Identification and characterization of small molecules targeting MYC function

Abstract: The MYC family genes (c-, N- and L-MYC) encode potent oncoproteins/transcription factors regulating fundamental cellular processes involved in cell proliferation, metabolism and survival, and they play an important role in tumor development. Overexpression of MYC often induce apoptosis as a failsafe mechanism to prevent tumor development and it is known to sensitize cells to genotoxic agents that induce DNA damage by triggered apoptosis. However, the MYC-regulated effectors acting upstream of the mitochondrial apoptotic pathway in response to DNA damage are still obscure. We focused on apoptosis activated by DNA damage responses in this study by comparing cell death induced upon ionizing radiation (IR), bacterial cytolethal distending toxin (CDT) and UV irradiation. We could demonstrate that phosphorylation of the ATM kinase and its downstream effectors, such as histone H2AX, were impaired in the MYC null cell line HO15.19, in comparison to the wild type parental cell line TGR-1 and MYC reconstituted HOMYC3 cells in response to IR or CDT. We also found that nuclear foci formation of the Nijmegen Breakage Syndrome (Nbs) 1 protein, which is essential for efficient ATM activation, was also reduced in the absence of MYC. Knocking down of endogenous MYC by siRNA in the HCT116 human colon cancer cell line resulted in decreased ATM and CHK2 phosphorylation in response to ionizing irradiation. However, the response to UV irradiation, which is known to activate the ATR dependent checkpoint, was functional in all of these cell lines, indicating MYC status did not play an important role in ATR signaling. In summary, we found that MYC is required for the activation of ATM- dependent checkpoint in response to IR and CDT; it contributes to DNA damage response by stimulating ATM phosphorylation and promoting NBS1 expression and nuclear translocation, thereby enhancing the apoptotic response, but potentially also stimulating DNA repair. Deregulated MYC expression is implicated in the development of a wide variety of cancers and is often strongly correlated with poor prognosis, underscoring the importance of finding ways to counteract MYC function. To exert its oncogenic activity, MYC must be able to interact with a number of cofactors that are essential for MYC function. For instance, dimerization with the partner Max enables the MYC to bind target gene promoters. This requirement for cofactors may allow for control of MYC activity with small molecules that interfere with interactions with these factors. We used bimolecular fluorescence complementation (BiFC) assay to visualize interactions between MYC and cofactors in living cells. Using BiFC we screened a 2000 compound library for molecules inhibiting the interaction between MYC and MAX, and found several interesting compounds. MYCMI-6 emerged among the top hits, and was further validated by split Gaussia luciferase (Gluc), in situ proximity ligation (isPLA), microscale thermophoresis (MST) and surface plasmon resonance (SPR) assays and was found to be a strong selective inhibitor of MYC:MAX interaction in cells and in vitro at single-digit micromolar concentrations without affecting MYC expression. SPR showed that MYCMI-6 binds to the recombinant MYC bHLHZip domain with a KD of 1.6 ± 0.5 μM. MYCMI-6 downregulated MYC-driven transcription and inhibited tumor cell proliferation and viability in a MYC-dependent manner in the low micromolar range, but was not cytotoxic to normal cells. In vivo studies using a xenograft mouse model of MYCN-amplified neuroblastoma revealed that daily intraperitoneal injections of MYCMI-6 led to reduced tumor cell proliferation, reduced microvascular density and induced massive apoptosis in tumor tissue without causing severe side effects for the mice. MYCMI-7 is another of the top hits identified in the BiFC screening. The efficacy and selectivity of MYCMI-7 was further validated with respect to inhibition MYC:MAX interaction, binding to MYC and effects MYC-driven tumor cell growth. Using a number of protein interactions assays which could demonstrate that MYCMI-7 efficiently blocks MYC:MAX interaction both in cells and in vitro. Using MST and SPR we showed that MYCMI-7 binds to recombinant MYC with an affinity of approximately 4 M. In contrast to MYCMI-6, MYCMI-7 downregulated the steady state levels of MYC protein subsequent to the inhibition of MYC:MAX interaction, suggesting that it could inhibit MYC in both direct and indirect ways. MYCMI-7 strongly inhibited tumor cell growth and induced apoptosis in a MYC dependent manner in a number of different tumor cell lines such as neuroblastoma, glioblastoma, Burkitt’s lymphoma, AML, lung cancer and several other epithelial tumors as well as patient-derived AML and glioblastoma tumor samples, while it only causing G1 arrest with cytotoxicity in normal cells. Moreover, MYCMI-7 blocked transformation of primary rat embryo fibroblasts by MYC together with activated RAS. Importantly, treatment with MYCMI-7 in vivo inhibited tumor growth and prolonged survival in mouse models of MYC-driven acute leukemia, triple negative breast cancer and MYCN-amplified neuroblastoma. Besides MYCMI-6 and MYCMI-7, yet another compound MYCMI-2 was identified from the BiFC screening. MYCMI-2 exhibited outstanding specific inhibition of heterodimerization of in vitro translated or recombinant MYC and MAX in vitro as determined by split GLuc, SPR and FRET assays, the latter showing an IC50 of 150 nM. Further, MYCMI-2 bound to MYC with extraordinary high affinity (KD 1.3 nM) as determined by SPR. We utilized cell based Gluc and isPLA to validate MYCMI-2’s MYC:MAX inhibitory efficacy in cells. The latter assay demonstrated an IC50 of about 5 μM in MCF7 cells. Further, MYCMI-2 inhibited MYC-driven tumor cell growth and viability in a MYC-dependent manner, in a number of tumor cell lines with an IC50 of 1.5-6 μM, while viability of normal cells was not affected. Due to the difference between MYCMI-2’s extraordinary activity in vitro and limited efficacy in cell cultures, we attempted to identify analogues with improved efficacy in cells with maintained activity in vitro. The analogue molecule MYCMI-2:7 showed lower but acceptable potency and maintained selectively towards MYC:MAX heterodimerization in vitro compared with MYCMI-2, but demonstrated slightly better MYC:MAX inhibitory effect in the cell based Gluc and the isPLA assays with inhibition down to about 40% of DMSO treatment. Unlike MYCMI-2, MYCMI-2:7 downregulated the endogenous MYC protein level in MCF7 cells, which indicated that MYCMI-2 and MYCMI-2:7 work via different mechanism. In conclusion, we have demonstrated that MYCMI-2 has an extraordinary potency in vitro binding at a KD of 1 nM, and an activity in cells in the lower μM range, while the analogue MYCMI-2:7 was less active in vitro and only marginally better in cells. Both molecules could potentially contribute to the development of bioactive MYC inhibitors of therapeutic interest in cancer therapy in the future. We now try to improve the bioactivity of these compounds by modifications, evaluate their efficacies in vivo, and to further elucidate their mode of action and selectivity. Our protein-protein interaction platforms could potentially be used for high-throughput screens of larger chemical libraries for inhibitors of interactions between MYC and MAX as well as other cofactors of therapeutic and biological interest. We hope this project will lead to better understanding of the biological functions of the MYC network during tumorigenesis and provide new therapeutic tools to combat cancer in the future.