Small molecules that affect the p53 pathway and their potential use in the treatment of cancer

Abstract

The tumor suppressor p53 was identified 35 years ago and has since then been studied extensively, but despite all efforts, no drug or therapy directly involving it has been clinically approved - yet! A lot of potential new drugs are on their way that can reactivate p53 function by various mechanisms. Even a whole new approach called cyclotherapy has been established, during which p53 is activated in normal cells to protect patients from the adverse effects of chemotherapy while tumor cells are still being killed efficiently. In this thesis, 16 drug combinations are being described in this context (paper I). Four individual p53-activating compounds, i.e. tenovin-6, leptomycin B (LMB), nutlin-3 and actinomycin D at low doses (LDactD), were used prior to the addition of each one clinically approved chemotherapeutic agent, i.e vinblastine, vinorelbine, cytosine arabinoside or gemcitabine. LDactD, which is clinically approved, showed the most promising results.

Unexpectedly, we identified two compounds that can inhibit p53’s ability to induce p21, i.e. the novel SirT2 inhibitor tenovin-D3 (paper II) and the widely used histone deacetylase inhibitor (HDACi) trichostatin A (TSA) (paper III). Inhibition of p21 in tumor cells might be desirable during cancer treatment to prevent tumor cells from undergoing cell cycle arrest, which would make them more vulnerable to classic chemotherapy. On the other hand, an inhibition of cell cycle arrest in normal cells might occur, which may worsen the side effects caused by chemotherapy. However, SirT2 plays a role in neurodegenerative diseases, and hence compounds like tenovin-D3 may be of use in the treatment thereof. Furthermore, the decrease in p21 levels may be a contributing factor in the previously observed increase in efficacy during the generation of induced pluripotent stem cells upon treatment with TSA; also tenovin-D3 could be useful in this context.

With the aid of a cell-based screen we identified two small molecules that can activate p53:

1) MJ05 was one of the most active hit compounds and was very selective (paper IV); it was highly cytotoxic in ARN8, especially when combined with nutlin-3, whereas it was cytostatic or had a very mild effect in other tumor cell lines and normal cells. It inhibited tumor growth in vivo, an effect that was enhanced upon co-treatment with nutlin-3. Furthermore, MJ05 selectively killed chronic myelogenous leukemia stem cells ex vivo while having milder effects in leukocyte stem cells derived from cord blood. Preliminary data strongly suggest that MJ05 acts by inhibition of pyrimidine (deoxy-) nucleotide synthesis.

2) Despite being a hit compound in our screen, MJ25 was not very potent at activating p53 (paper V). Nevertheless, its ability to inhibit thiredoxin reductase 1 (TrxR1) and its selectivity towards melanoma cell lines compared with normal cells were interesting features. We compared it with the TrxR1 inihibitor auranofin, which was very potent and selective at killing melanoma cells in cell viability assays. The insolubility of MJ25 at concentrations required for in vivo studies prevented us from testing it on xenografts in mice. Furthermore, MJ25 might not be specific for TrxR1, so the identification of additional targets could be investigated in the future. Auranofin, the other hand, has a more defined mechanism of action and is clinically approved for the treatment of rheumatoid arthritis. These traits combined with its potentially selective cytotoxic effect at low micromolar concentrations in melanoma cells may turn this compound into a potential drug candidate to be tested in patients suffering from malignant melanoma.

In the final study presented in this thesis (paper VI) we tested the small molecule tenovin-6 in zebrafish embryos The compound had been described previously by our group. The original aim of this study was to investigate if the activation of p53 in an organism could affect the ability of tumor cells to disseminate. Even though tenovin-6 did not activate wild-type p53 under the conditions tested, in vivo activity of the compound was still detectable, since embryos expressing mutant p53 (M214K) displayed an increase in p53 protein levels; furthermore, the compound was lethal in a dose- and time-dependent manner, and the embryos lost most of their brown/black pigmentation. The exact mechanism behind the latter observation could not be elucidated in the course of the project. However, tyrosinase, a key enzyme in melanogenesis, was not inhibited by tenovin-6, and the combination of data obtained by others on mutated or pharmacologically inhibited vacuolar H+-ATPase (V- ATPase) and yeast mutant strains suggested that the compound may target V-ATPase.