MDM2 antagonist Nutlin-3a potentiates antitumour activity of cytotoxic drugs in sarcoma cell lines

The TP53 gene, coding for the transcription factor p53, is thought to be the most frequently mutated gene in cancer, inactivated in about 50% of all tumours. However, aberrations of this pathway are probably even more wide-spread, as tumours retaining wild-type p53 (TP53 ^Wt ) might have defects in other parts of the pathway [1]. In sarcomas, malignant tumours resembling mesenchymal tissue, amplification of MDM2 (murine double minute 2) is relatively common (20%) in tumours having TP53 ^Wt , resulting in disabled p53 function because overexpressed MDM2 protein binds to and inactivates p53 [2, 3]. Remaining tumours may have other aberrations in their p53 pathway, either p53 mutations (TP53 ^Mut , 11-31% depending on subtype), or other changes in the downstream pathway that do not affect the level of MDM2 (MDM2 ^Wt /TP53 ^Wt , 11-88% depending on subtype) [4‐6].

Sarcomas are among the more frequent cancers among children [7], and both children and adults are treated with intensive surgery, chemotherapy or radiation, or a combination of these. Currently used chemotherapy (e.g. Methotrexate, Cisplatin and Doxorubicin) is frequently inadequate, with 50-80% long-term survival depending on tumour subgroup [8, 9] and associated with severe toxicity. Due to the frequent failure of prevailing therapy and unacceptable adverse effects there is an urgent need for new therapeutic modalities in sarcoma.

Nutlin-3a, a novel small-molecule inhibitor of the p53-MDM2 interaction, has been shown to be highly effective in killing osteosarcoma cells in vitro and reducing tumour burden in vivo [10, 11]. Nutlin-3a displaces p53 from the binding pocket of MDM2 and thereby releases p53 from inhibition and proteasomal degradation, leading to induction of its downstream targets, cell cycle arrest, and apoptosis. Tumours with amplification of the MDM2 gene (MDM2 ^Ampl ) are most responsive to Nutlin, most likely due to otherwise intact downstream p53 signalling [11]. However, the apoptotic response in cancer cells with normal levels of MDM2 can vary dramatically, suggesting that other mechanisms or modifying factors are involved in the response to MDM2 antagonists. Several studies have suggested both p53-related and unrelated factors, such as caspases, BAX, PUMA, p73 and other apoptotic factors to be involved [10, 12‐14], and the status of the related MDM4 (also known as MDMX) gene has also been proposed to explain the responses to Nutlin [15‐17].

Early studies suggest that MDM2 antagonists may be particularly effective in sarcomas because MDM2 is frequently amplified in these tumours [2, 5, 18, 19]. We previously confirmed the effectiveness of Nutlin-3a as a single agent in TP53 ^Wt /MDM2 ^Ampl liposarcomas [20]. However, since many sarcomas are TP53 ^Wt but do not have amplified MDM2 (are MDM2 ^Wt ), it would be of interest to investigate if Nutlin-3a could potentiate the response of MDM2 ^Wt /TP53 ^Wt tumours to conventional chemotherapy. Since p53 mutations are very heterogeneous, different mutations sites could also imply different response to combined therapy. Nutlin has been shown to be synergistic with genotoxic drugs (e.g. Fludarabine, Chlorambucil, Doxorubicin, Etoposide, Melphalan and Cisplatin) in haematological malignancies, lymphoma, neuroblastoma and hepatocellular carcinoma, and with radiation in lung cancer [12, 21‐26], but antagonistic with antimitotic agents (e.g. Paclitaxel) in colon cancer cell lines [27]. Paclitaxel was shown to be synergistic in rhabdomyosarcoma cell lines [28]. Here, we investigate the effect of Nutlin-3a on sarcoma cell lines in combination with the current standard of therapy (e.g. Doxorubicin, Cisplatin and Methotrexate). These are well known and frequently used genotoxic drugs that induce cell cycle arrest and apoptosis through both p53-dependent and-independent mechanisms [29‐31]. The latter group is represented by Methotrexate, which primarily inhibits dihydrofolate reductase (DHFR), but also glycinamide ribonucleotide formyltransferase (GARFT) and thymidylate synthetase (TS), all key components of nucleotide biosynthesis. Our studies show significant potentiation and/or reduction of effective dose of cytotoxic drugs by Nutlin, in both wild-type and mutated TP53 tumours, suggesting that clinical combination studies in sarcoma are warranted.

We investigated the effect of combining Nutlin-3a with the current standard of sarcoma therapy (Doxorubicin, Cisplatin, Methotrexate) on the growth of five sarcoma cell lines. The characteristics and sources of the cell lines used are described in Table 1. From our previous results we knew that the liposarcoma cell line T778 was sensitive to Nutlin-3a and we included the osteosarcoma cell line OSA with similar characteristics (MDM2 ^Ampl /TP53 ^Wt ) as a control cell line. As representative of MDM2 ^Wt /TP53 ^Wt , we chose the osteosarcoma cell line U2OS. Finally, SaOS-2 and RMS13 were chosen as examples of two different cell lines with mutated p53. SaOS-2 is known to be TP53 ^Null [38]. RMS13, previously described as TP53 ^Wt , did not respond to Nutlin-3a, and was found to have a transversion in codon 280, exon 8, changing the amino acid from arginine to serine in the resulting protein. Protein analysis uncovered high levels of probably dysfunctional p53 protein and constitutively high p21 and PUMA (see below). This is consistent with a report that the Rh30 cell line, established independently from a xenograft from the same patient, also had this mutation [39] and illustrates the usefulness of Nutlin as a probe for p53 status. The other cell lines with TP53 ^Wt were validated by sequencing.

Current treatment options for sarcomas, based on decades of optimising high dose chemotherapy, surgery and radiation, have reached a plateau with long-term survival limited to 50-80% depending on tumour subgroup [8, 9]. Although some new targeted treatments have been successful in other cancer types and also in some sarcoma subtypes (e.g. GIST) [40], these treatments rarely increase long-term survival. Since sarcomas are among the most frequent childhood cancers, the possibility of reducing the long-term adverse effects of high-dose chemotherapy is an important objective. Thus, new treatments targeting key cancer pathways are highly desirable. One such pathway is the p53 tumour suppressor pathway.

Although TP53 mutations are not so frequent in sarcomas [41], the pathway may also be inactivated by amplification of the MDM2 gene [2, 5, 18] which is observed almost exclusively in tumours where TP53 is wild-type. However, these two mechanisms do not account for all tumours, and although some complementary mechanisms have been suggested [6], it is still unclear if and how the pathway is inactivated in the remainder of sarcomas. If tumours contain defects of the p53 pathway that are upstream of p53 or suppress p53 levels, activation of p53 could bypass the defect and, provided the right signals are present, either induce apoptosis directly or sensitize the cells to induction of apoptosis by traditional therapy.

Small-molecule MDM2 antagonists such as Nutlin-3a are interesting in this aspect not only as therapeutic agents, but as tools to study p53 pathway functionality [11]. Nutlins efficiently activate p53 signalling and induce cell cycle arrest and apoptosis in sensitive cancer cells, while sparing most normal cells when tested in vitro [10‐13, 15, 23, 27, 42]. However, their safety in humans is still unknown, especially when combined with traditional cytotoxic therapy. Although it has been shown that Nutlin alone is very efficacious in MDM2 ^Ampl sarcoma cell lines in vitro and in vivo [11, 20], the effects of combinations with standard therapeutics in sarcomas is still unknown. The aim of this study was to investigate possible synergistic interactions between traditional sarcoma drugs and Nutlin-3a.

As expected, cell lines with wild-type TP53 responded to Nutlin-3a exposure by activation of p53 and its downstream targets, including p21 and MDM2, whereas tumours with mutated TP53, irrespective of mutation site, did not respond to Nutlin-3a alone. MDM2 ^Ampl cell lines were more sensitive to Nutlin-3a compared to the MDM2 ^Wt cell lines in spite of similar levels of induction of p21, MDM2 and PUMA. This is probably the result of other yet undiscovered defects in the downstream p53 signalling in the MDM2 ^Wt cells. MDM4 is an MDM2 homologue that shares some of the same p53 binding sites as MDM2, although there are some differences [43]. It has been suggested that high levels of MDM4 may counteract Nutlin because its binding to p53 is not inhibited by the treatment [15, 16, 44], although the hypothesis is still controversial [45].

The genotoxic agents tested in this study are commonly used in sarcoma treatment. Doxorubicin and Methotrexate were quite effective in these cell lines, whereas higher doses of Cisplatin and Nutlin-3a were required in the TP53 ^Wt lines. Neither Doxorubicin nor Cisplatin showed consistent differences in effect in relation to MDM2 amplification or TP53 mutation, but the TP53 mutated cell line (RMS13) was more sensitive to all three genotoxic agents tested than the TP53 null cell line (SaOS-2). Methotrexate was more effective in the lines with MDM2 amplification, suggesting some relation between MDM2 and the Methotrexate mechanism of action that warrants further investigation.

The DRI values clearly illustrated the benefit of the synergistic combinations, with 10-fold reductions in Doxorubicin and Cisplatin when combined with Nutlin-3a, but also combinations with less than additive effects showed promising DRI values. Nutlin-3a and Doxorubicin or Cisplatin were less than additive in the MDM2 ^Wt/TP53 ^Wt U2OS cell line. This is consistent with previous results in colon (HCT116) and ovarian (A2780) cancer cells, where Nutlin was cytoprotective in combination with Cisplatin [14]. Compellingly the TP53 ^Mut cell lines, RMS13 and SaOS-2, showed a significant sensitization towards Cisplatin and Doxorubicin at IC50 when combined with Nutlin-3a. This apparently p53-independent effect of Nutlin-3a is in line with previous publications [46‐48], and suggests efficacy against a wide spectrum of tumours. p73 has been proposed to be one of the factors responsible for the effect observed in TP53-null or-mutant cells. It was reported that when cells with dysfunctional TP53 were exposed to high concentrations of Nutlin (e.g. 20-30 μM), p73 was displaced from its binding to MDM2, resulting in cell cycle arrest and apoptosis [25, 46, 49]. Nutlin combined with Cisplatin or Doxorubicin has also previously been reported to induce apoptosis much more efficiently than either drug alone in cells with dysfunctional TP53, most likely due to E2F1-mediated up-regulation of p73 activity [48].

The third chemotherapy tested here, Methotrexate, acts through nucleotide metabolism, for which E2F1 is a central transcription factor [50, 51]. Additional targeting of the p53 pathway by Nutlin-3a could be complementary to this mode of action [52]. We found synergistic effects for both TP53 ^Wt /MDM2 ^Ampl and TP53 ^Wt /MDM2 ^Wt cell lines and DRI values that implied that Methotrexate could be reduced ten-fold when combined with Nutlin-3a. In the MDM2 ^Wt cell line U2OS, treatment with Methotrexate decreased MDM4 to undetectable levels both alone and in combination with Nutlin-3a. According to the current hypothesis, reduction of MDM4 could lead to more unbound p53 [53], but since this reduction also happened with Methotrexate only, it would not contribute to the observed synergy. On the other hand, Nutlin-3a can indirectly reduce the level of E2F1 through MDM2 mediated degradation [48], and recent results also document the reduction of DHFR through MDM2 [54]. It is tempting therefore to speculate on a double inhibition of the nucleotide metabolism when using Methotrexate together with Nutlin-3a. The inhibitory effect observed in the TP53 ^Mut cell line (RMS13), could not be explained by this and clearly needs further investigation.

Our in vitro data suggest that MDM2 antagonists could offer effective therapy for sarcoma either as single agent or in combination with prevailing chemotherapy, and may significantly reduce the genotoxic burden with the same or better antitumour effect. Further work with animal models of sarcoma and ultimately clinical studies needs to be done to evaluate and optimize combinations of Nutlin and current chemotherapies. Since clear biomarkers are still lacking, clinical trials should be followed up with detailed tumour profiling.