Portal de Periódicos da CAPES

A noncanonical FLT3 gatekeeper mutation disrupts gilteritinib binding and confers resistance

2021; Wiley; Volume: 96; Issue: 7 Linguagem: Inglês

10.1002/ajh.26174

1096-8652

Sunil K. Joshi, Setareh Sharzehi, Janét Pittsenbarger, Daniel Bottomly, Cristina E. Tognon, Shannon K. McWeeney, Brian J. Druker, Elie Traer,

Chronic Myeloid Leukemia Treatments

Acute myeloid leukemia (AML) is a genetically heterogenous disease with approximately 20 000 new cases per year in the United States.1 Patients with AML have a 5-year survival of <25%, and intense efforts are underway to develop new treatments to improve survival.2 Mutations in the FMS-like tyrosine kinase-3 (FLT3) gene are among the most common genomic aberrations in AML. Internal tandem duplication (ITD) in the juxtamembrane domain of FLT3 are present in approximately 20% of patients with AML. These mutations cause constitutive kinase activity, and lead to an increased risk of relapse and reduced survival. Another set of mutations in the tyrosine kinase domain (TKD) of FLT3 occur in 5–10% of AML patients and also result in activation of FLT3.3 Multiple FLT3 tyrosine kinase inhibitors have been developed and can be separated into two classes. Type I inhibitors are canonical ATP competitors that bind the ATP binding site of FLT3 in the active conformation and are effective against both ITD and TKD mutations. By contrast, type II inhibitors bind the hydrophobic region adjacent to the ATP binding domain in the inactive conformation. Type II inhibitors are effective against FLT3-ITD, but do not inhibit FLT3-TKD mutations. Quizartinib, a type II inhibitor, has potent activity against FLT3, KIT, and RET. Despite high response rates as a monotherapy in patients with relapsed/refractory AML, the duration of response to quizartinib is approximately 4 months, and resistance via FLT3-TKD mutations is common.4, 5 These mutations occur frequently at the activation loop residue D835 and less commonly at F691 which represents the "gatekeeper" position in FLT3.6 Gilteritinib is second-generation inhibitor that targets FLT3 and AXL.7 As a type I inhibitor, it is active against TKD mutations that impart quizartinib resistance. It was approved as monotherapy in relapsed/refractory patients with AML based upon the randomized phase 3 clinical study (ADMIRAL) which compared gilteritinib with chemotherapy.8 Despite the significant survival benefit in the gilteritinib arm, monotherapy is limited by the development of resistance, which typically occurs after 6–7 months. Resistance to gilteritinib most commonly occurs through acquisition/expansion of NRAS mutations, however a minority of patients with F691L gatekeeper mutations were also identified.7 To search for additional resistance mutations to gilteritinib, Tarver et al. used a well-established ENU mutagenesis assay and identified Y693C/N and G697S as mutations that confer resistance in vitro.5 These mutations appear to function similar to the gatekeeper mutation by blocking gilteritinib binding to FLT3, but have not been reported in patients. To more broadly investigate mechanisms of resistance to gilteritinib, we developed a two-step model of resistance that recapitulates the role of the marrow microenvironment (Figure 1A). In the first stage of resistance, or early resistance, the FLT3-mutated AML cell lines MOLM14 and MV4;11 are cultured with exogenous ligands, fibroblast growth factor 2 (FGF2) and FLT3 ligand (FL), that are normally supplied by marrow stromal cells. These culture conditions allow the cells to become resistant to gilteritinib without the need for resistance mutations.11 When ligands are removed, the cells regain sensitivity to gilteritinib, but ultimately become resistant, which we term late resistance. At this point, intrinsic resistance mutations were identified in all of the cultures via whole exome sequencing. Similar to clinical data,8 we found that the most common mutations are activating mutations in NRAS.12 One late resistant culture had an FLT3F691L gatekeeper mutation, and three cultures had an FLT3N701K mutation, which has not previously been reported (Figure 1B). Given its proximity to F691L (Figure 1C,D), we hypothesized that this mutation might also disrupt gilteritinib binding to FLT3. To determine whether the FLT3N701K mutation has oncogenic capacity, we evaluated this mutation in the Ba/F3 transformation assay. The Ba/F3 cells are normally IL-3 dependent but the presence of certain oncogenes transforms them to grow indefinitely in the absence of IL-3.13 The FLT3N701K mutation, similar to FLT3ITD and FLT3D835Y, is an activating mutation and promoted growth of Ba/F3 cells in the absence of IL-3, whereas the parental, empty vector, FLT3 wild type (FLT3WT), or FLT3F691L did not confer IL-3-independent growth (Figure 1E). In contrast to Ba/F3 cells expressing FLT3D835Y, Ba/F3 cells with FLT3N701K were much less sensitive to gilteritinib with an approximate 8.5-fold increase in IC50 (Figure 1F). To test whether FLT3N701K also promoted resistance to gilteritinib in the presence of FLT3ITD mutations (Figure 1B), we generated FLT3ITD+N701K and FLT3ITD+F691L double mutants and expressed them in Ba/F3 cells. Concordant with previous studies,6 the FLT3ITD+F691L mutant demonstrated an approximate 11-fold increase in IC50 to gilteritinib compared to FLT3-ITD alone. The FLT3ITD+N701K Ba/F3 cells were nearly identical to FLT3ITD+F691L cells in their resistance to gilteritinib (Figure 1G). As a control, FLT3WT Ba/F3 cells grown with IL-3 were insensitive to gilteritinib at comparable doses. Next, we assessed the impact of FLT3N701K mutations on downstream FLT3 signaling pathways. The Ba/F3 cells transformed with FLT3N701K, FLT3ITD, FLT3ITD+F691L, and FLT3ITD+N701K all resulted in phosphorylation of FLT3 (Y589/591) and STAT5 (Y694), AKT (S473), and ERK (T202/Y204) (Figure S1A). However, only FLT3ITD+N701K or FLT3ITD+F691L showed sustained phospho-FLT3 with increasing concentrations of gilteritinib (Figure 1H), indicating that both of these mutations prevent gilteritinib inhibition of FLT3, particularly at lower doses. The FLT3 kinase activity as reflected by FLT3 phosphorylation mirrored the viability assays in Figure 1G. Since F691L gatekeeper mutations are known to drive resistance to multiple FLT3 inhibitors,6, 7, 14, 15 we treated FLT3ITD, FLT3ITD+N701K and FLT3ITD+F691L Ba/F3 cells with midostaurin, crenolanib, and quizartinib. Although FLT3ITD+F691L and FLT3ITD+N701K were largely insensitive to type I inhibitors midostaurin and crenolanib, cells with FLT3ITD+N701K were notably more sensitive to the type II inhibitor quizartinib (Figure S2), suggesting that N701K blocks gilteritinib binding of type I inhibitors more effectively than type II. This was further apparent from our modeling of the FLT3N701K mutation. While the FLT3N701K mutation may sterically interfere with the binding of gilteritinib, quizartinib binding does not appear to be affected (Figure S3). Through our studies, we identified the novel FLT3N701K mutation in addition to the FLT3F691L gatekeeper mutation. We used the Ba/F3 system to demonstrate that N701K blocks gilteritinib binding to FLT3, similar to the gatekeeper F691L, and promotes resistance to gilteritinib. Our data fit nicely with recent data from a mutagenesis screen of Ba/F3 cells with FLT3-ITD that identified F691L in addition to D698N, G697S, and Y693C/N as mutations that drive resistance to gilteritinib.5 Modeling of these mutations indicates that they cause the loss of hydrogen bonding that accommodates the FLT3 side chain, leading to a steric clash between the tetrahydropyran ring of gilteritinib and FLT3.5 Given the proximity of N701K to these mutations, we speculate that the mechanism of resistance to gilteritinib imparted by this mutation is similar (Figure S3). Importantly, these complementary methods identify a common hotspot for gilteritinib resistance mutations (Figure S4). Given the increasing use of gilteritinib in the clinic, we anticipate that additional resistance mutations will likely be identified in patients. Of note, the N701K mutation appears to be more resistant to type I inhibitors but retains sensitivity to type II inhibitors such as quizartinib (Figure S2), implicating that TKI class switching could serve as a promising avenue to mitigate development of gilteritinib resistance. The use of type I FLT3 inhibitors following the acquisition of resistance to type II inhibitors is a well-established approach to overcome resistance. However, what makes the case with the N701K mutation interesting is acquired sensitivity to a type II inhibitor following development of resistance to a type I inhibitor, which is a largely underappreciated concept. This knowledge can be used to help rationally sequence FLT3 inhibitors upon development of resistance. We are extremely thankful to the Massively Parallel Sequencing Shared Resource for technical support. We thank Sudarshan Iyer for assistance with UCSF Chimera. This work was supported by the American Cancer Society (MRSG-17-040-01-LIB) to E.T., the Drug Sensitivity and Resistance Network (DRSN; U54CA224019) to B.J.D, S.K.M, and E.T, and Cancer Center Support grant (5P30CA069533-22) to B.J.D. Additional funding was provided by the Howard Hughes Medical Institute, and The Leukemia & Lymphoma Society to E.T. and B.J.D. S.K.J. is supported by the ARCS Scholar Foundation, The Paul & Daisy Soros Fellowship, and the National Cancer Institute (F30CA239335). E.T. is supported by grants from the Leukemia & Lymphoma Society, Hildegard Lamfrom Physician Scientist Award, American Cancer Society, and Cancer Early Detection Advanced Research Center. C.T., D.B., S.K.M., B.J.D, and E.T. were all supported by National Cancer Institute (U54CA224019). Study supervision: Elie Traer. Conception and design: Sunil K. Joshi, Elie Traer. Development of methodology: Sunil K. Joshi, Elie Traer. Acquisition of data: Sunil K. Joshi, Setareh Sharzehi, Janét Pittsenbarger. Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Sunil K. Joshi, Setareh Sharzehi, Janét Pittsenbarger, Daniel Bottomly, Cristina E. Tognon, Shannon K. McWeeney, Brian J. Druker, Elie Traer. Writing, review, & editing of the manuscript: Sunil K. Joshi, Janét Pittsenbarger, Setareh Sharzehi, Daniel Bottomly, Cristina E. Tognon, Shannon K. McWeeney, Brian J. Druker, Elie Traer. Development of prioritization framework to assist in rigor and reproducibility: Daniel Bottomly, Shannon K. McWeeney. B.J.D. potential competing interests – SAB: Aileron Therapeutics, Therapy Architects (ALLCRON), Cepheid, Vivid Biosciences, Celgene, RUNX1 Research Program, Novartis, Gilead Sciences (inactive), Monojul (inactive); SAB & Stock: Aptose Biosciences, Blueprint Medicines, EnLiven Therapeutics, Iterion Therapeutics, Third Coast Therapeutics, GRAIL (SAB inactive); Scientific Founder: MolecularMD (inactive, acquired by ICON); Board of Directors & Stock: Amgen; Vincera Pharma; Board of Directors: Burroughs Wellcome Fund, CureOne; Joint Steering Committee: Beat AML LLS; Founder: VB Therapeutics; Sponsored Research Agreement: EnLiven Therapeutics; Clinical Trial Funding: Novartis, Bristol-Myers Squibb, Pfizer; Royalties from Patent 6958335 (Novartis exclusive license) and OHSU and Dana-Farber Cancer Institute (one Merck exclusive license and one CytoImage, Inc. exclusive license). C.E.T. potential competing interests – Research funding from Ignyta (inactive). E.T. potential competing interests – Advisory Board/Consulting: Abbvie, Agios, Astellas, Daiichi-Sankyo, Clinical Trial Funding: Janssen, Incyte, LLS BeatAML. Stock options: Notable Labs. All other authors declare no potential competing interests. Figure S1. FLT3N701K activates FLT3 and results in increased downstream signaling. A. Expression of total and phosphorylated FLT3 is increased in mutant-transformed Ba/F3 cells relative to cells harboring empty vector. All mutants phosphorylate canonical downstream effectors – STAT5, AKT, and ERK. GAPDH served as a loading control. Prior to lysis, empty vector cells were grown in IL-3 supplemented media and all lines were starved overnight in 0.1% BSA RPMI. Figure S2. FLT3N701K is resistant to type I inhibitors, midostaurin and crenolanib but relatively sensitive to type II inhibitor, quizartinib. A-C. Six replicates of FLT3WT, FLT3ITD, FLT3ITD+F701K, and FLT3ITD+F691L Ba/F3 cells were plated with a dose gradient (0–1000 nM) of type I and II inhibitors, midostaurin (A, type I), crenolanib (B, type I), and quizartinib (C, type II) for 72 hrs. FLT3WT cells were plated in media supplemented with IL-3. Cell viability was determined using a tetrazolamine-based viability assay. Viability is represented as a percentage of the untreated control. The average mean ± SEM is shown. Figure S3. FLT3N701K sterically hinders binding of type I inhibitor gilteritinib (black). Binding of type II inhibitor quizartinib (green) is not affected by FLT3N701K. This modeling correlates with the drug sensitivity results presented in Supplemental Figure SS1. A. Front view. B. Side view. Ribbon diagrams were adapted from PDB 4RT71 (quizartinib) and PDB 6JQR2 (FLT3 and gilteritinib) and visualized with the UCSF Chimera software.3 Figure S4. FLT3N701K is found in a region where multiple noncanonical mutations that confer gilteritinib resistance reside. A. Gene schematic depicts location of all noncanonical gilteritinib-resistant mutations.4 B-C. Noncanonical mutations mapped onto the crystal structure of the FLT3 kinase domain. FLT3N701K is in close proximity to these mutations. Diagram was adapted from PDB 1RJB5 and visualized with the UCSF Chimera software.3 Table S1. List of all site-directed mutagenesis (a) and Sanger sequencing (b) primers used in this study. Table S2. List of all antibodies used in this study. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.