of Flt3 variant proteins on the cell surface. We successfully generated rabbit polyclonal antiserum against a unique peptide sequence present in the most commonly expressed abnormal splice variant, which we termed Flt3Va. Immunoblots performed with the polyclonal antibody identified a ~160 kDa protein expressed by TF-1 cells transfected with FLT3Va, and the antibody did not react with untransfected TF-1 cell lysate. Using standard techniques, we generated rabbit hybridomas and evaluated the clones by flow cytometry and western blotting experiments. Based on these data, we selected one antibody clone (15-7) for further experiments.

The 15-7 anti-Flt3Va rabbit monoclonal antibody identified Flt3Va protein expressed on the cell surface and within the cytoplasm of transfected TF-1 cells by flow cytometry and western blotting. However, no Flt3Va protein was detected in untransfected TF-1 cells or normal CD34+ bone marrow cells. The 15-7 antibody bound to 26 of 52 primary AML samples and 5 of 10 primagraft samples (PDX models) of human AML. Immunoblotting analyses of PDX models and patient samples confirmed binding to a protein of the expected size (130-160 kDa). Additionally, multi-parameter flow cytometry in 10 PDX models and 52 primary demonstrated that putative AML stem cells (as defined by the CD45dim, CD34, CD38, CD33, c-Kit cell surface expression) co-expressed Flt3Va antigen in 50% samples evaluated.

An analysis of Flt3Va protein localization by live cell imaging showed a punctate distribution of Flt3Va on the cell surface. Furthermore, we observed that overexpression of Flt3Va in TF-1 cells led to GM-CSF growth factor independence. Analysis of TF-1 cells in the absence of GM-CSF and Flt3 ligand demonstrated constitutive activation of STAT5, an important mediator of Flt3 signaling, in Flt3Va overexpressing cells. In addition, Erk1/2 phosphorylation was also increased in Flt3Va overexpressing cells, another downstream effector of Flt3.

In an effort to determine if Flt3Va+ cells had tumor repopulating ability, we sorted 0.3X10^6 Flt3Va+ and Flt3Va- cells from a PDX sample and injected the sorted populations or unsorted bulk tumor cells into NSG mice. The human cell engraftment in the mice was detected by the expression of human CD45, CD33, CD34, CD38, and c-kit antigens in the peripheral blood. In two experiments, mice injected with Flt3Va+ cells had detectable circulating leukemic cells by ~18 days after injection, while those injected with Flt3Va- cells had detectable circulating leukemic cells after the 4th week. These results suggest both Flt3Va+ and Flt3Va- cell populations are able to reconstitute leukemia after transplantation in NSG mice. However, Flt3Va+ may be expressed by an aggressive AML clone that facilitate early tumor engraftment.

Overall, these studies suggest that Flt3Va is a leukemia-specific neoantigen and is an attractive potential immunotherapeutic target in AML. Proteins such as Flt3Va generated by alternative splicing are common in AML and may be targets for of novel blocking antibodies or ADCs, minimizing effects on normal tissues.