IGF1R-IRS1/2 Pharmacological Inhibitors Act By Distinct Cellular and Molecular Mechanisms and Reveals Vulnerabilities for Treatment of Acute Myeloid Leukemia

Background: Preclinical rationale for targeting the insulin-like growth factor 1 (IGF1R)-Insulin Receptor Substrates 1 and 2 (IRS1/2) signaling in acute myeloid leukemia (AML), particularly in cells harboring the FLT3-ITD mutation, has been recently provided [Blood (2018) 132 Supp: 1512 and [Blood (2019) 134 Supp: 2689]. However, little is known about the non-canonical molecular mechanisms regulated by IGF1R-IRS1/2 signaling and pharmacological inhibition of this pathway in AML. Aims: To depict distinctive non-explicit molecular effects of linsitinib (IGF1R tyrosine kinase inhibitor) and NT157 (IGF1R-IRS1/2 allosteric inhibitor) treatment in FLT3-ITD-mutated AML cells. Material and methods: The MOLM-13 (homozygous) and MV4-11 (heterozygous) FLT3-ITD-mutated AML cell lines were treated with linsitinib (10 µM) or NT157 (1 µM) for 24 hours and used for label-free proteomic quantification analysis (n=3). Raw MS/MS data were processed using the SORCERER system and proteins were identified with built-in Andromeda search engine based on the human Uniprot proteome database. False discovery rate cutoffs were set to 1% on peptide, protein, and site decoy level, only allowing high quality identification to pass. Expression values were normalized across experimental conditions by quantile normalization based on the Limma-Voom pipeline, and then systematically compared similarities and differences in protein expression across experimental conditions by applying the Benjamin-Hochberg correction for multiple comparisons. To depict pathways associated to IGF1R, IRS1 and IRS2 gene expression related to processes identified by the proteomic data, we performed a gene-set enrichment analysis (GSEA) using the curated genesets for oncogenic events and molecular functions (MSigDB, Broad Institute) from RNA-seq data of the Cancer Genome Atlas AML cohort (n=173). Results: Considering a ≥ 2-fold change difference in both directions, linsitinib treatment downregulated 6 and 18 and upregulated 13 and 116 proteins in MOLM-13 and MV4-11 cells, respectively. Likewise, NT157 downregulated 12 and 126 and upregulated 204 and 297 proteins. When compared directly, linsitinib reduced expression of 11 and 35 and increased expression of 110 and 70 proteins in MOLM-13 and MV4-11 cells, respectively. Gene ontology identified that linsitinib resulted in upregulation of 7 molecular functions, while the NT157 ensued the upregulation of 18 and downregulation of 17 molecular functions pathways in a consistently manner between all comparisons. Of note, linsitinib activates post-transcriptional regulatory mechanisms, RNA metabolism (RNA binding P=1.15E-12; RNA processing P=8.64E-7) and reduced the protein and macromolecule metabolism (cellular protein metabolism P=3.86E-6). NT157 affected several of mitochondrial functions (increasing proton transmembrane transport activity P=1.55E-12, reducing oxidoreductase activity P=9.11E-10, and oxidative phosphorylation P=5.19E-8). Altogether, these data highlighted that NT157 profounder cytotoxic effect is a result of reprogramming of cellular energetics metabolism, and that linsitinib altered transcription and translation processes, probably as a result of autophagy, a mechanism originally described by our group [Blood (2017) 130 Supp: 3966]. GSEA analysis revealed that high IGF1R expression is positively enriched with RPS14 signature (Normalized Enriched Score [NES]=2.23; FDR-q<0.001), a ribosomal protein related to pathophysiology of myeloid neoplasms related to chromosome 5q deletion. Both IRS1 and IRS2 transcriptional signatures were associated with cellular growth signaling, such as AKT (NES=1.86; FDR-q= 0.006) and MYC (NES=1.67; FDR-q= 0.005), mitochondrial function [mitochondrial gene expression (NES=1.71; FDR-q= 0.001)]. Conclusion: Our proteomic data shed light on new and non-explicit mechanisms related to IGF1R-IRS1/2 inhibitors. Linsitinib modulates molecular processes related to RNA transcription and translation, while NT157 profoundly affect the cellular energetics, and, at least in part, explain the differential pre-clinical efficiency. Moreover, allosteric pharmacological inhibition of IGF1R-IRS1/2 pathway seems a more promising strategy than the tyrosine kinase inhibition, especially for AML subgroup more dependent of mitochondrial metabolism, such as AML with FLT3 mutation.