Pro-senescence therapy by targeting MYC and its network as a strategy to combat cancer

Abstract

Cancer is defined as abnormal cell proliferation with the potential of metastasis and is an increasing threat to global health as the population becomes older. Cancer is triggered by gain-of-function mutations in oncogenes and loss of function mutations in tumor suppressor genes. This leads to the breakdown of two important barriers against tumorigenesis-cellular senescence, which is permanent cell cycle arrest, and apoptosis, which is a programmed cell death. MYC is one of the classical oncogenes and its deregulation is observed in various types of cancer where it plays an essential role in tumor development. It has been more than three decades since MYC was discovered; however, the precise role of MYC in normal cells and in different types of cancer is still under intense research and debate. For example, it has been shown that MYC can overcome senescence induced by another oncogene, RAS, and RAS can override apoptosis triggered by MYC, leading to oncogenic transformation of primary rodent cells. However, there is no evidence indicating to which extent, this cooperation between MYC and RAS occurs also in human primary cells, since transformation of human cells requires several steps in addition to overcoming senescence and apoptosis.

Therefore, in Paper I, to investigate cooperation between MYC and RAS in human primary cells, we used human BJ fibroblast cells where we established a double inducible RAS and MYC system (BJRAS-MYC). In this system RAS expression is induced by doxycycline while MYC is expressed as a MycER fusion protein controlled by 4-hydroxytamoxifen (OHT) treatment. As expected, RAS induction triggered cellular senescence, as evidenced by increased senescence-associated β-galactosidase (SA-β-GAL) activity, enlarged cell size, arrested cell proliferation, increased expression of histone H3 lysine 9 trimethylation (H3K9me3), 21 and p16, and decreased phosphorylation of pRB. MYC induction preferentially triggered apoptosis, measured as increased cell death and cleaved PARP. Dual induction of RAS and MYC decreased H3K9me3 and SA-β-GAL but enhanced p16 level compared to RAS induction alone, suggesting MYC failed to completely overcome RAS-induced senescence. On the other hand, MYC-trigged cell death was not rescued by RAS induction, even when MYC activity was tuned down or MYC activation was scheduled at different times after RAS induction. Thus, MYC and RAS failed to cooperate to overcome each other’s fail-safe mechanisms, in which the main obstacle seemed to be MYC-induced cell death. Given the importance of p53 in regulation of apoptosis and that p53 was triggered upon MYC induction in our system, we knocked down p53 to try to rescue cells from cell death. Unexpectedly, p53 depletion (BJ-RAS-MYC-shp53) failed to overcome MYC-induced cell death although it rescued RAS-induced senescence. In conclusion, MYC and RAS did not cooperate to abrogate apoptosis and senescence, respectively, in human primary cells, even after p53 depletion, suggesting that additional oncogenic events are required to overcome senescence and apoptosis in this system.

In has been reported that cancer treatments such as chemotherapy and radiation can trigger senescence in tumor cells. Hence, the concept of pro-senescence therapy has emerged as an alternative of anti-cancer therapy, especially for tumors resistant to apoptosis-inducing therapies or where conventional therapies cause severe side-effects. In Paper II, we explored the potential of prosenescence therapy in melanoma treatment. First we investigated the efficacy of eleven clinical and preclinical drugs as monotherapy in a senescence screen, using a panel of melanoma cell lines with different driver or mutations, such as BRAFV600E, NRASQ61R, PTEN, PI3K, p53 and CDK4. Following treatment, cell number, nuclear/cell size, EdU incorporation, intensity of H3K9me3, p53 as well as HLA class I staining was measured using an Olympus scanR high-content imager system. We found that vemurafenib and trametinib induced senescence in some but not all BRAF-mutated cell lines, while palbociclib, crizotinib and BKM120 induced senescence most of cell lines irrespective of BRAF/NRAS mutation status. These results were confirmed using a additional senescence markers, such as analysis of SA-β-GAL activity, cell cycle distribution and phosphorylation of pRb. Moreover, palbociclib and BKM120 also increased expression of several SASP factors and palbociclib and crizotinib enhanced expression of HLA class I, implying the potential of increased immunogenicity.

We next addressed the question whether combination treatment could synergize with vemurafenib to overcome intrinsic and acquired vemurafenib resistance. Therefore, we performed a combination treatment senescence screen, combining vemurafenib and trametinib with the other drugs. The results showed that palbociclib, crizotinib and BKM120 synergize with vemurafenib to trigger senescence in both vemurafenib-sensitive and –resistant cells. In addition, we found that the combination of palbociclib with crizotinib further enhanced senescence including SASP factors of relevance for immune responses, again independent of BRAF/NRAS mutation status. We also looked at expression of molecules associated with recognition by T and NK cells. Palbociclib increased the expression of HLA class I and HLA class II, but also of PD-L1 and the NKG2D ligands ULBP 2/5/6 in several melanoma cell lines, and the expression of these molecules were in some cases further enhanced by combining palbociclib with crizotinib. This indicates that palbociclib induced both positive and negative immune markers for CD8+ T cell and NK cell recognition in melanoma cells, implying possible benefit of using anti-PD-1 combination treatment to eliminate such cancer cells.

Although it is clear that MYC is one of the most important players in tumorigenesis, there are, unfortunately, no selective MYC inhibitors available in clinic today. The main reason for this is that transcription factors like MYC are difficult to target due to their lack enzymatic activity and intrinsically disordered features. Since MYC function is dependent on interaction with its partner MAX, this motivates us to try to identify and characterize MYC inhibitors directly targeting the MYC:MAX interaction. Thus, in Paper III and Paper IV, we used a cell-based bimolecular fluorescence complementation (BiFC) screen, where MYC and MAX were fused with two parts of yellow fluorescent protein (YFP), only the close proximity of MYC and MAX can restore the fluorescent signal. We found six compounds (MYCMI-2, MYCMI-6, MYCMI-7, MYCMI-9, MYCMI-11 and MYCMI-14) potentially targeting the MYC-MAX interaction. This was validated in cells using the following assays: 1) split Gaussia luciferase (GLuc), where MYC and MAX are fused with two complementary fragments of GLuc; 2) in situ proximity ligation assay (isPLA), in which after primary antibodies binding, secondary antibodies conjugated with oligo were bound to the primary antibodies, and the close proximity assisted oligos ligation and followed by PCR, fluorescent probe to PCR products can be visualized; 3) U2OS-MycER system showed us that these compounds repressed MYC target genes. Among the six compounds, MYCMI-6 and MYCMI-7 exerted the highest efficacies. We next investigated whether these molecules directly target MYC:MAX interaction in vitro using recombinant proteins in microscale thermophoresis (MST) assay, which is based on changes in the mobility of a fluorescent molecule in a temperature gradient upon binding of ligands/inhibitors). We also used surface plasmon resonance (SPR), which is a highly sensitive assay to determine the affinity between protein and ligand and to measure the kinetics of interactions. With both these assays, MYCMI-6 and MYCMI-7 were confirmed to inhibit MYC: MAX interaction via directly binding to MYC bHLHZip domain in the low micromolar range, which is more potent than previously reported MYC:MAX inhibitors. A difference between MYCMI-6 and MYCMI-7 was that the former did not affect MYC expression, while MYCMI-7 decreased MYC protein, but not mRNA, level. The mechanism behind this is not clear. Furthermore, both MYCMI-6 and MYCMI-7 inhibited growth and induced apoptosis in a number of MYCdependent tumor cell lines, such as MYCN-amplified neuroblastoma and MYC-driven Burkitt’s lymphoma. MYCMI-6 reduced tumor cell proliferation in a mouse xenograft model of MYCNamplified neuroblastoma. In addition, MYCMI-7 reduced tumor burden in mouse models of MYCdriven acute leukemia, triple negative breast cancer and MYCN-amplified neuroblastoma, and prolonged survival in the latter two models.

It has been reported previously that MYC depletion by shRNA induced cellular senescence in melanoma with BRAF or NRAS mutation, suggesting that MYC inhibition could be used in prosenescence therapy. In Paper V, we evaluated the senescence-inducing potential of the MYC inhibitors MYCMI-6 and MYCMI-7, reported in Paper III and Paper IV, as well as the previously reported MYC:MAX inhibitor 10058-F4 (F4) and one BET inhibitor JQ1, which represses MYC transcription, in a panel of melanoma cell lines. In a senescence screen performed as in Paper II, MYCMI-7 triggered senescence in the majority of cell lines, and was more potent than F4. This was further validated in selected melanoma cells with BRAFV600E mutation, and the effects of F4, MYCMI-7 and JQ1 was compared. JQ1 mainly induced senescence; MYCMI-7 a mix of apoptosis and senescence and F4 preferentially triggered apoptosis. Among the three MYC inhibitors, only MYCMI-7 exerted a persistent proliferation inhibitory effect after drug removal. We also performed a senescence screen for the combination of the MYC inhibitors and the BRAFV600E inhibitor vemurafenib, MEK inhibitors (trametinib/selumetinib) and the CDK4/6 inhibitor palbociclib. The combination treatment of MYC inhibitors with vemurafenib/trametinib/selumetinib synergistically induced senescence and overcame vemurafenib/trametinib/selumetinib resistance. In addition, MYC inhibitors synergized with palbociclib to induce senescence in BRAFV600E-mutated cells. Thus, combining MYC inhibitors with BRAF, MEK and CDK4/6 inhibitors synergistically induced malignant melanoma senescence. The data presented in this thesis has shed further light on the function of MYC in normal and tumor cells, but also opening up the possibility of targeting MYC in pro-senescence therapy.